Home / LEV Testing / Case study
Case study · Car parks · Vehicle exhaust and CO
An underground car park relied on mechanical extraction and carbon monoxide detection to keep vehicle fumes to safe levels. The airflow largely met design - but the sensors that are supposed to switch the fans up had quietly drifted out of calibration. Here is why that is the dangerous kind of fault.
The site
The car park sits below a busy building, enclosed on all sides, with cars circulating and idling throughout the day. Its ventilation is mechanical: extract fans that clear vehicle fume, driven by a carbon monoxide detection system that is meant to run the fans harder as the air gets dirtier and ease them back when it clears.
That demand-controlled arrangement is efficient and sensible - there is no point running big fans flat out over an empty car park at three in the morning. But it moves the whole burden of safety onto the sensors. The fans do not know the air is foul; the carbon monoxide detectors tell them. If the detectors are wrong, the fans do the wrong thing, and nobody in the car park has any way of knowing.
The examination set out to answer two questions: do the fans move the air they were designed to move, and does the detection that controls them actually work?
The hazard and the law
The hazard here is vehicle exhaust in an enclosed space, and the headline substance is carbon monoxide - colourless, odourless, and dangerous precisely because the human body gives no warning of it. Nitrogen dioxide and fine particulate come with it. Building guidance for car parks aims to hold carbon monoxide down to a low average, with tighter limits on the short-term peaks that occur near ramps and exits, and modern car parks link the ventilation rate directly to carbon monoxide readings so the fans match the pollution.
The duty to control exposure to a substance hazardous to health applies to the people who use and work in the car park, and the way you discharge it is a ventilation system that demonstrably holds the concentration down. "Demonstrably" is the operative word: the system has to be examined, its airflow measured, and - because this is a controlled system - its control proven to work. A set of fans without working sensors is not a control; it is a hope.
What we tested
We measured each extract fan against its design airflow and inspected the ductwork, grilles and discharge, confirming fume was taken clear of the building rather than recirculated. That is the half of the job people expect.
The more revealing half was the control system. We checked the carbon monoxide detectors for calibration - were they reading true against a known reference - and we tested the logic: when carbon monoxide rose, did the fans actually ramp up as they should, and did they ease back as it cleared? On a demand-controlled car park, this is where safety lives or dies, and it is the part a quick visual inspection cannot see.
What we found
| Check | Finding | Result |
|---|---|---|
| Extract fans vs design airflow | Most to design; one fan running around 15% short of duty | Remedial |
| CO detection calibration | Sensors drifted - reading low against reference | Fail |
| Demand-control response | Fans slow to ramp up because sensors under-reading | Remedial |
| Ductwork and discharge | Clear of the building, no recirculation | Pass |
On paper the car park looked fine - the fans largely met their design airflow, and the ductwork and discharge were sound. But the carbon monoxide sensors had drifted and were reading low, which meant the demand control was not calling the fans up when it should. The ventilation had the capacity to clear the air; its controller simply was not asking it to. One extract fan was also running about 15% short of its design duty. The overall assessment was a pass on airflow but with remedial actions required on the control system - the part that actually keeps people safe.
The remedial actions
The outcome
With the sensors recalibrated, the demand-control response proven and the short-running fan brought back to duty, the car park's ventilation now responds to the air it is actually dealing with. Just as importantly, the maintenance regime gained a recurring carbon monoxide calibration check - the one task whose omission the examination had exposed.
Demand-controlled ventilation is efficient, but it concentrates all the risk in the sensors. A car park can have healthy fans, clean ducts and a clear discharge and still let carbon monoxide build, simply because the detectors have drifted and are telling the fans to stay quiet. The gas gives no warning, so the only protection is a control system that is tested and sensors that are calibrated. Checking the fans alone is checking the easy half; the half that matters is whether the system knows when to run.
Sources and further reading
The duties, exposure limits and control standards described here come from the Health and Safety Executive. For your own process, always work from the current HSE guidance:
Questions
Because vehicle exhaust builds up in an enclosed space. Petrol and diesel engines give off carbon monoxide and nitrogen dioxide - carbon monoxide is colourless and odourless and can reach harmful levels without anyone noticing. Mechanical extraction, usually linked to carbon monoxide detection, keeps the concentration down to safe levels for the people using and working in the car park.
Building guidance aims to keep carbon monoxide low - commonly an average in the region of 30 ppm over an eight-hour period, with short-term peaks near ramps and exits limited. The exact figures should be read from current guidance, but the principle is the same: the ventilation has to hold the concentration down, and carbon monoxide detection is how the system knows when to work harder.
We measure the extract fans against their design airflow, check the ductwork and discharge, and - critically - test the carbon monoxide detection system: are the sensors reading accurately, and do the fans actually ramp up when carbon monoxide rises? A car park with demand-controlled ventilation is only as safe as the sensors that drive it.
Because the sensors are the brain of the system. If they have drifted and read low, they will not call the fans up when fumes build, so the ventilation sits idle while carbon monoxide climbs - and nobody can smell it. Airflow that meets design means nothing if the control that triggers it has gone deaf. A quiet calibration drift is exactly the kind of fault that hides until it matters.
It is general mechanical ventilation rather than a capture hood at a single source, but it is examined on the same principles - measured airflow, a working control system, a clear record - and it controls exposure to a substance hazardous to health. Whether you file it under LEV or workplace ventilation, the duty to keep people from harmful carbon monoxide, and to prove the system works, is the same.
Extract performance should be examined on a regular interval in line with the control-of-exposure duty, and carbon monoxide sensors need routine calibration - they drift over time and are the one component whose failure is invisible. Building a sensor-calibration check into the maintenance schedule, not just an annual glance at the fans, is what keeps the system honest.
Tell us your capture points and the process they control, and we will quote the thorough examination and test to COSHH and HSG258.
