In 2024, California cannabis product recalls increased 800% year-over-year. Most of the recalled inventory failed for the same reason: Aspergillus. Most of the Aspergillus came from the same root cause — dehumidification that couldn’t keep relative humidity below the 60% threshold under cultivation load.
Operators got blamed. Cultivation teams got blamed. Genetics got blamed.
The forensic cause was none of those things. It was the building.
A 10,000 square foot flowering canopy releases more than 2,000 gallons of water vapor into the air every day — 90% to 100% of the irrigation volume the plants receive. If the facility’s mechanical system can’t pull that volume back out, the cultivation room sits above the Aspergillus threshold for hours every day. The result is the inventory you watched the state destroy on the news.
That’s a construction decision, not a cultivation decision. It was made the day the GC sized the HVAC system.
”HVAC” isn’t the cannabis term. It’s HVACD.
Most commercial GCs size cooling by tons per square foot. That’s the right approach for an office building, where the dominant load is sensible (heat from people, lights, equipment). It’s the wrong approach for a cannabis facility, where the dominant load is latent — water vapor.
Cannabis-experienced engineers add the “D” deliberately — HVACD — because handling humidity is a separate engineering problem from handling temperature. Adding the loads together and buying one oversized unit is a documented common error. When the lights go off and cooling cycles down, the dehumidification stops with it. RH spikes. Aspergillus sets in. The cultivation team didn’t change anything; the building did its job exactly as it was built.
Designing for cannabis means sizing equipment for 125% of peak irrigation volume, splitting sensible and latent loads, and maintaining dehumidification independent of cooling cycles. The cost difference between right-the-first-time and an emergency retrofit isn’t 20% or 30%. It’s an inventory destroyed and an audit finding on the operator’s record.
The architect designs for today’s facility. The GC builds for the next one.
Architects translate the operator’s current methodology into a building. That’s their job, and they’re good at it. But the methodology you build for in 2026 isn’t the methodology you’ll be operating with in 2030.
LED replaced HPS in roughly five years. Vertical replaced flat-bench in roughly five years. Living soil and aeroponic systems are the next wave. The cannabis facilities being built right now are being depreciated over 15 to 30 years.
The math doesn’t work unless the building can adapt — and adaptability is decided in construction, not design.
Adaptability lives in things an architect can’t fully control:
- Slab loading sufficient for vertical conversion. Adding two- and three-tier vertical to a flat-bench room adds significant point load. Designed in on day one, the cost is marginal. Reinforced after the fact, the room comes out of production for months.
- Electrical infrastructure sized for the amperage swings between lighting technologies. A facility built for 1000W HPS uses different conduit and panel architecture than one built for LED. Building for the higher load and installing extra conduit runs preserves the option to switch — without re-trenching the floor.
- MEP routing through service corridors adjacent to grow rooms, not above them. Future expansion or methodology change doesn’t require breaching the controlled environment. A layout decision, not a cost decision — but it determines whether year-three changes take weeks or months.
- Pressure cascade architecture between clean zones, anterooms, and corridors. Borrowed from pharma. Cannot be retrofitted without demolishing wall-floor-ceiling junctions. Most commercial facilities are built at neutral pressure, which lets odor escape (a regulatory issue) and lets contamination in (a test-failure issue).
- Wash-down floor systems with integral cove base — seamless epoxy or urethane, radius cove tying floor to wall. Standard in pharma and food. Frequently skipped on cannabis to save $8 to $15 per square foot, then required at three to five times the original cost when the operator pursues GMP certification.
- Camera and security conduit designed before drywall. Running cabling through a sealed cleanroom after the fact is a structural problem, not a punch-list item.
None of this is on the architect’s deliverable. All of it is the GC’s daily decision-making.
In Florida, you can’t separate the retrofit.
Every Florida MMTC operates cultivation, processing, and dispensing under one authorization. When the Department of Health adjusts standards at any one stage, the inspection that follows often touches the others. A cultivation-room finish problem can put the processing room and the dispensary under simultaneous review.
That structural tie means a Florida cannabis GC has to know all three rule sets while the slab is being poured — not as separate compliance check-ins, but as one integrated build. Most generic commercial GCs working in Florida have never had to think about cannabis as a vertically integrated obligation.
This is why the right partnership for a Florida operator looks different from the right partnership for a Colorado or Massachusetts operator. The construction firm has to be regulator-aware, not just code-compliant. Florida already imports food-establishment construction standards into its MMTC framework — DOH inspection plus national-accreditation certification plus food-safety practices. Healthcare-grade in Florida isn’t an upgrade. It’s the floor.
The construction decision compounds at the rate of methodology innovation.
Every cheap-build decision compounds. The wrong slab limits vertical conversion. Undersized panels limit lighting upgrades. Drywall walls become the year-three rebuild that should have been FRP from the beginning. Each individual shortcut is small. The aggregate is a facility on a five-year obsolescence clock against a 15-to-30-year depreciation schedule.
When the license eventually changes hands — and most do — the buyer inherits the building. A facility with documented N+1 mechanical redundancy, GMP-grade finishes, and over-spec’d electrical commands a multiple. A facility with patched HVAC and exposed drywall is a discount asset. The GC’s work is implicitly an asset-valuation decision, made years before the sale.
What Galleon was built to do.
A general contractor who understands cannabis brings the operational fluency of someone who’s worked in the industry, the regulatory awareness of someone who’s read the rules and watched them change, and the construction-execution mastery of someone who’s built to GMP-grade standards in healthcare. When all three line up, the building stops being a one-shot solution. It becomes a platform that can carry the operator across regulatory shifts, methodology changes, and ownership transitions — without the retrofit bills that quietly bankrupt operators who picked the wrong builder.
That’s the work. That’s why we exist.