Dew-point margin: the metric that prevents humidor mold

TL;DR. Mold doesn’t grow because your RH is wrong. It grows because something inside the humidor crossed its dew point — a cool corner near the cooling element, a drip on the inside of a wineador door, the back wall of a tower cabinet on a January morning. Dew-point margin (the number of degrees between current temperature and the dew point) is the metric that catches that before the mold does.

The first cigar I ever lost to mold was perfectly stored. April, three years back. Cohiba Siglo VI, bottom shelf of the wineador. 65% RH and 19°C, every reading honest, every metric green for the entire prior week. I picked it up to smoke after dinner and felt the foot. A flare, barely a millimeter, bell-shaped — the unmistakable bloom where a single drop of water had landed on the wrapper and pushed the leaf outward. The cigar on the shelf above it, same shipment, had a faint dusting of green at the foot. I almost didn’t notice.

The dashboard hadn’t been lying to me. It was just missing the one number that mattered.

So this is the post I’d write to my past self. It’s about dew point, the number your hygrometer doesn’t show you, and dew-point margin, the buffer between “your humidor is fine” and “you’ve got 48 hours before things get visible.”

What dew point actually is

Air can hold a temperature-dependent maximum of water vapor. We covered that in the VPD post. Dew point is the inverse question: given the water currently in the air, at what temperature would that air become saturated?^[1]

Cool the air below that temperature and the excess water vapor has nowhere to go. It condenses. Inside a wineador door. On the metal hinge of a wooden humidor. On the cooler back wall of a tower cabinet. On the foot of a Cohiba sitting too close to the spot where the compressor cycles cold.

Mathematically, dew point is recovered from temperature and RH via the Magnus-Tetens approximation^[2]:

γ = ln(RH/100) + (17.625 · T) / (243.04 + T)

T_d = (243.04 · γ) / (17.625 − γ)

You don’t have to compute this. The Cabinet does. What you need to remember is the rough, useful approximation that survives most discussions among working farmers, brewers, and wineador owners:

Td ≈ T − (100 − RH) / 5

At 19°C and 65% RH, dew point is roughly 12°C. At 22°C and 70% RH, it’s roughly 16°C. At 25°C and 75% RH, it’s roughly 20°C, and a small temperature dip to 19°C is enough to bring you to saturation. That is when mold gets its day.

The margin, and what 48 hours costs you

Dew-point margin is the simple subtraction:

margin = T − T_d

It’s how many degrees the air would have to cool before it starts dropping water onto your cigars. Bigger margin, safer. Smaller margin, condensation risk.

The EPA’s residential mold guidance rests on the physics every cigar smoker eventually learns the hard way: under the right conditions, visible mold appears within 24 to 48 hours of a moisture intrusion^[5]. Tobacco is more generous to mold spores than drywall is. A single drip of condensation on a cigar wrapper is, biologically, a mold’s birthday party.

Margin matters in two regimes.

1. Steady state. A humidor with a 6–8°C margin at 19°C/65% RH is fine. Air that close to saturation needs to cool meaningfully before condensation forms. Most overnight temperature fluctuations stay inside that buffer.

2. Local excursions. This is where everything goes wrong. The average reading from your hygrometer can show a comfortable 7°C margin while a corner of your humidor (the one nearest the cooling element, or the one farthest from any air circulation) is sitting at 1°C of margin overnight, every night, all winter. That corner is where mold appears first.

The mold that almost cost me a row of Cohibas

The wineador case I kept finding in research, mirrored almost exactly to my own near-miss, is documented in painful detail on the CigarPass forum^[3]. A collector running 65–68% RH at 68–72°F (numbers that look excellent on any dashboard) lost roughly twenty cigars to mold concentrated in his Cuban drawer, with one cigar showing a flared foot from a direct condensation drip. When he investigated, he found water pooling at the bottom-back of the wineador, oversaturated humidity beads slowly leaking moisture, and a plugged drain hole that trapped the puddle his fan was circulating around the box.

His average dashboard reading was honest. The asymmetry inside the box was not. Local dew-point excursions in dead-spots near the cooling assembly were probably hitting saturation overnight; his hygrometer, sitting in the well-ventilated middle of the box, never saw it.

My near-miss was the same story in miniature. The wineador’s compressor had been cycling on a heavier schedule than usual for about a week (warm spring weather). The compressor’s inner panel was running cooler than the rest of the box at night. Air at 65% RH and 19°C has a dew point near 12°C. The panel was probably sitting at 11°C for a few hours each morning. Single drop. One shipment of cigars on the shelf below it. One flared foot on a Cohiba I’d bought to age for ten years.

I caught it before it spread. I’ll never go without a dew-point-margin alarm again.

Where dew-point risk hides

Three classes of humidor reliably concentrate dew-point excursions where the hygrometer doesn’t see them:

Wineadors and electric coolers. The compressor’s cool side is colder than the box average, and it cycles. Even a small thermal asymmetry (2-3°C between the cooler back and the warmer front) pushes the back into condensation territory hours before the average reading would warn you. The Prometheus 50-stick mold thread is a good adjacent example^[4]: the storage was running at 77% RH and 78°F, where the dew point is already close to room temperature; tiny local cool-spots tipped it over.

Walk-ins and large cabinets. Stratification. Warmer at the ceiling, cooler at the floor. A walk-in running 65% RH and a 21°C average can be sitting at 18°C near the floor, where the dew point of the ceiling air (being slightly warmer and holding more moisture) might already be uncomfortably close. Anyone with a walk-in needs sensors at top and bottom and the dew-point margin computed for each.

Wooden humidors near cold walls or windows. A traditional wooden humidor sitting on an exterior wall in winter has one face running 5°C colder than the other. The wood doesn’t insulate well enough to prevent this. The interior surface of that cold panel is where condensation forms first.

The pattern across all three is the same: the bulk-air dashboard reads fine; something inside the box is much cooler than that bulk reading; condensation forms there; mold follows.

What to actually do

1. Compute dew-point margin alongside RH and temperature, not instead of. Treat margin as a third primary metric. The Cabinet displays all three because skipping any one of them hides a failure mode the others can’t catch.

2. Set a margin alarm at 4°C. Below that, you’re entering the regime where minor temperature dips can cross saturation. 4°C is conservative; 3°C is the absolute floor for storage conditions where you’re aging cigars you care about.

3. Multi-sensor any box bigger than a desktop. This is the same advice as the previous post and it bears repeating. Walk-ins, wineadors, and large cabinets stratify. A single sensor in the friendly part of the box reports the friendly part of the box. It can’t tell you what the cold corner is doing at 4 AM.

4. For wineador owners specifically: clean the drain, check the fan, watch for puddles. Three of the most common mold case studies on every cigar forum I’ve read trace back to one of these three failures. The hygrometer keeps reporting a sensible RH right up until you open the box and find what’s been brewing.

The hard thing about humidor monitoring is that the failures don’t show up in any single number. RH alone misses VPD. RH and temperature together miss dew-point margin. Dew-point margin alone misses spatial gradients across a big box. The Cabinet computes all three on every reading from every sensor I add to it, because the morning I almost lost a Cohiba in April taught me the dashboard is only as good as the worst metric on it.

The payoff that surprised me: most mornings, the dashboard tells me nothing is wrong. Which is why most mornings nothing is wrong. After enough months of watching three numbers hold steady through compressor cycles, AC swings, the radiator coming back on in October, the underlying anxiety of “is the box actually okay or am I about to find another flared foot” just gets replaced. By something I’d describe, accurately, as quiet nerd satisfaction. The cigars are aging. The chart is steady. I have time to drink the coffee.

— Roger