What humidity should you actually aim for?

For decades the answer was a single number: 50 percent relative humidity, plus or minus five, held permanently. It appeared in specifications, drove the design of a great deal of expensive plant, and was rarely questioned. It has since been substantially revised, and understanding why matters, because the revision changes what you need to buy.

Where the fixed number came from

The figure emerged from mid-twentieth-century museum practice, at a point when air conditioning was newly available and the prevailing instinct was that a constant environment must be the safe one. It was reasonable as a working assumption. What it was not was an experimentally derived optimum, and the confidence with which it was later stated outran the evidence behind it.

The pushback came from two directions. Conservation science produced evidence that many materials tolerate wider bands than assumed, particularly when change is gradual and the object has already equilibrated to its historic conditions. And the energy cost of holding a narrow band in an old building became impossible to ignore.

What the current frameworks say

ASHRAE sets out classes of control, from a tightly specified class through progressively looser ones down to unconditioned. Its useful contribution is not any particular number but the structure: each class is paired with an explicit statement of the risks that class accepts. It makes the trade-off visible instead of pretending a specification is free.

PAS 198 takes the opposite approach. Rather than prescribing values, it asks the institution to establish what change is acceptable for its own collection and to manage towards that. For a small museum this is frequently more useful, because it does not begin by demanding a class the building cannot deliver.

The Bizot Green Protocol, agreed among major lending institutions, widened the conditions commonly written into loan agreements, on the explicit grounds that the narrow specification was not well evidenced and carried a large carbon cost. Its significance is practical: it changed what borrowers can be required to provide.

These frameworks are not identical and it would be misleading to blend them into one number. What they agree on is more useful than what they dispute.

The points of agreement

Rate matters more than range. A collection that drifts slowly across a season is far better off than one that cycles weekly across the same span. Hygroscopic materials respond to change by moving, and slow movement distributes stress that fast movement concentrates.

Historic equilibrium counts. An object that has sat in one set of conditions for two hundred years is in equilibrium with them. Moving it to a textbook figure is an intervention with its own risk, and the burden of proof sits with the change.

Extremes have thresholds, middles do not. Above roughly 65 to 70 percent sustained, mould becomes a real risk on organic substrates. Below roughly 30 to 35 percent, mechanical damage to hygroscopic material becomes likely. Between those, the precise figure matters far less than the stability.

Different materials want different things. No single room specification suits metals, photographs and furniture simultaneously. That is an argument for enclosures rather than for a compromise that suits nothing, and the materials section covers the individual cases.

Choosing a number you can defend

The sequence is the same as everywhere else in this field, and it starts with data rather than a specification.

1. Log the space for a full year and establish what it actually does, including the rate of change. Reading climate data covers the analysis.
2. List what is in it and identify the most vulnerable material rather than the most valuable one.
3. Set a band, not a point, and write down the reasoning. A defensible position is one that names the framework it follows and the reason it departs from it.
4. Set the band around what the building already does where that is safe. Aiming for 55 percent in a building that naturally sits at 55 costs almost nothing; aiming for 50 costs continuously, forever.
5. Handle the outliers with enclosures rather than by tightening the room.

The one figure worth being firm about

Sustained high humidity is the exception to all of this nuance. Mould germination on organic substrates becomes a real risk somewhere above roughly 65 to 70 percent held over days rather than hours, and once it starts it is a treatment problem rather than an environmental one. If your record shows sustained readings in that region, that is the problem to solve first, and dehumidifiers is where to go next.