Project Case Study · Bangkok & Vicinity, Thailand
In April 2026 the Thai Ministry of Public Health changed what a cannabis building has to be. Ministerial Regulation (No. 2) B.E. 2569, effective 30 April, raised the conditions attached to a cannabis flower licence: lawful possession of the premises, dedicated storage able to preserve flower quality, staff holding prescribed medical or cultivation qualifications, at least one DTAM-trained person on site during operating hours, and an effective system for eliminating cannabis odour and smoke.
Our client already held a licensed warehouse in the Bangkok metropolitan region and a cultivation programme with nowhere to put it. The question was never really what to build. It was whether the building they already owned could be made to answer that list without buying land, re-entering permitting, and losing a year to construction.
The short version: cultivation went into volume they had already paid for. A steel mezzanine split the existing warehouse into two growing levels, and the envelope around them was specified against the regulation rather than against a catalogue.
Thailand removed cannabis from its Category 5 narcotics list in June 2022. What followed was four years of retail growth with almost no framework around it, and then a steady tightening that is still running. Anyone planning a cultivation asset in Thailand today is designing against a moving target, and the direction of travel has been consistent.
We supply steel buildings. We are not a Thai licensing adviser, and none of the above is legal advice. What matters on this page is narrower than the law: every one of those steps moved a requirement out of the paperwork and onto the building.
Three options were on the table: build a new cultivation hall on new land, convert an open floor at grade, or go vertical inside the shell they already had. The third won, and it is worth being honest that it did not win on price alone.
A mezzanine moves cost out of the ground and into the structure. You pay for a floor system, a stair, edge protection, and a load check on a frame that was engineered for a different job. It pays back when the shell is sound and tall enough, when the slab can take the new column loads, and when the alternative is a year of permitting on land you have to buy first. Bangkok industrial land and the permitting clock are exactly the conditions that make vertical expansion rational.
Where a mezzanine is the wrong answer: low eaves, a tight column grid, or foundations with no reserve. Squeezing two levels into four metres of headroom gives you two rooms nobody can work in, and no amount of clever framing fixes that. We ask for the eave height, the column grid and the slab build-up before we quote a mezzanine, and if the numbers say build new, we say build new.
The interesting part of this project is not that it has a mezzanine. It is that three zones in the same building got three different specifications, and that the differences follow the work each room does rather than a standard package.
| Zone | Walls | Ceiling | Why this and not the default |
| Cultivation rooms, both levels | 75 mm PIR | 100 mm PIR | The ceiling is the air barrier and the surface the humid room meets first. Walls get washed and hit; the ceiling does not. |
| Chemistry / laboratory room | Stainless-faced PIR | Stainless-faced PIR | Solvents and disinfectants, wiped down daily. Organic facings lose that fight within a couple of seasons. |
| Rest of the warehouse | Existing cladding retained | Unchanged | It was doing its job. Replacing it would have added cost and waste for no performance gain. |
Why PIR. Polyisocyanurate is a polyurethane chemistry with a higher cross-link density, which in practice means a closed cell structure above 90%, a core thermal conductivity around 0.022 W/(m·K), and better reaction to fire than standard PUR. Core-only, 75 mm sits near 0.29 W/m²K and 100 mm near 0.22 W/m²K. We quote the figure including facings and fixings, because the number on a datasheet is not the number the building delivers.
Why the ceiling is thicker than the walls. Warm, humid air rises, so the ceiling is where the room meets its own envelope first. It is also the air barrier: every joint up there is a route for room air into the plenum, where it can condense on a cooler surface and drip back onto the canopy below. Interstitial condensation in a roof void is the failure that costs a crop, because you do not see it until the ceiling panel stains. The walls carry a different load — they get washed, they take trolley impacts, and they are the surface you can actually inspect.
Why stainless in one room only. The chemistry room is where organic coatings lose. Solvents, aggressive disinfectants and daily wiping will dull a painted facing and eventually expose the steel underneath. Stainless-faced PIR costs more per square metre and is worth it in exactly that room. Specifying it across the whole grow area would have been an easy upsell and the wrong call.
The new steel is the straightforward half. The interface between the new steel and the old building is where schedules slip, and it is the part suppliers talk about least.
What this costs you in time: usually the second week on site, when the building turns out to be a few centimetres out from the drawings. It is not a steel delivery delay. It is the difference between a building that was measured and one that was assumed.
Canglong engineers travelled to Thailand for this one, and that is worth stating plainly because a large share of projects in our industry are delivered by remote guidance and then written up as though an engineer had been standing there. Here our team was on site through the build: component assembly, joint work, and commissioning, working alongside the client’s own crew rather than in place of them.

The reason to send people rather than drawings is the joint. A panel schedule tells you what to install; it cannot tell you whether the crew in front of you has sealed the previous three metres correctly. On a cultivation envelope, the joint is the product. Gaps left at a ceiling-to-wall junction are where humidity escapes into the void, where odour escapes into the rest of the building, and where a wash-down finds its way behind the lining.

Case studies in this industry tend to grow claims on the way to publication. Four things this project was not, so the scope is clear before you read a number into it.
Engineering is done in-house rather than outsourced to a drawing office, which matters more on retrofit work than on new build: when the site turns out to be out of square, the change has to come from the people who modelled the frame. Every component ships numbered against that model with a packing list and a loading plan, so the sequence at the container matches the sequence on site.
If you have a warehouse and a cultivation or processing programme with nowhere to put it, the first question is not price. It is whether the shell you have can take a second level, and what the regulations in your market ask the room to do. Send us the eave height, the column grid, and what you intend to grow or process, and you will get a straight answer on both.
Abby coordinated this project from the first enclosure question through to handover on site. Send her your building’s eave height and column grid, what you plan to grow or process inside it, and where it is.
Thailand
Cannabis Factory
2026-6-6
Bangkok, Thailand
3060㎡

No




Your email address will not be published. Required fields are marked *