CO2 Enrichment in Commercial Cannabis

Supplemental CO₂ is one of the few inputs in indoor cultivation that reliably increases yield, and it is one of the easiest to spend money on without ever collecting that yield. It only works when carbon is the thing limiting the crop, which means light, climate and irrigation have to be right first. It also pushes the room toward a humidity range that favors bud rot, so on the humidity side it is the disease ceiling, not the VPD target, that should pick your setpoint.

Setpoints by stage

Everything below assumes a sealed room. CO₂ by stage, keyed to the light level that makes it worth spending:

The floor underneath all of it is light. Enrichment does very little below roughly 800 PPFD PPFD Photosynthetic photon flux density: how many usable light photons land on a square meter of canopy each second, in µmol·m⁻²·s⁻¹. It measures intensity at an instant, not the daily total. Full entry → , because a light-limited canopy is not short of carbon in the first place. Fix PPFD, or the daily light integral Daily light integral (DLI) The total number of usable photons delivered to a square meter of canopy over a full day, in mol·m⁻²·d⁻¹. Intensity multiplied by time, so it captures what PPFD alone cannot. Full entry → behind it, before spending anything on gas.

Climate that goes with those setpoints

Injection runs from 15 to 30 minutes after lights on until 60 to 90 minutes before lights off, and sits at zero overnight. The VPD Vapor pressure deficit The gap between how much moisture the atmosphere holds and how much it could hold. When VPD collapses, transpiration stalls and calcium stops reaching developing tissue. Full entry → target is the one line worth restating: the general 1.0 to 1.4 kPa range still applies, but in an enriched room disease risk picks the end of it, and the end you want is the dry one. Above roughly 70 percent relative humidity Relative humidity The share of moisture air is holding against the most it could hold at that temperature. Because it moves with temperature, cooling air raises RH with no water added. Full entry → you are in the range botrytis Botrytis cinerea The fungus behind bud rot, also called gray mold. Its spores germinate in prolonged leaf wetness at high humidity, at roughly 13 to 24 °C (55 to 75 °F). Full entry → prefers. Dryback Dryback The share of its water content a substrate loses between irrigations, usually given as a percentage of volumetric water content. It is the central measurement in crop steering. Full entry → slows too, so re-baseline the substrate curves before trusting any steering decision.

The relative humidity figures in these tables are calculated from the VPD target rather than measured, using the Magnus equation with leaf temperature taken as equal to the temperature of the surrounding room. That assumption holds better under enrichment than at ambient, for the reason in the humidity section below, which also walks through the arithmetic if you want to run the numbers for a setpoint that is not listed here.

Before you turn the gas on

In order. Each one is what makes the next worth doing.

1. Seal the room. Mini-splits and dehumidification do the climate work. If you are exhausting during injection, you are buying CO₂ for the neighborhood.
2. Get PPFD above 800 at canopy height, measured on a grid rather than assumed from fixture specs.
3. Get humidity control working with the dehumidification capacity to hold 62 to 68 percent at 82 to 85 °F (28 to 29 °C), and alarm on dew point Dew point The temperature at which saturation occurs and water condenses. Sizing dehumidification to a target dew point is more reliable than targeting a humidity percentage. Full entry → rather than on relative humidity alone.
4. Put the CO₂ monitoring in first, before the first injection rather than after. Specifics in Safety, below.
5. Raise canopy temperature and CO₂ together, on the same day. Not the gas first and the thermostat sometime later.
6. Re-baseline drybacks once the gas is running, before making any generative or vegetative call.

Why it works

Cannabis is a C3 plant C3 photosynthesis The carbon fixation pathway used by cannabis and most crops, named for the three-carbon compound it makes first. C3 crops gain from added CO2 because ambient levels leave them carbon-limited. Full entry → , which is the whole reason supplemental carbon does anything. In a C3 plant, carbon dioxide is captured directly by RuBisCO RuBisCO The enzyme that captures CO2 and commits it to sugar. It is slow, hugely abundant, and it also reacts with oxygen, which is the flaw that makes photorespiration possible. Full entry → , and that enzyme carries a flaw: it accepts oxygen about as readily as it accepts carbon dioxide.

When it grabs carbon the reaction is carboxylation Carboxylation The reaction in which RuBisCO attaches CO2 to a sugar backbone and commits that carbon to the plant. It is the productive half of the enzyme's behavior, competing directly with oxygenation. Full entry → and the plant gains sugar. When it grabs oxygen instead, the plant runs photorespiration Photorespiration A wasteful side reaction in which RuBisCO binds oxygen rather than CO2, costing energy and releasing carbon the plant had already fixed. It worsens as temperature rises and eases when CO2 is elevated. Full entry → , a salvage pathway that burns energy and hands back carbon it had already fixed. At the roughly 420 ppm Parts per million (ppm) A ratio rather than a quantity. For CO2 it is by volume: 1,000 ppm means 1,000 CO2 molecules for every million molecules of the surrounding gas. Full entry → outside, a real share of a cannabis plant's photosynthetic capacity goes out this way. Raising CO₂ in a sealed room tilts the competition at the enzyme toward carbon, photorespiration falls, and net photosynthesis Net photosynthesis The carbon a leaf fixes minus what it loses to respiration and photorespiration. It is the figure gas exchange studies report, and the one enrichment is meant to raise. Full entry → rises.

The second consequence matters more than most growers expect. With more carbon available, the plant can meet its needs with its stomata Stomata Adjustable pores on the leaf surface, each framed by a pair of guard cells, that let CO2 into the leaf and water vapor out. The plant opens and closes them constantly. Full entry → less far open. Stomatal conductance Stomatal conductance A measure of how readily gas passes through a leaf's stomata, reflecting how far those pores are open. It falls under elevated CO2, which is why enrichment lowers water use. Full entry → falls, transpiration Transpiration Water moving up through the plant and evaporating out through the stomata. It drives nutrient uptake, and carries calcium, which moves almost entirely by that flow. Full entry → falls with it, and water use efficiency Water use efficiency (WUE) How much carbon a plant fixes per unit of water it transpires. Elevated CO2 raises it sharply, because the plant takes in more carbon while opening its stomata less. Full entry → climbs. That one change drives most of the section below.

What changes when the gas goes on

CO₂ is not a setting you bolt onto an existing recipe. It changes the plant's water relations, and climate and irrigation have to move with it.

Canopy temperature: 76 to 78 °F (24 to 26 °C) becomes 82 to 85 °F (28 to 29 °C)

What to do. Move the canopy setpoint up by about 6 °F (3 °C), on the same day the gas comes on.

Why. Photorespiration gets worse as temperature rises, which is the main reason conventional rooms settle at 76 to 78 °F (24 to 26 °C): push warmer without extra carbon and you lose more to the oxygenation reaction than you gain from faster enzyme kinetics. Elevated CO₂ removes that penalty. With photorespiration suppressed, the temperature at which net photosynthesis peaks moves up, and the crop can finally use the speed that warmth buys. The mechanism is reviewed across C3 species by Dusenge, Duarte and Way (2019), though the exact size of the shift has not been measured directly in cannabis.

If you finish cool for color. A cool finish for anthocyanin expression pulls the opposite way, and under enrichment it costs more than it does at ambient: you give up the warmer optimum and you keep paying for gas a cool canopy cannot use. What that trade is worth, and the bud rot risk that comes with it, is in why cannabis turns purple.

What it costs to skip. Running 1,200 ppm at 76 °F (24 °C) is the single most common reason enrichment underperforms. The gas gets bought, the thermostat never moves, and most of the benefit stays in the tank.

Relative humidity: up, but disease risk sets the ceiling

What to do. Hold VPD at the dry end of the range, 1.3 to 1.4 kPa, which puts relative humidity at 62 to 68 percent anywhere in the 82 to 85 °F (28 to 29 °C) band. Do not let the VPD math walk you up to 70 percent and above.

Why humidity has to rise at all. Warmer conditions hold more moisture, so the same VPD at a higher temperature takes a higher relative humidity. At 77 °F (25 °C), 1.2 kPa works out to roughly 62 percent. At 84 °F (29 °C) the same 1.2 kPa needs roughly 70 percent. Rooms that raise temperature while holding the old humidity setpoint end up running the canopy far drier than intended, which is its own problem.

What too dry actually costs. Hold 62 percent while the room sits at 84 °F (29 °C) and the canopy is seeing about 1.5 kPa, not the 1.3 to 1.4 you meant to run. The dial never moved; the temperature moved underneath it. Stomata close against a gradient that steep, and under enrichment they are already part closed, so the two effects stack. That throttles CO₂ uptake, which is the one thing the gas was bought to increase, so the room pays for enrichment and then shuts the door the carbon comes through. The tells are leaf edges cupping upward, growth flattening, and irrigation demand climbing while the root zone struggles to keep pace. None of which argues for a wetter room: dry is still the safer direction to miss in, because the wet miss is bud rot. The error is drifting past the range, not sitting at the dry end of it.

A measurement subtlety. Leaf temperature normally sits below room temperature because transpiration cools the leaf. Under enrichment the plant transpires less, so it cools itself less, and leaf temperature runs closer to room temperature than your VPD math assumes. Measure the canopy with an infrared thermometer rather than trusting the wall sensor.

One thing that gets easier. Dehumidification load falls, since the plant is transpiring less and the target itself allows more moisture in the room. The equipment problem is smaller under enrichment, not larger.

Running the humidity numbers yourself

Air can only hold so much water vapor before it begins to condense, and that ceiling depends on temperature alone rather than on how much moisture happens to be present. The ceiling is the saturation vapor pressure Saturation vapor pressure (SVP) The ceiling on how much water vapor the atmosphere can hold at a given temperature, in kPa. It climbs steeply as a room warms, so the same VPD needs a higher relative humidity in a warmer room. Full entry → , or SVP. Relative humidity is how far up toward that ceiling a room currently sits. Vapor pressure deficit is how much headroom is left above it. They are two readings of one quantity taken from opposite ends, which is why either one converts into the other the moment you know the temperature.

Once you know SVP, the conversion is one line:

RH = 1 - (VPD / SVP)

SVP itself comes from an empirical curve fit called the Magnus equation. It is not derived from first principles, it is fitted to measured data, and across every temperature a flower room will ever see it lands within a fraction of a percent of the true value:

SVP = 0.61094 × exp(17.625 × T / (T + 243.04))

T is the only input the equation takes. Saturation vapor pressure depends on temperature and nothing else: how humid the room already is does not move the ceiling, it only decides how much of the ceiling is already used up. Feed it canopy temperature rather than the reading off the wall, since the leaf is the surface a VPD target is about, and note that converting back to relative humidity assumes those two are the same number, which is exactly the assumption the measurement note above is warning you about. SVP comes out in kPa, and T is in °C, so convert a Fahrenheit reading first with (°F - 32) × 5/9.

The 0.61094 is nothing more than the saturation vapor pressure at freezing. The other two, 17.625 and 243.04, are curve-fitting numbers and nothing else: someone measured saturation vapor pressure across a wide span of temperatures, then went looking for the pair that made this equation trace those measurements most closely. They are not constants of nature, which is why other references quote slightly different pairs, each fitted against different data. Across the range a flower room occupies, those versions agree with this one to within about a fifth of a percent, so which pair you use does not matter.

Worked through at the enriched setpoint: at 84 °F (29 °C) the equation returns an SVP of 4.00 kPa, so holding VPD at 1.3 kPa puts the room at 1 - (1.3 / 4.00), or 67 percent. Push to the dry end of the range at 1.4 kPa and it falls to 65 percent.

The exponential is the part worth carrying around. SVP at 84 °F (29 °C) is roughly 26 percent higher than at 77 °F (25 °C), which is the whole reason an enriched room needs a higher humidity setpoint than the one it ran at ambient, and the reason holding the old percentage lands the canopy at the 1.5 kPa described earlier. The same curve is why the disease ceiling arrives so quickly under enrichment: at 1.2 kPa the 84 °F (29 °C) row already sits at 70 percent, which is where botrytis starts to prefer the room, and that is why the VPD target gets pushed to the dry end rather than the middle.

Powdery mildew and bud rot: what the warmer room actually changes

The temperature change works in your favor here. The humidity change works against you. Running the dry end of the VPD band is how you keep both.

Temperature moves you out of the bud rot window. Mahmoud et al. (2023) puts Botrytis cinerea Botrytis cinerea The fungus behind bud rot, also called gray mold. Its spores germinate in prolonged leaf wetness at high humidity, at roughly 13 to 24 °C (55 to 75 °F). Full entry → at its most destructive on greenhouse cannabis above 70 percent relative humidity and at 63 to 75 °F (17 to 24 °C). A conventional room at 76 to 78 °F (24 to 26 °C) is sitting right on the top edge of that window. An enriched room at 82 to 85 °F (28 to 29 °C) is a clear 7 to 9 °F (4 to 5 °C) past it. The temperature you raise for carbon reasons is also temperature that botrytis likes less.

Humidity is the half that bites, and the naive VPD math lands on the threshold. That same 70 percent is where the generic advice puts you. Holding 1.2 kPa at 85 °F (29 °C) needs 71 percent. Holding 1.0 kPa needs 73 to 76 percent anywhere in the band. Both hand botrytis its preferred humidity while the temperature is busy keeping it away, which is not a trade worth making for the small VPD difference involved.

At 1.3 to 1.4 kPa the whole band clears it. Every temperature from 82 to 85 °F (28 to 29 °C) lands between 62 and 68 percent, under the threshold at every point, and the dew point runs 68 to 73 °F (20 to 23 °C):

Powdery mildew Powdery mildew A fungal disease that coats leaf and stem surfaces in white powdery growth. Unlike most fungal pathogens it needs no free water to infect, only high humidity. Full entry → needs no free water, only humid conditions, which is what makes it track the humidity setpoint more directly than botrytis does. The evidence is thinner on this side: the peer-reviewed cannabis work describes its conditions qualitatively, as high humidity and moderate temperature, rather than giving the numeric thresholds the botrytis work gives. Hold the same humidity ceiling, and do not assume temperature is doing the work for you that it does for bud rot.

Alarm on dew point Dew point The temperature at which saturation occurs and water condenses. Sizing dehumidification to a target dew point is more reliable than targeting a humidity percentage. Full entry → , not on a percentage. Relative humidity moves with temperature, so the same reading means different things in a room that has drifted warm or cold. Dew point does not move, which is why it catches a dehumidifier falling behind that an RH number can mask.

And room humidity is not bud humidity. A dense cola holds a wetter microclimate than the sensor on the wall reports, and that microclimate is where bud rot actually starts. Canopy density and air movement through the canopy, not the room average, are what decide whether it stays under the threshold. If you can only do one thing, thin the canopy before you chase the last two points of humidity.

Dryback slows down

What to do. Re-baseline your substrate moisture curves after the gas comes on, before you trust any generative or vegetative decision.

Why. In a coco coir Coco coir A coconut-husk growing substrate. It holds little nutrient charge of its own, so it is fed every irrigation and reacts fast to changes in temperature or EC. Full entry → or rockwool Rockwool An inert substrate spun from molten basalt into fibers and pressed into cubes and slabs. It holds a large volume of water and is used almost entirely in drain-to-waste systems. Full entry → drain to waste Drain-to-waste An irrigation strategy where runoff is discarded rather than recirculated. Runoff readings become the main window into what the root zone is doing. Full entry → system, dryback is driven by transpiration. Cut transpiration by roughly a quarter and the substrate takes longer to give up its water. If your crop steering Crop steering Deliberately pushing a crop toward vegetative or generative growth by manipulating irrigation, EC, temperature and climate rather than leaving the plant to set its own balance. Full entry → strategy is built on hitting a target dryback percentage before the first shot of the day, enrichment shifts that timeline underneath you without announcing itself.

What it does to the crop. A room that was steering generatively at a given shot size and frequency can drift vegetative once transpiration drops, which puts growth into leaf and stem rather than flower. That is the opposite of what the gas was bought for.

Calcium delivery falls

What to do. Watch total daily irrigation volume, and calcium in particular. Feed EC unchanged does not mean calcium delivered unchanged.

Why. Calcium reaches the leaf by mass flow Mass flow The bulk movement of water and dissolved nutrients from substrate to leaf, pulled along by transpiration. Calcium reaches the canopy almost entirely this way. Full entry → , carried along with water. It has almost no phloem Phloem The living vascular tissue a plant moves sugars and dissolved nutrients through. It runs in whichever direction demand pulls, unlike xylem, which only carries water upward. Full entry → mobility, so it depends on transpiration to get where it is going. Less water moved means less calcium delivered.

Where it shows is less settled than it looks. The textbook answer is the newest, fastest-expanding tissue, and Cockson et al. (2019) found exactly that in vegetative plants: stunted growing tips and marginal necrosis on newly expanding leaves. Llewellyn et al. (2023) found the reverse in flowering plants, with margin necrosis starting on lower fan leaves and only reaching the upper canopy two weeks later. They call it a departure from how calcium normally behaves, and their tissue tests make it stranger still: the symptomatic lower leaves held about three times the calcium of the clean upper canopy. Two studies, two patterns. Location alone will not diagnose this, so check both ends of the canopy and confirm with tissue or runoff testing.

Daily schedule

Injection tracks the photoperiod Photoperiod The hours of light a crop receives in a 24 hour cycle. Cannabis flowering responds to the length of the uninterrupted dark period rather than to the light hours themselves. Full entry → , because carbon is only fixed while the lights are on.

Tapering out at the end of flower

The table above covers a single day. The last 7 to 10 days need the other axis, because injection and canopy temperature have to come down together. The 82 to 85 °F (28 to 29 °C) setpoint is only defensible while the extra carbon is suppressing photorespiration, so a room still held at 85 °F (29 °C) after the gas is off is paying the photorespiratory cost with nothing offsetting it. That is worse than either setting on its own. Count the schedule back from harvest rather than from a week number, so it does not assume a cycle length.

To run it in 7 days instead of 10, drop the 650 ppm step and hold each of the remaining four for 2 days. The ppm column is the least load-bearing part of the schedule, since nothing goes wrong if a step lands at 700 instead of 650; the figures exist to keep injection falling ahead of the heat rather than behind it. Night temperature is deliberately not in this table, because how far you open the day/night differential depends on whether you are finishing cool to move pigment. Those night setpoints, and the dew point ceiling each one needs, are in why cannabis turns purple.

Where these numbers come from

The measured cannabis work is Chandra et al. (2008), who put Cannabis sativa in a gas exchange system across a range of light levels, temperatures and CO₂ concentrations. At 750 ppm, which the paper writes as 750 µmol·mol⁻¹, they recorded:

That study's tested range topped out at 750 ppm, and assimilation was still climbing when it got there. It never identifies a saturation point because it never reached one, and it is routinely cited as though it did.

So the familiar 1,000 to 1,200 ppm figure does not come from a published cannabis dose-response curve Dose-response curve A plot of how a response changes across a range of an input. For CO2 it means photosynthesis measured at a series of concentrations, which is the only way to locate a saturation point. Full entry → , because there is not one. It comes from the general C3 literature, the cost of gas, realistic leak rates, and worker safety. It is an economic convention that has held up well in practice, and it is better to say that plainly than to dress it up as a plant science finding.

The temperature shift is the same kind of claim. Dusenge, Duarte and Way review the mechanism across C3 species, so the direction is solid. The size of the shift in cannabis specifically has not been measured.

What the evidence does not cover

• Potency. Enrichment raises grams, not cannabinoid Cannabinoid The class of compounds including THC, CBD and CBG, produced in the glandular trichomes of cannabis flower and responsible for its psychoactive and medicinal effects. Full entry → concentration. Potency per gram tends to hold flat or dip slightly while total grams go up. Specific yield-lift percentages at specific ppm values circulate widely, but they trace back to vendor copy rather than to a study you can read.
• Anything above 750 ppm. There is no cannabis data either way. What is well established across C3 crops is that the response curve flattens, and that sustained very high CO₂ can trigger photosynthetic acclimation Photosynthetic acclimation The downward adjustment many C3 plants make under sustained high CO2, reducing RuBisCO content so the early gain in photosynthesis shrinks over time. Also called down-regulation. Full entry → . Higher is not better. It is more expensive, and it widens your exposure risk.
• Tapering and terpenes Terpene The volatile aromatic compounds a cannabis plant produces in its trichomes alongside cannabinoids. They account for how a cultivar smells and tastes, and they evaporate readily. Full entry → . We have not found published cannabis work testing whether a late taper changes terpene retention, in either direction. The taper is recommended here on cost, and on the fact that plants entering senescence Senescence The natural winding-down at the end of a plant's life cycle, when nitrogen moves out of the leaves and chlorophyll breaks down. Full entry → assimilate less carbon.

That is where the literature stands as of August 2026, as far as we can find it. One well-run dose-response trial above 750 ppm would settle the setpoint question. If there is published work we have missed, or you have replicated data of your own, send it over and we will update this section.

Sourcing the gas

Sizing starts from what a room consumes. A 40 by 20 ft flower room holds about 226 m³ over 74 m² of canopy. Lifting it from 420 to 1,200 ppm takes about 318 g of CO₂, which feels like the answer but is not, because the canopy then fixes carbon all day long. Two independent estimates, one from canopy assimilation rate and one from the carbon content of the biomass you harvest, put daily uptake between roughly 1.8 and 2.8 kg. That is the number a source has to meet, before accounting for leakage.

A 50 lb cylinder holds 22.7 kg, so about nine days at 2.5 kg per day, and bulk liquid scales that up without changing anything else. Capacity is not what you are buying. A regulator and controller hold a setpoint and shut off at lights out, and that is what a commercial room needs.

Safety

OSHA's permissible exposure limit is 5,000 ppm as an eight hour time weighted average, so a 1,200 ppm setpoint is nowhere near dangerous in normal operation.

Normal operation is not the scenario worth designing for. A failed regulator or a stuck solenoid in a sealed room with someone working inside it is. Whatever the source, that case is the one that justifies hardwired monitors with audible alarms and a written evacuation protocol.

References

1. Chandra, S., Lata, H., Khan, I. A., & ElSohly, M. A. (2008). Photosynthetic response of Cannabis sativa L. to variations in photosynthetic photon flux densities, temperature and CO₂ conditions. Physiology and Molecular Biology of Plants, 14(4), 299–306. https://doi.org/10.1007/s12298-008-0027-x
2. Cockson, P., Landis, H., Smith, T., Hicks, K., & Whipker, B. E. (2019). Characterization of nutrient disorders of Cannabis sativa. Applied Sciences, 9(20), 4432. https://doi.org/10.3390/app9204432
3. Dusenge, M. E., Duarte, A. G., & Way, D. A. (2019). Plant carbon metabolism and climate change: elevated CO₂ and temperature impacts on photosynthesis, photorespiration and respiration. New Phytologist, 221(1), 32–49. https://doi.org/10.1111/nph.15283
4. Llewellyn, D., Golem, S., Jones, A. M. P., & Zheng, Y. (2023). Foliar symptomology, nutrient content, yield, and secondary metabolite variability of cannabis grown hydroponically with different single-element nutrient deficiencies. Plants, 12(3), 422. https://doi.org/10.3390/plants12030422
5. Mahmoud, M., BenRejeb, I., Punja, Z. K., Buirs, L., & Jabaji, S. (2023). Understanding bud rot development, caused by Botrytis cinerea, on cannabis (Cannabis sativa L.) plants grown under greenhouse conditions [review]. Botany, 101(7), 200–231. https://doi.org/10.1139/cjb-2022-0139
6. Occupational Safety and Health Administration. Annotated PELs, Table Z-1. https://www.osha.gov/annotated-pels/table-z-1