Preventive conservation, without the mystique

Most damage to collections is slow. A painting does not fail on a Tuesday; it responds, over decades, to a room that swings between damp winters and dry heating seasons. A textile does not fade in an afternoon; it accumulates light. Preventive conservation is the discipline of changing those surroundings rather than treating the consequences, and its central claim is unglamorous but well supported: the cheapest intervention is almost always the one that stops damage from starting.

The agents of deterioration

The framework most widely used comes from the Canadian Conservation Institute, which sorts causes of loss into ten agents. It is worth knowing because it stops you from over-focusing on the one risk you happen to have a meter for.

Agent	What it looks like	Can you measure it?
Physical forces	Handling damage, shock in transit, shelf collapse	Partly, via shock and tilt loggers
Theft and vandalism	Loss, deliberate damage	No, this is a security question
Fire	Total loss	Detection, not measurement
Water	Flood, leaks, condensation	Yes, leak detectors and RH
Pests	Insect and rodent damage	Yes, trap counts over time
Pollutants	Tarnish, corrosion, embrittlement	Yes, though not cheaply
Light and UV	Fading, yellowing, embrittlement	Yes, lux and UV meters
Incorrect temperature	Accelerated chemical change, softening	Yes, easily
Incorrect humidity	Dimensional movement, mould, corrosion	Yes, easily
Dissociation	Object survives, information about it does not	No, this is a records question

Three of those, light, temperature and humidity, are cheap and easy to measure. That is precisely why they get most of the attention, and it is also why collections sometimes end up with beautiful climate graphs and an unnoticed moth infestation. Measure what you can, but keep the list in front of you.

Which agents to work on first

There is no universal order, but there is a defensible one, and it follows from how fast each agent acts and how reversible its damage is.

First, the catastrophic and the fast. Water and fire can destroy in an hour what climate takes fifty years to do. A leak under a storage rack, a blocked gutter above a store room, a display case directly beneath a heating pipe. None of this needs instrumentation, it needs someone walking the building with a clipboard.

Second, pests. An unnoticed webbing clothes moth population can go through a textile store in one season, and the damage is not recoverable. Monitoring traps cost very little and tell you within a few weeks whether you have a problem.

Third, light. Fading is cumulative and completely irreversible. Every lux-hour is spent permanently, and unlike humidity there is no bad year that a good year makes up for. If objects are on open display, work out the exposure budget before you buy anything else.

Fourth, humidity and temperature. These matter enormously, but they act more slowly, and fixing them usually costs the most. That is an argument for measuring them first and acting second, not the other way round. A year of data from a data logger will tell you whether you have a real problem or an assumed one.

Fifth, pollutants. Real, sometimes severe for metals and photographs, but harder and more expensive to diagnose. Worth attention once the rest is under control, or immediately if you have visible tarnish or corrosion.

Measure before you spend

The most common expensive mistake in small collections is buying environmental equipment to solve a problem nobody has quantified. A dehumidifier bought on the strength of a damp smell may run for years in a room that was within an acceptable range all along, costing electricity and drying objects it was meant to protect.

The sequence that avoids this is dull and works:

1. Log the space for a full year, or at minimum through one heating season and one summer. Anything shorter misses the swing that matters.
2. Look at the shape of the data, not just the extremes. A room that sits at 58 percent all winter is a very different problem from one that oscillates between 35 and 70 every week.
3. Decide what range you are actually aiming for, and be able to say why. The historic fixed target has been substantially revised, and the target ranges page covers the argument.
4. Only then choose equipment, and size it to the measured load rather than the room volume.

What the standards say, and where they disagree

Two documents dominate practice in Europe. The ASHRAE chapter on museums, galleries, archives and libraries sets out classes of environmental control, from tightly specified through to unconditioned, and pairs each class with the risks it accepts. PAS 198, published in the UK, takes a different approach: rather than prescribing numbers, it asks the institution to define acceptable change and manage towards it.

Alongside these sits the Bizot Green Protocol, agreed by a group of major lending institutions, which widened the loan conditions many museums had treated as fixed. Its stated range is broader than the classic specification, on the grounds that the narrow target was never well evidenced and carries a large carbon cost.

These are not in perfect agreement, and you should be suspicious of any source that presents a single number as settled fact. What they do agree on is more useful than what they dispute: sudden change is worse than steady deviation, and a stable imperfect environment beats an unstable ideal one.

Where to go next

If you are starting from nothing, start with climate monitoring and insect traps. Both produce evidence within weeks. If you already have data and want to act on it, humidity control and light and UV are where most of the decisions get made. If your problem is specific to one kind of object, the materials section approaches it from that end instead.