Cooling load calculator: sizing AC to lighting and transpiration
In a sealed room every watt you plug in becomes heat to remove; transpiration decides how much of it arrives as moisture. Size to both, then convert to TR.
On this page
A 100 m² flower room with 15.5 kW of LED lighting needs about 5.7 tonnes of cooling, not the 2 tonnes an office rule of thumb gives it, and not the 12 tonnes you get if you add the canopy's transpiration on top of the lights. Both mistakes are common. The first leaves a room that never holds its setpoint through the hottest hour of the photoperiod; the second buys a plant that short-cycles, dehumidifies badly and costs twice what it should. This page gives the one number that sits between them and the split inside it that decides which equipment can actually deliver it.
Why a square-footage rule undersizes a flower room
Office rules assume a modest, steady heat density: a person, a laptop, some overhead lighting. A generic commercial figure puts about a tonne of cooling on every 500–600 ft² of floor [1], which gives a 100 m² (1,076 ft²) room roughly 2 tonnes. The same contractor resource rates even a data centre in a hot climate at one tonne per 200–300 ft² [1]. The worked example below needs one tonne per 190 ft² or so, denser than that data centre, because a flower room running 600–1,000 µmol·m⁻²·s⁻¹ at the canopy draws several times the electrical load per square metre that any floor-area rule was built around. The fix is to stop estimating from area and add up what the room actually draws.
Everything that plugs in becomes heat, and nothing else does
In a sealed room, every watt an appliance draws ends up as heat inside the envelope. Light that leaves an LED strikes leaves, media and walls, and almost all of it is absorbed and turns to heat; a small share is stored as sugar by photosynthesis, which is too small to matter for sizing. The sizing guide most growers meet first puts it simply: multiply total watts by 3.41 to get BTU per hour, with no discount for an "efficient" fixture [2]. Efficiency buys you more light per watt, not less heat per watt.
Standalone dehumidifiers are the line growers most often leave out. A compressor dehumidifier condenses water by moving heat, and it rejects all of it, its own electrical draw plus the heat released as the water condenses, straight back into the room through its hot coil [2]. So a dehumidifier inside the room removes moisture but adds its full electrical input to the cooling load. Fans, pumps and controllers commonly add 1–2 kW on a mid-sized commercial room; each person working the room during the lit period adds roughly 100 W; a combustion CO₂ burner adds its full rated heat output plus water vapour.
That list is the whole energy input. Transpiration is not on it, and that is the point most sizing sheets get wrong.
Transpiration moves heat into the water, it doesn't add any
A canopy under high light evaporates a lot of water, and every litre that leaves a leaf as vapour carries its latent heat of vaporisation with it: about 2.45 MJ per kilogram at 20 °C, or roughly 0.68 kWh per litre [3]. That energy has to come from somewhere, and in a sealed room it comes from the light and warm air already in it. Leaves that transpire run cooler, and the room air warms less than it would over bare concrete. The energy has not grown; part of it has changed form, from warm air (sensible heat) to water vapour (latent heat).
Trade-press engineers writing on cannabis HVAC name the mistake directly: adding a room's full electrical heat to its full transpiration latent heat is a common sizing error, because the second is carved out of the first [4]. The latent figure still matters, just not as an addition. It tells you how much of the total arrives at the coil as moisture, and that decides whether a given unit can remove it.
Total (kW) = lighting + standalone dehumidifiers + other equipment and people
Latent (kW) = litres transpired per day × 0.68 kWh/L ÷ lit hours per day
Sensible (kW) = Total − Latent
SHR = Sensible ÷ Total
Tonnes (TR) = Total ÷ 3.517
BTU/h = Total × 3,412
Divide the latent energy by lit hours rather than 24 because a lights-off canopy transpires only a fraction as much. Measure transpiration as water applied minus runoff over a few days; don't guess it from plant count.
Interactive version on the way. The formula below is what it will compute.
Total (kW) = lighting_kw + dehumidifier_kw + other_kw. Latent (kW) = transpiration_l × 0.68 kWh/L ÷ lit_hours (capped at Total). Sensible (kW) = Total − Latent. SHR = Sensible ÷ Total. Tonnes = Total ÷ 3.517. BTU/h = Total × 3,412. Envelope, fresh-air and combustion gains are extra.Worked example: a 100 m² flower room, start to finish
Take a flower room running 24 LED fixtures at 645 W each: 15,480 W, call it 15.5 kW. Add two standalone dehumidifiers drawing 1.5 kW each (3.0 kW) and 1.5 kW of circulation fans, controllers and staff. The canopy, dense and in late flower, takes 190 L of irrigation a day and returns about 20 L as runoff, so it transpires roughly 170 L/day, or 1.7 L per m² of floor. That water figure is illustrative; use your own log.
- Total heat to remove: 15.5 + 3.0 + 1.5 = 20.0 kW.
- Latent share: 170 × 0.68 = 115.6 kWh a day, spread over 12 lit hours = 9.6 kW arriving as moisture.
- Sensible share: 20.0 − 9.6 = 10.4 kW arriving as warm air.
- Sensible heat ratio: 10.4 ÷ 20.0 = 0.52.
- Capacity: 20.0 ÷ 3.517 = 5.7 TR, or 20.0 × 3,412 ≈ 68,200 BTU/h.
The latent share is 62% of the lighting input: high, as it should be for a dense late-flower canopy, but inside the energy the room actually receives.
- Step 1
Electricity in: 20.0 kW
Every watt drawn inside the sealed room ends up as heat inside it.
- Step 2
The canopy turns part of it into vapour
Leaves use absorbed light and warm air to evaporate water. No new energy is added.
- Step 3
Condensing the vapour releases that heat again
A standalone dehumidifier releases it back into the room as warm air; an HVAC-D coil carries it outside with the rest.
- Step 4
Total to reject: 20.0 kW
Where the dehumidification happens changes the total
The same room can be served two ways, and the choice moves the number.
Standalone dehumidifiers plus air conditioning. The dehumidifiers condense the 170 L/day inside the room and hand the heat back as warm air, so the air conditioner sees all 20.0 kW as sensible load and can be an ordinary high-SHR unit. The cost is the dehumidifiers' own 3.0 kW, which the AC then has to remove as well.
An integrated HVAC-D system. One coil condenses the moisture and removes the heat, and its compressor heat is rejected outside. Take the in-room dehumidifiers out of the example and the total drops to 17.0 kW (4.8 TR), but that coil now has to handle 9.6 kW of it as latent: an SHR of about 0.43.
That second number is the one that catches people out. A packaged air conditioner is typically built for an SHR of 0.75–0.85, mostly dry cooling, while a dense flower room commonly runs 0.5 or lower [4]. A unit with the right tonnage and the wrong SHR holds temperature while humidity keeps climbing, because its coil was never designed to condense that much water relative to its cooling capacity. Specify the split, not just the tonnage.
From kilowatts to tonnes and BTU/h: checking a contractor's quote
Air conditioning is still sold in tonnes of refrigeration (TR) and BTU/h almost everywhere outside continental Europe. One TR is defined as 12,000 BTU/h, which is 3.517 kW [5]; one kilowatt is about 3,412 BTU/h.
Use this table, or the calculator above, to check a proposal before you sign it.
| Total load (kW) | Tonnes of refrigeration (TR) | BTU/h |
|---|---|---|
| 10 | 2.8 | 34,100 |
| 17 | 4.8 | 58,000 |
| 20 | 5.7 | 68,200 |
| 30 | 8.5 | 102,400 |
| 50 | 14.2 | 170,600 |
Three readings of a quote for a room like the worked example:
- About 2 TR: sized from floor area. It will not hold setpoint.
- About 12 TR: almost certainly adds transpiration on top of the lights. Ask to see the calculation; you are paying for capacity the room cannot use, and an oversized unit dehumidifies badly (next section).
- 5–7 TR: the right order of magnitude once envelope and fresh-air gains are added. Now ask what SHR the unit delivers at your design conditions, and who handles the 170 L/day.
What the calculator leaves out, and what oversizing costs
This method covers the heat generated inside a sealed room. It does not include heat through walls and roof, solar gain in a greenhouse, outside air brought in for ventilation or CO₂ control, or rooms sharing plant. A licensed HVAC engineer adds those, applies a design margin, and sizes the installed equipment; bring this number to that conversation as a check, not as a substitute for it [5].
Oversizing is not the safe side it looks like. An oversized unit satisfies the thermostat within minutes, shuts down before its coil has run cold long enough to pull real moisture out, and cycles all lit period without settling into steady dehumidification. An undersized one runs flat out and still loses the hottest hour. That is why commercial rooms above roughly 15–20 kW are usually built on several smaller staged units: capacity steps with the load, and one failed compressor costs a fraction of the total rather than all of it, which is the whole subject of HVAC redundancy when a unit fails in flower.
Layout choices that reduce the load in the first place are in facility design for heat and humidity and commercial HVAC design basics. Splitting installed capacity between sensible and latent duty is in HVAC-D sizing: latent vs sensible load; criteria for packaged units, split systems and chillers are in the commercial AC and chillers comparison; sizing dehumidification on its own is the dehumidifier sizing calculator, and running the load through a power bill is the electricity cost calculator. Building permits, refrigerant handling and plant-room sign-off are set locally; the law section covers where cultivation licences add their own conditions.
Sources
- AirFixture (2024). Proper tons per square foot for commercial HVAC systems Accessed 2026-09-27.
- Hydrobuilder Learning Center (2026). Grow room air conditioner sizing guide and BTU calculator Accessed 2026-09-27.
- Allen RG, Pereira LS, Raes D, et al. (1998). Crop evapotranspiration: guidelines for computing crop water requirements. FAO Irrigation and Drainage Paper 56 Accessed 2026-09-27.
- Schurk D (2023). Latent loads matter: HVAC for cannabis grow facilities. HPAC Engineering Accessed 2026-09-27.
- Engineering ToolBox (n.d.). Calculating cooling loads Accessed 2026-09-27.