Dew point as a protocol.

A confession

The first time one of us proposed a setpoint regime, it was for a Squirrel Hill rowhouse with a four-Audubon-plate library, in 2017, and we had the calculator open on a phone but not on a separate monitor, and we proposed 20 °C / 50% RH from memory because that is what the books we had read said. The client agreed. The walls sweat at the inside face for two consecutive winters. The plates were never affected because the boxes the plates were in were vapor-tight and the room held setpoint, but the brick wall behind the bookshelves bloomed two visible mortar weeps, and the homeowner asked us to come back and look.

We came back, opened the calculator on a laptop on the dining-room table, ran the building’s realistic equilibrium against the proposed setpoint, and saw that the wall’s inside face was tracking within 0.8 °C of the dew point during three-week cold snaps. We had not run the calculation in 2017. We had not run it in 2017 because we knew, intuitively, that 20/50 was reasonable for a residential collection. The calculation, run in 2018, told us that 20/50 was reasonable for the room and unreasonable for the wall. We lowered RH to 45%. The walls have not bloomed since.

The story has been a fixture of every junior-engineer onboarding since. It is the practice’s short-form way of saying: the calculator is the protocol; everything else is conversation.

The calculator, briefly

The IPI Dew Point Calculator is a free, web-based tool published by the Image Permanence Institute at the Rochester Institute of Technology. It accepts a temperature and a relative humidity and computes the corresponding dew point. It also reports the IPI environmental ratings — Natural Aging factor, Mechanical Decay risk, Mold Growth risk, and Metal Corrosion risk — for the input conditions. The interface has not changed materially since roughly 2008, which is one of the things we love about it.

The calculator’s value is not the dew-point computation. The dew-point computation is a 1962 psychrometric chart in software; any HVAC engineer can do it on the back of an envelope. The calculator’s value is the four ratings reported alongside the dew point. They are computed from peer-reviewed research by IPI on natural aging of cellulose, on mechanical decay of paper and plastic, on mold-growth thresholds, and on metal corrosion. Reading them in the same plane as the dew point lets you see the trade-offs without doing the algebra.

The four numbers

1. Natural Aging factor. A multiplier on the room-temperature baseline. NA = 1 means the collection ages at the same rate as it would at 21 °C / 50% RH. NA = 12 means twelve times slower. Cool storage projects often target NA > 8.
2. Mechanical Decay risk. A categorical assessment (low, medium, high). Driven principally by RH stability rather than RH absolute value.
3. Mold Growth risk. A categorical assessment driven by RH and dwell time at high RH. Sharp threshold around 60%.
4. Metal Corrosion risk. A categorical assessment driven by RH absolute value at moderate temperatures. Less sensitive to swings.

For most institutional collections, no single regime is optimal across all four. A 5 °C / 30% RH regime gives the best Natural Aging factor but a Metal Corrosion risk near zero is overkill for a paper collection. A 20 °C / 50% RH regime gives a comfortable Metal Corrosion outcome but a Natural Aging factor of 1.0 (no improvement). The collection’s composition decides where in the trade-off space the conservator should land.

What it tells you

What the calculator tells you, in our protocol-of-use, is three things:

1. The wall’s dew-point clearance. Run the proposed setpoint into the calculator, read the dew point. If the wall’s inside face will reach that dew point under any plausible cold-snap scenario, the wall sweats. The proposed setpoint is wrong, or the envelope improvement is wrong, or both.
2. The collection’s preservation outcome. Run the proposed setpoint into the calculator, read the four IPI ratings. If the Natural Aging factor is 1.0, the proposed regime offers no preservation benefit and the chiller load is a luxury. If the Mold Growth risk is high, the regime is unsuitable for the collection regardless of the conservator’s preferences. If both are wrong, the regime is wrong.
3. The trade-off space. Run the proposed setpoint plus four neighbors (±2 °C, ±5% RH) into the calculator. Read all five sets of numbers. Pick the one that makes the most sense given the collection, the building, and the budget. Do not pick the one your memory tells you is right; pick the one the calculator tells you is right.

An example: a private library, 2024

Late 2024, a Mt. Lebanon residence with a 14-piece collection of printed books from the 18th and early 19th century, plus a small set of regional manuscripts. The conservator had recommended Class A: 21 °C / 50% RH. The homeowner, on the basis of an article in The New York Times they had read about preservation conditions at the Bodleian, asked us about Class AA at 21 °C / 50% RH ±2 / ±5.

Petra opened the calculator. Class A at 21/50: NA factor 1.0 (baseline), MD low, MG OK, MC OK, dew point 10.2 °C. Class AA at 21/50: same numbers (the AA distinction is a tightness-of-band question, not a setpoint question). The calculator said the proposed regime offered no preservation benefit over a much laxer regime — what the collection actually needed was lower temperature, not tighter band.

We proposed an alternative: 16 °C / 45% RH on a Class-A band. The calculator returned NA factor 2.6 (a 2.6× improvement), MD low, MG OK, MC OK, dew point 4.0 °C. The homeowner’s walls would clear the dew point during three-week cold snaps with 1.4 °C to spare; the chiller load was lower; the energy bill would be 22% lower; the preservation outcome was meaningfully better; and the room was still comfortable enough for the homeowner to read in for an hour at a time. We commissioned the room at 16/45 in March 2025. It has held setpoint within band for fourteen months as of this writing.

The Bodleian regime, in any case, was a New York Times story whose source was a 1972 paper by Garry Thomson. It was not a regime intended for a Mt. Lebanon residence. The calculator told us this in eleven seconds.

Why the protocol

The shop’s “dew point on a separate monitor” rule is, on its face, a small bureaucratic discipline. The deeper reason for it is that mechanical engineers, by training, are good at sizing equipment to a setpoint and bad at questioning the setpoint. The setpoint arrives, in most projects, as a number on a spec sheet handed down from a curator, an insurer, or a board. The mechanical engineer’s instinct is to design the equipment to deliver the number. The conservator’s instinct is to write the number into a contract. Neither of those instincts asks whether the number is the right number.

The calculator asks whether the number is the right number. It is the only piece of software in our practice that does this consistently. It is therefore the first piece of software we open at the start of every Gate M2 risk register. The screen at Petra’s desk runs the calculator on the right monitor, all day, and she has a small frame around it with a quote from the IPI guide: “Knowing the dew point can help achieve long-term preservation of collection materials for libraries, museums, and archives.” This is the practice’s working summary of why the protocol exists.

If you take one thing from this entry, take this: before you propose a setpoint, run it through the calculator. Run the four neighbors. Read the dew point. Read the natural-aging factor. Decide on the basis of the numbers. Do not decide on the basis of the books you have read.

— PA & JNO, Pittsburgh, 2026