commercialhvacuk
Get a Quote

Demand-controlled ventilation: how CO2 sensors cut ventilation energy

Updated 23 October 2026 · SEO Dons Editorial

What counts as demand-controlled ventilation?

The ecodesign regulation for ventilation units defines demand control as a device or set of devices that measures a control parameter and uses the result to regulate automatically the flow rate of the unit or ducts. The parameter can be humidity, volatile organic compounds, carbon dioxide, or presence, motion or occupancy detection. A unit that uses demand control is a demand-controlled ventilation unit.

The same text separates central control (one sensor for the whole building or part of it) from local control (more than one sensor on a ducted unit, or one sensor on a non-ducted unit). From 1 January 2016, Annex III also requires every non-residential ventilation unit other than a dual-use unit to have a multi-speed or variable speed drive, which is what lets a fan follow a sensor. CIBSE Journal's indoor air quality module names variable speed fans, possibly with volume control dampers, as the hardware. The signals can feed a building management system, covered in the BMS guide, and the system types are on the commercial ventilation and MVHR page.

Where does demand control save the most, and where little?

CIBSE Journal's demand-controlled ventilation module says the saving depends on how occupancy varies, the minimum flow the system can turn down to, the indicator used and the supply air conditions. Simple occupancy sensors tend to move a room between just two steps, minimum and maximum, which gives more airflow overall than CO2 or VOC sensing. It also cites studies showing that, in cellular offices, typically fewer than 50% to 65% of rooms are occupied at once. The table is a reading of those factors, not a measured result.

SpaceProspectsWhy
Meeting and training roomsGoodEmpty for long spells, full in bursts
Cellular officesGoodRooms are rarely all in use together
ClassroomsModerateFull during lessons; the saving is in breaks and out-of-hours (see school ventilation)
Rooms with air cleaners or CO2-producing processesPoor fitHSE says CO2 monitors are unsuitable there
High-ceilinged halls and warehousesLimitedHSE says monitors are of limited use where air is not fully mixed

Which CO2 levels do HSE and the DfE use?

HSE's guidance on CO2 monitors quotes its Approved Code of Practice: the fresh-air supply rate should not normally fall below 5 to 8 litres per second per person, with 10 recommended in many guides for most commercial buildings. HSE says 1000ppm is equivalent to about 10 litres per second per person, that levels consistently above 1500ppm in an occupied room indicate poor ventilation, and that CO2 is a broad guide rather than a demonstration of ‘safe levels’.

SourceWhat it says
HSE, workplacesConsistently above 1500ppm in an occupied room: poor ventilation
DfE, education settings in England (published 24 February 2026)Under 800ppm good, 800 to 1500ppm adequate, over 1500ppm poor
CIBSE Journal summary of BB101 (2018)1,000ppm for mechanical ventilation; no more than 20 minutes above 1,500ppm in a new classroom

These are operating and school-design figures, not one legal CO2 limit for every workplace. Design guidance for building work in England is Approved Document F. As gov.uk states it, Approved Document F (2026) applies to buildings subject to the 2026 standards, and earlier versions still apply to buildings subject to earlier standards. The page lists a 2026 Volume 1 for dwellings and Volume 2 for buildings other than dwellings separately, so confirm which edition applies. Whatever the target, the control needs a minimum flow so an occupied room is still ventilated.

Where do CO2 sensors go, and how are they kept accurate?

  • Position. HSE says place monitors at head height, away from windows, doors and air supply openings, and over 50cm from people, because exhaled breath reads misleadingly high. CIBSE Journal says fixed sensors are typically in return air ducts or within the occupied spaces. The DfE says one sensor is enough for a typical classroom.
  • Space size. HSE says monitors suit spaces where a consistent number of people stay for over an hour; quickly changing numbers give unreliable results.
  • Drift. NDIR sensors, well calibrated, can read within ±1 to 5% of a reference, but tend to drift over time. The DfE says to test or calibrate periodically in line with the manufacturer's recommendations.

HSE writes for portable monitors, but a fixed sensor faces the same physics. A contractor should record each sensor's commissioning check and the manufacturer's re-check interval.

What might demand control save? A worked example

For example, take a training room for 30 people at 10 litres per second each, so 300 litres per second at the design rate. Assume it runs 2,600 hours a year, 1,500 of them in the heating season at an average 10 K difference between inside and outside. These are placeholders.

  • Heating: air carries about 1.21 kW per cubic metre per second per kelvin, from the constants in CIBSE Journal's heat recovery module (specific heat 1.005 kJ per kg K, specific volume 0.83). Heat is priced at gas with no boiler losses, which understates the value.
  • Fan: 1.5 watts per litre per second, assumed and held constant, so fan power falls in step with flow. A real fan curve can differ.
  • Prices: the DESNZ annual report (updated 19 August 2026) gives 2025 (provisional) non-domestic averages of 5.3 pence per kWh for gas and 24.3 pence for electricity, including the Climate Change Levy.
Average airflow with demand controlHeat savedAt 5.3pFan energy savedAt 24.3pTotal a year
40% of design3,270 kWh£173702 kWh£171£344
60% of design2,180 kWh£116468 kWh£114£230
80% of design1,090 kWh£58234 kWh£57£115

The saving falls as the room gets busier. If a heat-recovery unit at the 73% minimum for most exchanger types also serves the room, only the 27% of heat not recovered is left to save, so the heating line at 40% drops from £173 to about £47. The measures overlap rather than add; the MVHR payback guide sets out that sum.

Payback is the extra cost of sensors, drives or dampers, controls and commissioning divided by the annual saving. This page gives no installed costs, so ask for each as a separate line. For scale only, a CIBSE Journal case study predicted, for one mixed-use building in Falkenberg, Sweden, a saving of more than 50% of operational energy against constant air volume, for about 10% extra investment and a payback under two years.

How does demand control fit with filters and other pollutants?

  • Filters. CIBSE Journal says dust loading cannot be predicted accurately under demand control without more data. The ecodesign regulation requires a unit with a filter to signal when pressure drop exceeds the maximum allowed, so filter changes can follow that signal as well as the calendar.
  • Other pollutants. CO2 is a useful proxy where occupant emissions dominate. CIBSE Journal says particle and gas sensors can add a more detailed picture where occupants are not the main source.
  • Handover. Ask for sensor ranges, set-points, minimum flows and fault behaviour in writing, and for BMS logging of CO2 and airflow.

You can send an enquiry through the form on this site if you want to talk through a ventilation upgrade.

Sources

Common questions

It should not. Regulation 6 requires a sufficient quantity of fresh or purified air in every enclosed workplace, and HSE's Approved Code of Practice says the supply rate should not normally fall below 5 to 8 litres per second per person. A sound design sets a minimum flow that holds whenever the room is occupied.

Does a CO2 sensor show whether the air outside is polluted?

No. CO2 tracks people, not traffic fumes or particles. CIBSE Journal describes CO2 as a useful proxy where occupant emissions dominate, so outdoor pollution needs a particle or gas sensor or a suitable filter. HSE also says CO2 monitors do not suit rooms with air cleaning units.

How often should CO2 sensors be calibrated?

No single interval appears in the HSE or DfE guidance. Both say to follow the manufacturer's instructions, and CIBSE Journal notes that sensors tend to drift. Ask the supplier for the interval and a written re-check procedure.

Is demand control worth it in a school?

It depends on the timetable. Classrooms are close to full during lessons, so the saving comes mostly from breaks and out-of-hours periods. The DfE's operating bands and BB101's criteria are the targets the control has to meet.

Ask for commercial HVAC quotes

SEO Dons Ltd publishes this site and replies to every enquiry by email. It does not install, service or maintain HVAC.

Get a quote
Get a quote