A seated adult breathes out somewhere around 20 litres of carbon dioxide an hour. Put forty of them on a floor with the fresh air turned down to save on heating, and by three in the afternoon the room is running at two and a half times the concentration of the street outside.
That is the number behind every stuffy meeting room complaint I have ever been asked about, and it has nothing whatever to do with cleaning. I get asked anyway, because the cleaner is the only supplier most tenants see, and because the word hygiene has stretched over the last few years to cover things it was never meant to cover. So this is a piece about where the line falls between my trade and the mechanical engineer’s, written by someone who has an obvious commercial interest in drawing that line generously and is going to draw it straight instead.
What is a CO2 reading telling you?
Carbon dioxide at office concentrations is a proxy rather than a hazard. People exhale it, ventilation dilutes it, so the level in a room tells you how much outdoor air is arriving relative to how many people are in there. It is the cheapest useful measurement in building services.
The reference points are well established. CIBSE takes 600 to 800 parts per million as the mark of a reasonably well ventilated room, and treats anything above 1,500 ppm as a very poorly ventilated space. The HSE’s position runs alongside it: readings consistently over 1,500 ppm in an occupied room mean the ventilation needs attention, and roughly 1,000 ppm corresponds to about 10 litres per second per person. That figure is also the whole-building standard in Part F, which asks for 10 litres per second per person or one litre per second per square metre, whichever comes out higher. The HSE’s Approved Code of Practice sets a lower floor still, saying fresh air should not normally fall below 5 to 8 litres per second per occupant.
Outdoor air in London sits somewhere over 400 ppm, so you are never starting from zero.
1,500 ppm is not a poison reading
Tenants who buy their first monitor tend to panic at the wrong thing. A reading of 1,400 ppm is not making anyone ill in the toxicological sense. Occupational exposure limits for carbon dioxide sit at thousands of parts per million, far above anything an office will produce.
What that reading tells you is that the air in the room is being recycled through the lungs of the people in it, and that everything else in the air is concentrating at the same rate. Volatile compounds off the furniture, cooking smells, whatever came in off the road, the aerosols from the person with a cough by the window. Carbon dioxide is the receipt for all of it.
The air is not a surface, and that distinction runs through the rest of this article.
Can cleaning change the air in an office?
Yes, at the margins, and the margins are worth having.
Most of what floats around an office floor has been on the floor first. Settled dust is a reservoir: skin cells, fibres, the fine grit that walks in on shoes, and the particulate that came off the South Circular and through the door. Footfall lifts it, and a badly run cleaning operation lifts more of it than the occupants do.
Damp dusting rather than dry keeps particulate captured instead of redistributed. Microfibre holds fines that a duster flicks into the air. Regular attention to the horizontal surfaces nobody sees, particularly above head height, removes the reservoir before it can be disturbed. Kitchen and washroom extract grilles clogged with grease and lint reduce the extract rate they were designed for, and cleaning them restores something a mechanical contractor would otherwise be called out to diagnose.
None of that changes the ventilation rate by a single litre per second. It changes what is available in the room to be circulated by it.
The vacuum is the machine that matters
If a client wants one cleaning intervention that shows up in a particulate reading, it is the vacuum specification.
A cylinder machine with a tired bag and a poor exhaust filter collects the coarse material and fires the fine fraction straight back out at ankle height, which is precisely the fraction that stays airborne longest and reaches deepest into the lung. HEPA-filtered machines with sealed bodies do not do this. The seal matters as much as the filter grade, since air that bypasses a filter through a badly fitting housing has not been filtered at all.
Ask your contractor what machines are on your site and what the filtration is. It is a fair question, most specifications do not answer it, and the difference in cost between the right machine and the wrong one is a few hundred pounds spread over several years.
Where does cleaning make the air worse?
Here is the part of the trade that does not get discussed at facilities exhibitions.
Cleaning chemicals are a source of volatile organic compounds. Solvent-based products, aerosol sprays, and anything strongly fragranced put measurable quantities of material into the air of a room, and a floor cleaned at seven in the evening with the air handling switched off at six sits in its own emissions until the plant restarts in the morning. The first arrivals breathe the concentrated version.
Trigger sprays make it worse than it needs to be by aerosolising product that could have been applied to a cloth. Decanting into a damp cloth and wiping, rather than misting a surface at head height, is a change that costs nothing and takes a real quantity of airborne material out of the room. Where a building runs air cleaning devices as well, the interaction between what those produce and what we spray deserves more scrutiny than it gets.
The scent that gets mistaken for hygiene
My industry sells smell as evidence of work, and I think it is one of the worse habits we have.
Plug-in air fresheners, scented washroom units, citrus-heavy floor products: all of them exist because a client walks in and judges the clean by the nose. What they are doing chemically is adding compounds to the air of a room that was already short of fresh air, and calling the result improved. Limonene, the compound behind most citrus fragrance, is reactive, and the chemistry it takes part in indoors is not something anyone can wave away as pleasant.
A properly cleaned office smells of nothing. That is the target, and it is a harder one to hit than a floor that smells of lemons.
What should be measured, and what should be ignored?
Buy a CO2 monitor and take it seriously. Buy one with a non-dispersive infrared sensor, which measures carbon dioxide directly. The cheap end of the market sells devices reporting something labelled eCO2 or equivalent CO2, which is inferred from a volatile compound sensor rather than measured, and the numbers those produce in a room where somebody has just cleaned or opened a marker pen are close to meaningless.
Log over a fortnight rather than glancing at a display. What you want is the shape of the day: where the curve sits at nine, at eleven, after lunch, at four, and whether it ever comes back down. A room that starts at 500 ppm and finishes at 1,600 ppm every day has a ventilation rate that was designed for fewer people than are now sitting in it.
Ignore anything sold on the promise of sterilising your air, at least until it comes with dose figures and independent room-scale data rather than a laboratory certificate.
Where the sensor goes
Not on a desk beside somebody’s face, where their own breathing dominates the reading. Not next to a window or a door, where dilution flatters it. Not on top of a warm server cabinet.
Roughly breathing height, a metre or so off the floor, in the occupied body of the room, away from supply diffusers. One sensor per distinct space, since a meeting room and the open floor outside it behave nothing alike. Put a sensor in the wrong place and you will get a number that is precise, repeatable and about the wrong volume of air.
Who should you call about an air quality complaint?
Not me, most of the time.
Persistent stuffiness, headaches through the afternoon, a floor that everybody describes as airless: those are ventilation problems and they belong to a mechanical services contractor. The questions to ask are whether the air handling unit is running at its designed fresh air fraction or has been dialled back for energy, when the supply filters were last changed and to what grade, whether the extract is balanced against the supply, and whether the occupancy of the floor has risen since the system was commissioned. That last one catches a lot of London buildings, where a floor designed for eighty people now holds a hundred and thirty desks.
Spend the money there before you spend a penny with me. The air is not a surface, and a cleaning contract is not a ventilation strategy however it is dressed up.
What we changed off Clarence Street
A tenant on a second floor off Clarence Street in Kingston, KT1, called us about air quality last autumn, convinced the cleaning was the problem. It was not. Their monitor was showing 1,700 ppm by mid-afternoon in a room with sealed windows and an air handling unit somebody had set back years earlier.
We did make three changes on our side. Machines with sealed HEPA filtration replaced two elderly vacuums. The plug-in fragrance units came out. And the cleaning slot moved to start while the plant was still running rather than an hour after it shut down, which cost the landlord about forty minutes of fan power a night.
The 1,700 ppm did not move at all. That took a controls engineer and a fresh air damper that had been shut since 2019. The air is not a surface, and the cleaner is rarely the person holding the answer.
