The sealed case paradox

A display case has one specification that governs almost everything it does: how fast its interior air is replaced by room air. Reduce that rate and the case becomes a better humidity buffer. Reduce it and the case also becomes a better accumulator of whatever its own materials emit.

These are the same action. There is no setting at which one improves and the other does not degrade, which makes the case seal a genuine trade-off rather than a quality to be maximised.

How the failure happens

The sequence is depressingly consistent, and it usually follows a well-intentioned improvement.

A collection notices that a case tracks the gallery humidity closely. Someone fits gaskets to the access panel and seals the cable entry. The case now holds a stable band, the logger inside confirms it, and everyone is satisfied.

Eighteen months later a lead object in the case has developed a white powdery corrosion, or the silver needs polishing every few months, or a shell object has begun to flake. The case deck is chipboard, faced with a laminate that never covered the underside. It had been emitting acetic acid all along, and the room’s air exchange had been carrying it away.

Improving the case removed the ventilation that had been protecting the object from the case.

Which risk dominates

The answer depends on what is inside, and it can be reasoned about rather than guessed.

Objects at risk from humidity swings are the hygroscopic organic ones: wood, ivory, parchment, textiles, panel paintings. For these, sealing is a clear benefit provided the materials are sound.

Objects at risk from organic acids are lead, shell, calcareous material and some glasses. Objects at risk from sulphur compounds are silver and copper alloys. For these, a sealed case built from untested material is worse than an open shelf.

A mixed case containing both a wooden object and a lead one is the difficult case, and the resolution is not a compromise on the seal. It is to fix the material problem so that sealing becomes unambiguously good.

The resolution

Find out what the case is made of. This is the step that gets skipped, and everything else depends on it. The construction, the deck, the mount board, the fabric covering, the adhesive, the paint. Any of them can be the source.

Test or specify. Materials going into a sealed enclosure should have an Oddy test result or come from the established inert list: glass, powder-coated steel, anodised aluminium, stainless steel, polyester, polyethylene, polypropylene.

Where replacement is impossible, barrier and absorb. Sealing internal surfaces reduces emission substantially, and pollutant sorbents take up what remains. Both are partial and the sorbent is a consumable with a replacement schedule, but together they make an existing case usable.

Verify with metal coupons, not with a sensor. Small polished lead, silver and copper coupons left inside the case for some months and then inspected tell you directly whether corrosive vapours are reaching a level that matters. That is a far more meaningful answer than a total VOC number, for the reasons set out in VOC monitoring.

Then seal, and log the interior. Once the material question is settled, sealing is straightforwardly good and the humidity buffering can do its work.

The general form of the mistake

The pattern here is not specific to cases. It appears whenever an intervention is optimised for one agent of deterioration without checking what it does to the others.

Humidification that produces condensation inside a wall. A tight building envelope that concentrates indoor pollutants. Cold storage that introduces condensation risk on retrieval. Sealed bags around degrading plastics that accelerate the degradation. In each instance the intervention is correct for the agent it targets and creates a problem elsewhere.

The habit that catches it is to run through the list of agents after every proposed change rather than before only the one that prompted it. It takes ten minutes and it is the difference between an improvement and an expensive lesson.