Storing the unstorable.

The brief

A regional history museum — named privately on request — held a 380-piece mixed-media collection that had been stored, at the time we were retained, in a single 1,400 ft² basement room at general-archival conditions of 20 °C / 50% RH. The collection was a thematic one: the regional iron-and-steel industry, late 19th to mid 20th century. Object types included cast-iron architectural fragments, copper-alloy machine parts, painted-and-japanned ironwork, framed photographic prints (silver-gelatin and a small set of color), original engineering drawings on linen, large-format glass-plate negatives, two cellulose-nitrate negative sets, a small set of plywood-and-hide drafting tables, oak company-history ledgers, and a 14-piece collection of trade-association banners on cotton.

The 2023 condition survey found that some of the objects were doing well at 20/50 (the iron, the bronze, the painted ironwork) and most were not. The drawings on linen were curling at 50% RH. The ledgers were drying at 50% RH. The cellulose nitrate was actively decomposing on visible inspection. The glass-plate negatives were mirroring. The cotton banners were dimensionally stable but dust-soiled. The cast iron was rusting at the points where it had previously been painted with linseed-oil-based paint, where the paint had failed.

The conservator’s recommendation: separate the collection into compatibility groups, give each group a regime, store each group in its own envelope. Easy to write. The institution did not have the budget, the floor area, or the political appetite to add four new dedicated vaults. The institution had budget for one room. The brief became: store the unstorable, in one room.

The collection’s requirements, by material

Where the requirements collide

The collisions are immediate. Iron at 35% RH is incompatible with linen drawings at 55% RH in the same room. Cellulose nitrate at 2 °C is incompatible with cast iron at 18 °C in the same room. Color photography at 5 °C and 30% RH is incompatible with linen drawings at 20 °C and 55% RH at every parameter that matters.

What the literature gets right is each individual material’s ideal envelope. What the literature does not address — or addresses only obliquely, in occasional papers, in the AAM conservation FAQ, and in scattered AIC working-group threads — is what to do when one room must hold them all. The most-cited piece of writing on the subject is Garry Thomson’s 1986 paragraph in The Museum Environment recommending compromise at 20/50 for mixed collections. Thomson’s paragraph is honest about the compromise; the literature has, in the four decades since, mostly carried over Thomson’s 20/50 number without carrying over his honesty.

Rooms-within-rooms

The approach we settled on, after Gate M2 and a six-week back-and-forth with the conservator and the institution’s curator, is one we now call “rooms-within-rooms.” The 1,400 ft² institutional storeroom is held at a single Class-A regime: 18 °C / 40% RH. Inside the storeroom, three smaller envelopes are constructed:

1. A 220 ft² vapor-tight cool-storage walk-in at 5 °C / 30% RH for the glass-plate negatives, the color photography, and (after a 12-month equilibration) the silver-gelatin prints.
2. A 60 ft² fire-rated nitrate vault at 2 °C / 30% RH, vented to outdoor, with a separate dedicated air-handling skid. The 22 nitrate negatives live there.
3. An 80 ft² “humid alcove” at 20 °C / 55% RH for the linen drawings and the small wood-and-hide tables. Vapor-tight envelope, dedicated humidification skid, separate sensor mesh.

The remaining 1,040 ft² of the room holds, at the room’s baseline 18/40, the iron, the bronze, the painted ironwork, the cotton banners (in vapor-tight cabinets), and the oak ledgers (in vapor-tight document boxes that buffer against the room’s lower RH).

The architectural footprint of the three inner envelopes is approximately 360 ft², or 26% of the original floor area. The institution did not lose net usable storage, because the inner envelopes are denser-packed than the open-floor storage they replaced. The mechanical cost was approximately 3.4× the cost of a single-class A renovation. The energy cost is roughly 2.1× a single-class A renovation. The preservation cost is, on the conservator’s post-commissioning assessment, dramatically lower — the cellulose nitrate is now stable, the color photography is improving, and the linen drawings are flat for the first time in a decade.

What we built

The mechanical:

• A primary AHU on the building’s 35-ton chiller, sized to deliver the 18/40 baseline. Trane CenTraVac, single-stage. ASHRAE Class A on the room.
• A glycol-loop split system on the cool-storage walk-in: outdoor air-cooled condenser, 12 tons, glycol secondary, indoor evaporator with low-temperature ECM blower. Bally walk-in box with custom Niagara controls overlay.
• A two-stage cascade with desiccant assist on the nitrate vault: pre-conditioning desiccant (Munters MX), cascade R-1234ze chiller, vapor-tight envelope, vented to outdoor through a 4-inch sleeved chase.
• A small dedicated DOAS skid on the humid alcove: 320 cfm, conditioned to 20/55, with humidifier on a polished-RO water feed.
• A LoRaWAN sensor mesh of 31 nodes across all four zones, with two in-cabinet sensors for the cotton banners and four in-box sensors for the oak ledgers.
• A Niagara BMS handling the four zones as a federated cluster, with a single cross-zone alert dashboard.

Equilibration was staged. The cool-storage walk-in took five months. The humid alcove took six weeks. The nitrate vault took three weeks. The 18/40 baseline room took two months. The whole project ran 14 months Gate-M1 to commissioning.

What it cost, and what we’d do again

The total construction cost was approximately $1.06M, which compared to a single-class A renovation at approximately $310,000 is a 3.4× multiple. The institution’s board accepted the multiple after the conservator presented the alternative — namely, accepting the deterioration of approximately 38% of the collection over the next two decades — in a 90-second slide. The slide was the most cost-effective slide in the project.

What we’d do again, on the next mixed-media commission:

1. Fire the rooms-within-rooms scheme as the default. It works. The capital cost is high; the energy cost is manageable; the preservation outcome is dramatically better than any compromise regime; the institution does not lose floor area.
2. Make the cool-storage envelope the largest of the inner rooms. Photography and glass plate dominate most mixed-media collections by piece count, and they share an envelope.
3. Treat the nitrate vault as a separate building. It is fire-rated, vented, dedicated. It must be independently commissioned and independently retainered.
4. Buffer the iron and the paper with cabinets, not with regimes. Vapor-tight cabinets do most of the work. The room is the second line.

What we’d do differently:

1. Build the humid alcove with a smaller-than-our-original-design dedicated DOAS. We oversized at 320 cfm; 220 cfm would have been enough and the over-cycling cost about 8% in energy.
2. Pause the equilibration of the silver-gelatin prints longer at the 13-degree waypoint. We brought them down too fast and saw a brief uptick in mechanical decay risk at the 11-degree pause.
3. Specify the fire-rating of the nitrate vault to a tighter standard than NFPA 40 minimum. We met the minimum; we should have exceeded it for residual capital protection.

The institution is in its fourteenth month of post-commissioning operation. The 2026 condition survey is scheduled for August. We are looking forward to the report.

— SM, Pittsburgh, 2026