Carbon Dioxide Levels as a Proxy for Ventilation Adequacy

Carbon dioxide builds up indoors because people breathe, and in most occupied rooms, human exhalation is the only meaningful source of it. No combustion, no decay, no large animals adding to the mix: just lungs. That single fact turns CO₂ into something more useful than a pollutant to worry about. It becomes a measurable stand-in for a question that's otherwise hard to answer in real time: how much of the air in this room has people already breathed, and how much of it is fresh?

The outdoor atmosphere is currently around 420 ppm. Treat that as the baseline. Every ppm measured indoors above that number represents exhaled breath that hasn't yet been replaced with outside air. A room reading 420 ppm is, for practical purposes, breathing outdoor air directly. A room reading 1,200 ppm has accumulated a lot of rebreathed air, and the gap between those two numbers tells you how well the ventilation system is doing its job.

That's also what makes CO₂ a cleaner signal than most other indoor air quality measurements. Particulate matter comes from cooking, from outdoor infiltration, from dust disturbed by foot traffic. Volatile organic compounds leach out of furniture, paint, and cleaning products on their own schedule, independent of how many people are in the room. CO₂ doesn't have that noise problem. With rare exceptions, like a room with a gas appliance running, the molecule's concentration tracks occupancy and air exchange almost exclusively.

The practical value of this goes beyond CO₂ itself. At the concentrations found in occupied buildings, CO₂ isn't toxic. The reason it matters has nothing to do with the molecule's own health effects and everything to do with what travels alongside it. The air exchange rate that lets CO₂ accumulate is the same air exchange rate that lets VOCs build up, that lets exhaled pathogens linger, that lets bio-effluents from skin and breath concentrate in the room. Measuring one molecule gives a building manager a reasonable read on the whole occupancy-linked pollution load, without needing a sensor for every compound in the air.

ASHRAE Standard 62.1-2022, the ventilation standard most U.S. commercial buildings are designed against, builds its guidance on exactly this logic. The standard doesn't set a single hard ceiling for indoor CO₂. Instead, it uses the outdoor baseline plus 700 ppm as a reference point for ventilation adequacy. Given today's outdoor level near 420 ppm, that reference works out to a level substantially above normal outdoor conditions indoors, corresponding to a modest amount of outdoor air supplied per person. Under a 2023 addendum to the standard, that 700 ppm differential stopped being merely advisory for buildings using demand-controlled ventilation: it became a mandatory maximum. The number carries weight in actual system design now, not just in guidance documents.

Effects of elevated CO₂ on occupants

None of this would matter much if elevated CO₂ readings were just an abstract compliance detail, and they are not. Research on cognitive performance, infection risk, and sleep quality all point to the same conclusion: when CO₂ climbs because ventilation is inadequate, people in that room measurably suffer for it.

Cognitive performance carries the deepest evidence base. Harvard's COGfx Study 1 put people through a battery of cognitive function tests in conventional office conditions, then in simulated green building conditions with enhanced ventilation. Test scores roughly doubled in the better-ventilated environment. A follow-up study, the global COGfx project, took the question out of the lab and into real office buildings, tracking workers across six countries. Elevated CO₂, measured as the stand-in for ventilation quality, lined up with slower response times and more mistakes on cognitive tests. The relationship scales with dose: each 500 ppm increase in CO₂ slowed response times by 1.4 to 1.8%, with a measurable drop in throughput alongside it. No floor showed up below which more ventilation stopped helping. Buildings already meeting standard guidelines still produced sharper cognitive performance when ventilation improved further.

Infection risk tells a parallel story, with CO₂ serving as a surrogate for infection risk and not as its cause. A pathogen like SARS-CoV-2 or the bacterium behind tuberculosis gives no visible warning before it spreads through shared air. A 2026 Springer book chapter makes the case for reframing CO₂ away from a comfort metric and toward a biologically meaningful proxy for airborne infection risk: since occupants can't see pathogen load, a measurable number that reflects how much air is being diluted and replaced becomes the practical tool for managing that risk.

Sleep quality and aging bodies extend the stakes past the office. A bedroom with the door closed overnight will likely see CO₂ climb well past comfortable levels by early morning, and most bedrooms have no monitoring at all, despite occupants spending more hours there than in any commercial space they visit. A closed-door home office can cross an elevated threshold within the first hour of focused work. Age adds another layer of risk on top of that. Experimental work published in 2025 found that cerebrovascular reactivity to CO₂, the body's ability to dilate blood vessels in response to rising CO₂, is impaired in healthy older adults: oxidative stress suppresses the internal carotid artery's normal dilation response during CO₂ exposure, an effect not present in younger people. The commercial office remains the center of this discussion, but the mechanism doesn't stop at the office door.

The thresholds building managers work with

Knowing that elevated CO₂ signals a problem is only half the job. Building managers need numbers they can act on, and three threshold bands structure how the industry reads a monitor. The band a space falls into sets whether the right response is urgent, advisory, or simply preventive maintenance.

Below roughly 1,000 ppm sits the acceptable zone under most current frameworks. The WELL Building Standard sets its bar at 900 ppm or below for acceptable ventilation adequacy, with enhanced credit available at 750 ppm. ASHRAE 62.1 doesn't actually anchor itself to a single number in this range anymore: its old 1,000 ppm threshold was removed after 1989, once it became clear the figure was being misread as a health limit rather than a ventilation indicator, and the standard now relies on ventilation-rate procedures instead, with differentials for demand-controlled systems that vary by space type. A study across 18 London school classrooms adds a concrete illustration of why this band matters beyond comfort: once dust reservoirs were removed from the rooms, keeping average CO₂ below 1,000 ppm through increased ventilation also kept airborne particulate matter below WHO guideline levels. Managing one pollutant well managed another for free.

The 1,000 to 1,500 ppm band is where ventilation starts falling visibly behind demand. Conference rooms during a packed meeting and classrooms at peak occupancy are the most common spaces to land here, often with no mechanical failure involved. Enough people breathing in a fixed volume of air, for a long enough stretch, is sufficient on its own to push the reading into this advisory range.

A tighter target comes out of recent school-focused modeling. A Bayesian stochastic CO₂ model applied to Canadian primary schools put the recommended ceiling at roughly 690 ppm when aerosol exposure, not comfort, is the priority. That's well below ASHRAE's comfort-oriented reference point, and the gap between the two numbers makes a useful point on its own: the right threshold depends on what a building manager is actually trying to control. A space managed for general comfort and one managed to suppress airborne infection risk are not the same design problem, even though both get measured with the same molecule.

ASHRAE Standard 241-2023, Control of Infectious Aerosols, changes the math for that second case directly. The standard grew out of a hard lesson from COVID-19: existing ventilation standards weren't built to handle airborne pathogen control, and 241 introduces the concept of equivalent clean airflow to close that gap. Outdoor air, filtered recirculated air, standalone air cleaners, and verified natural ventilation all count toward the same per-person airflow target under this framework, rather than treating outdoor air as the only currency that counts. The EPA has updated its own ventilation guidance to point to Standard 241 for schools, offices, and commercial buildings, and the CDC has confirmed that 241 complements its existing mitigation strategies for cutting transmission of COVID-19, influenza, and RSV.

The three bands together give a single CO₂ reading real operational meaning. A number under 1,000 ppm says a space is in good shape. A number between 1,000 and 1,500 ppm says ventilation needs attention soon, even without equipment failure. A number above that, especially in a space where infection control matters, says the gap between current airflow and the target is serious enough to act on now.

Measuring CO₂ reliably enough to act on the readings

A threshold is only useful if the number feeding into it can be trusted, and that depends on equipment choices most building managers never think to question. The gap between a reliable CO₂ reading and a misleading one is wide enough to produce false confidence in one direction and needless alarm in the other.

Sensor technology is the first fork in the road. Non-dispersive infrared, or NDIR, sensors measure CO₂ directly and remain the right technology for building ventilation monitoring. A cheaper alternative, often labeled eCO₂, estimates CO₂ indirectly by reading VOC levels and inferring a CO₂ number from them. That estimate is not a reliable substitute for a direct measurement, and treating it as equivalent risks building decisions on a number that was never really measuring what it claims to measure. Calibration drift adds a second, quieter risk: even a properly chosen NDIR sensor needs periodic calibration checks, since a drifted unit can report numbers that run systematically high or low for months without tripping any alarm.

Placement matters just as much as the sensor itself. A unit mounted near a supply-air diffuser samples air that just arrived from outside rather than air that's been breathed, so it reads artificially low. A unit placed in the breathing zone, roughly 3 to 6 feet above the floor in a seated space, reads something closer to what occupants are actually inhaling. Ceiling-mounted sensors have their place in system diagnostics, where the goal is tracking how the HVAC system as a whole is performing, but they're the wrong choice for assessing what people in the room are exposed to. Had an elevated reading in a poorly ventilated room been caught by a well-placed, continuously logging sensor rather than a ceiling unit checked once a quarter, the problem would have shown up in the data long before anyone in the room noticed the air felt stale.

That distinction between spot-testing and continuous monitoring matters because a single morning measurement, taken before a room fills up, says nothing about what that same room looks like during a packed afternoon meeting. Occupant density changes through the day, outdoor air conditions change, and operational activity changes, all of which shift CO₂ independently of each other. Continuous monitoring captures threshold exceedances as they happen, builds a trend record over time, and produces documentation that holds up for regulatory or audit purposes. The U.S. Department of Energy's Ventilation Assessment and Action Guide recommends pairing that data stream with a physical building walkthrough, not relying on the sensor alone: once a reading flags a problem, someone still needs to go look at the dampers, filters, and diffusers to find the actual source.

Reading the number correctly takes some judgment too. Handing an uncalibrated monitor to untrained occupants and asking them to react to every fluctuation can backfire: someone chasing a number down on a cold day might over-ventilate a space and create a comfort problem, or waste energy doing it, when the actual issue was more modest. The UK's Scientific Advisory Group for Emergencies has flagged a related trap from the other direction: low CO₂ doesn't automatically mean a space is well ventilated, particularly in large-volume or low-occupancy rooms, where the sheer size of the space can keep concentrations down even when fresh-air exchange is genuinely poor.

Where CO₂ fails as a ventilation proxy

CO₂ earns its place as the default ventilation indicator because it tracks occupancy so cleanly, but that same narrow focus is also its limit. It reads occupancy-driven air exchange reliably. It says very little about pollutants that don't share that source pattern, and treating a good CO₂ reading as proof that the air is clean overall creates blind spots that let those unmeasured pollutants go unaddressed.

The low-occupancy, large-volume room is the clearest case. A warehouse, an atrium, or a half-empty auditorium can post a reassuringly low CO₂ reading simply because there aren't enough people breathing into a large enough volume of air to push the number up, even while the actual air exchange rate is genuinely inadequate. A manager who reads that low number as confirmation of good ventilation, rather than as a side effect of low occupancy and high ceiling volume, is drawing the wrong conclusion from the right data point.

Pollutants with non-human sources present the same problem from a different angle. Particulate matter from a nearby construction site, VOCs off-gassing from new furniture or recent painting, mold spores from a hidden moisture problem: none of these track human exhalation, so none of them register reliably in a CO₂ reading. A room could hold an unremarkable, textbook-normal CO₂ reading and still carry a VOC or particulate load that no CO₂ monitor would ever flag. The co-benefit relationship that makes CO₂ useful, where fixing ventilation for CO₂ also tends to improve other pollutants, only holds when those other pollutants originate from occupants. Once a pollutant's source sits outside the building's occupant load, CO₂ stops saying anything useful about it.

The practical approach treats CO₂ monitoring as one layer in a small stack of measurements, rather than a single number standing in for air quality as a whole. Particulate sensors catch what CO₂ misses from combustion, dust, and outdoor infiltration. VOC sensors catch off-gassing from materials and products that have nothing to do with how many people are in the room. CO₂ keeps doing the job it's suited to: giving a fast, cheap, direct read on whether the air occupants are breathing is being adequately replaced with fresh air from outside. Used that way, alongside instruments built for what it can't see, CO₂ remains the single most practical entry point into understanding what's actually happening in the air of an occupied room.