CO2 in HVAC Systems and Indoor Air Quality

Demand-controlled ventilation (DCV) reduces outside air intake when a space is unoccupied and increases it when people are present, instead of ventilating at a constant rate at all times. Compared to a constant air volume system, a CO2-based DCV system can save more than 50% of HVAC energy use — the impact is greatest in small offices, retail stores, and supermarkets compared to other building types.

The principle is simple: CO2 level in a space indicates human presence, and that data can be used to modulate ventilation in real time. But that energy saving depends entirely on the CO2 reading being accurate — a sensor that gets it wrong produces the wrong ventilation decisions, either wasting energy or compromising air quality.

A single-beam, single-wavelength infrared sensor suffers from substantial long-term drift: the incandescent bulb it uses as a light source degrades with use — tungsten evaporates and blackens the glass, reducing emitted light intensity. The sensor interprets that loss of intensity as a rise in CO2, producing increasingly unreliable readings.

To compensate, many manufacturers use automatic background calibration that assumes the lowest recorded reading corresponds to fresh air (400 ppm). That assumption fails in two common cases: in occupied buildings, where CO2 rarely drops below 600–800 ppm; and in new buildings, where concrete carbonation can push ambient CO2 below 400 ppm. In both cases, automatic recalibration introduces a systematic error that ends in inadequate ventilation and worse air quality.

Dual-beam sensors (two light sources) add a second lamp to try to compensate for drift, but that introduces an extra potential failure point and uneven dust accumulation between the two optical paths — in practice, they turn out to be relatively unreliable over the long term.

Vaisala's CARBOCAP® technology uses a single beam and a single light source, but measures two wavelengths: the CO2 absorption wavelength and a non-absorbing reference wavelength. A tunable Fabry-Perot filter switches between both measurements, so any change in light source intensity or optical contamination affects both measurements equally — and cancels out in the comparison, with no need for complex compensation algorithms or assumptions about a fresh-air baseline.

Vaisala's microglow light source, used in the GMW80 and GMW90 wall transmitters, extends sensor lifetime by 50% and cuts power consumption to a quarter of a traditional incandescent source — exactly the transmitters designed for green building projects and demand-controlled ventilation.