Cordage Paper Vault.

Brief

The client — a regional historical society holding the institutional papers, marketing collateral, ledgers, and photograph albums of the late-19th- through mid-20th-century Pittsburgh-area cordage industry — had purchased the second floor of a 1903 cordage warehouse on the 4100 block of Butler Street in 2018, intending to consolidate three smaller collection-storage spaces into a single climate-controlled archive. The building’s ground floor was a long-running coffee roaster; the third floor was a furniture-finishing studio. The second floor had been vacant for a decade. The walls were soft red brick, the roof was a 1970s-era flat membrane in late life, and the windows were original 1903 wood-sash single-glazed. The floor was structurally a 12-inch heavy-timber-and-iron deck.

The collection: approximately 12,000 bound volumes, 480 archival boxes of loose paper, 31 oversize map drawers, and a previously-unsorted set of 1,400 silver-gelatin photographs. Total cellulose mass on the order of 8.4 metric tons. The collection had been at the previous storage site for forty-eight years at uncontrolled conditions averaging 24 °C and 68% RH summer, 18 °C and 35% RH winter, with daily swings of ±5 °C and ±15% RH. A condition survey by an independent paper conservator in 2022 had identified mechanical decay across approximately 18% of the bound volumes (cracking spines, weakened joints), and incipient mold blooms in approximately 60 archival boxes.

Diagnosis (Gates M1 + M2)

We were retained in March 2023. The Gate M1 walk — sixteen sensors deployed for fourteen days — produced the unsurprising finding that the building was tracking outdoor weather with about a 4-hour lag and a 30% damping. The Gate M2 risk register named 9 of the 10 Michalski agents of deterioration as at unacceptable risk and the tenth (criminals) as at controllable risk. The dew-point envelope (calculated against the building’s realistic equilibrium and against the client’s budget) named Class As at 18 °C / 45% RH (with a winter floor of 35% RH) as the most preservation-favorable point inside the envelope of achievable. Class A was tested and rejected: the brick wall could not hold a 50% RH internal regime in February without a wall-cavity sweat point that the masonry would not survive over a 30-year design life. Class AA was not tested.

The risk register also named the building’s envelope as the lead engineering problem. We declined to design a mechanical that the wall would not survive, and instead proposed a two-phase scope: (1) Build a new vapor-tight inner shell inside the existing brick envelope, leaving a 14-inch ventilated cavity between the historic brick and the new envelope; (2) Run the mechanical against the new shell, treating the original brick as weather skin only.

Schematic (Gate M3)

The mechanical: a single 80-ton magnetic-bearing chiller (Trane CenTraVac), a 12,000-cfm AHU with a hot-gas-reheat coil and a desiccant-assist post-cool, an iso-steam humidifier sized for 60 lb/h with a winter rest mode, a Munters MX desiccant on the storage line, a 2,400-cfm DOAS/ERV for outside air, a sensor mesh of 31 LoRaWAN nodes, and a Niagara BMS handling controls. N+1 redundancy on the chiller (a second 80-ton unit, in-line at standby). Two control loops: a primary AHU loop tied to a return-air sensor, and a trim loop tied to the densest section of the archive. The full schematic is on file with the client and (in redacted form) in the shop library.

The new inner shell: 4-inch closed-cell polyurethane spray foam on the inside face of the existing brick (with a service cavity outside the foam), a vapor barrier on the warm side of the foam, a 5/8-inch fire-rated drywall finish, and a continuous service plenum at the ceiling. The 14-inch cavity between the new shell and the historic brick was left ventilated to outdoor weather, with weep openings at the bottom and ridge vents at the top. The brick is now allowed to behave as it was designed to behave: a hygroscopic mass that breathes seasonally, decoupled from the controlled environment inside.

Build (Gate M4)

Construction took eleven months, from December 2023 to November 2024. The shell took six months. The mechanical took five. Three notable details:

• Window openings were retained; the original wood sashes were restored on the brick face (the cordage industry being part of the building’s historic character) and a flush vapor-tight glazed panel was installed on the inside face of the inner shell, sealed at the perimeter, with the cavity vented to outdoor.
• The chiller had to be lifted to the second floor by a 50-ton mobile crane on a Saturday morning, with Butler Street closed by city permit between Mile-Marker 41 and 43. The lift took 38 minutes. We had budgeted 90.
• The DOAS air-intake was placed on the rear (north) wall, away from the coffee-roaster exhaust, after a six-week air-quality study showed measurable VOC contamination from the roaster on the south side. The roaster’s operator was, gracefully, not at all surprised by the result.

The 18-month ramp (Gate M5)

The collection had been at an effective average of 24 °C / 68% RH for almost five decades. The new vault was designed to hold 18 °C / 45% RH. The delta — 6 °C and 23 percentage points of RH — would be the longest equilibration the practice has run. The plan was for 0.4 °C and 1.5% RH per week, with a four-week pause at the midpoint to let the wall and the collection both catch up.

The collection was moved into the new vault, in stages, beginning April 2025 (still at the previous-site setpoint of 24 °C / 68% RH, transferred in conditioned trucks). The ramp began the day the last box was on the shelf. Five months in, we paused for four weeks. Eleven months in, we paused again. At 18 months — September 2025 — the vault held setpoint, the dew-point trace was inside the IPI envelope, and Petra signed the binder. The total elapsed time from Gate M1 to commissioned: 30 months.

Today

The vault is in its eighth month of post-commissioning operation as of this writing (May 2026). It has held setpoint, within band, every day since the ribbon was signed. The energy bill is roughly 14% above the original estimate — a delta we attribute to the unusual 2025–26 winter cold snap (twelve consecutive days below −12 °C in late January) and not to a system flaw. The 2022 condition survey will be repeated in 2027 by the same outside conservator, who is on retainer to provide an independent five-year follow-up. We are looking forward to that report.

This case taught the practice three things, written into our shop manual for future projects:

1. The 14-inch cavity is the right answer for soft-brick masonry. An inner shell that lets the historic wall breathe is structurally and conservatorially the sound move in a building that was not designed for a tight envelope.
2. Pause the ramp twice, not once. The midpoint pause is the easy one. The wall’s catch-up pause — at roughly two-thirds of the ramp — is the one that buys the collection the cushion it needs against a winter cold snap or a summer humidity event.
3. The chiller lift is on a Saturday. The 50-ton mobile crane is on the morning shift. The street-closure permit is non-negotiable. We have learned this twice now.