I nearly sold my air purifier. In Port-Leucate, in a relatively clean environment, I noticed little difference outside pollen season and cooking. Then I installed it two weeks after arriving in Grenoble for my electrical engineering studies. I have asthma (though it has improved), allergies to pollen, cypress and roughly everything on Earth, even grass: so I sneeze very often outdoors or in class. At home, the sneezing has almost disappeared since the air purifier arrived.
The second difference appears when I cook. Fatigue (due to autism) already gives me a dry cough. Without a hood in my flat, some cooking makes it worse, especially red meat. When my purifier keeps its own indicator around 60 to 80 during less polluting cooking, I still cough less. I do not know the sensor's unit or accuracy, so I will not turn that number into a scientific measurement. The observation itself repeats.
of 0.3 µm particles captured by the US definition of a HEPA filter
US EPA
one-minute PM2.5 peaks measured after cooking in an experimental kitchen and living room
Liu 2021
covered area when the same airflow moves from one to five air changes per hour
CADR / volume calculation
99.9% tells you almost nothing about speed
Two units can both claim 99.9% and clean the same room at completely different speeds. That percentage measures what the filter captures in one pass. It does not say how much air crosses the filter each minute. A small unit with excellent filter media can clean a volume of air that is simply too small for the room.
The useful figure is CADR, or Clean Air Delivery Rate. The AHAM standard measures it separately for smoke, dust and pollen. A CADR of 250 m³/h means the unit delivers the equivalent of 250 m³ of air cleaned of the tested particles each hour. It measures neither gases nor odours.
Why 130 m² can become 26 m²
My unit is advertised for 130 m². It sits in a room of about 20 m² and, despite that gap, an air pollution spike does not disappear instantly: it takes five to ten minutes. There is no contradiction. Maximum coverage depends on how many times you expect the room's air to pass through the filter each hour, an assumption large marketing labels rarely show.
Take a 2.5 m ceiling. One hundred and thirty square metres contain 325 m³. If the unit processes that volume once per hour, its CADR is 325 m³/h. With the same airflow and a target of five air changes, coverage falls to 26 m². Both areas can be honestly calculated from the same motor. They simply do not promise the same speed.
The CDC advises aiming for five air changes per hour when possible to reduce airborne viral particles in buildings. That is not a universal standard for allergies or cooking, but it is a useful sizing reference. Oversizing can bring you closer to that airflow, or let you use a quieter setting each day and keep full power for spikes.
| Check | What it answers | What it does not answer |
|---|---|---|
| Smoke CADR | Delivery against fine particles | Gases and odours |
| Pollen CADR | Delivery against larger particles | Your clinical benefit |
| Advertised area | Coverage under one assumption | Air changes if the assumption is missing |
| HEPA class | Single-pass filter efficiency | Room cleaning speed |
| Carbon weight | Likely capacity for some gases | Exact time to saturation |
| Noise in dB | Level at a stated setting | Noise in your room at another distance |
The smaller-particle trap
It is easy to picture a filter as a sieve: anything smaller than the hole passes through. HEPA does not work that way. Particles are captured when they meet fibres through interception, impact or diffusion. The smallest move more erratically through the air, increasing their chance of hitting a fibre.
That is why 0.3 µm is the worst case, the most penetrating particle size in the EPA definition. The filter captures at least 99.97% at that size. Larger and smaller particles are trapped even more efficiently. Comparing a claim of 99.9% at 0.1 µm with 99.97% at 0.3 µm, without knowing the protocol, cannot establish that the first filter is better.
HEPA for particles, carbon for some gases
HEPA targets airborne matter: pollen, dust, spores, smoke and fine particles. It does not properly capture the gaseous molecules behind many volatile organic compounds and smells. The EPA guide therefore separates particle filters from gas filters. To address both, a unit generally combines HEPA with activated carbon.
Carbon adsorbs some molecules onto its surface. Capacity depends on carbon mass, the compound, humidity and time to saturation. A thin black sheet does not have the reserve of a deep bed of pellets. In my home, the displayed particles fall while the cooking smell often remains. Your nose and the particle sensor are not measuring the same pollution.
Cooking creates a spike before the unit catches up
In an apartment experiment, one-minute PM2.5 concentrations peaked between 200 and 1,400 µg/m³ in the kitchen and living room one to seven minutes after cooking. The figure varies widely with food, oil, temperature and ventilation. For context, the WHO guideline is 15 µg/m³ over twenty-four hours. A brief peak and a daily average do not represent the same risk, but the comparison shows how intense the source can be for a short time.
Another study pan-fried 100 g of pork for nine minutes in thirteen homes. Median PM2.5 over two hours reached 4.5 mg/m³ without ventilation and 0.5 with the hood. That precise protocol does not mean every kitchen becomes the most polluted place in the world. It supports a narrower statement: high-emission cooking without extraction can briefly produce an extremely high particle exposure.
The order is therefore simple. Cover the pan, lower temperature when the recipe allows, run a hood that vents outdoors, then open a window if outdoor air is cleaner. The purifier treats what has already escaped the source. Without a hood in my home, it limits accumulation and speeds the decline, but it does not stop pollution during cooking.
What I actually observed
In Port-Leucate, I mainly noticed an effect during cooking and pollen season. The rest of the time, it felt dispensable enough that I considered selling it. Near Grenoble, in a basin where pollution can linger, the difference became obvious: I sneeze often away from home and much less in the filtered room.
This is a repeated observation, not a trial. Many things changed between the two cities, and I did not alternate a working unit with a sham unit blindly. It remains consistent with what the filter can remove. In a randomised trial among children with asthma in an agricultural area, two HEPA units reduced PM2.5 by 60% in the bedroom and 42% in the living room, with no meaningful reduction in ammonia, a gas.
Moving from particles to symptoms requires more caution. In 90 adults with allergic rhinitis, a double-blind trial found lower medication use in the purifier group, without significant improvement in every symptom score. I can say that I sneeze less at home. I cannot promise that a purifier will treat someone else's asthma or allergies.
What it costs me
Mine was listed at about €300 and I bought it for €200 during a sale. The filter costs around €60 and lasts about a year in my use. I also vacuum the pre-filter occasionally. These are not reference prices, only my actual cost: roughly €5 a month after purchase if the filter does last twelve months.
That expense does not have the same value everywhere. In a low-pollution area, without allergies and with effective kitchen extraction, the gain may be hard to notice. With asthma, multiple allergies, pollen or a city where particles accumulate, the exposure reduction is more likely to be visible. The right purchase does not depend on the unit alone. It depends on what actually enters the room.
What this article does not say
- That an air purifier treats asthma or allergies. It reduces some airborne pollutants. Medical treatment answers a different question.
- That my sensor measures PM2.5 accurately. I report its indicator as an internal trend, without a confirmed unit or calibration against a reference instrument.
- That a hood is unnecessary. Extraction and ventilation remove pollution at the source. Filtration acts after emission.
- That carbon removes every smell. Its effectiveness varies by gas and quantity, then falls as it saturates.
- That advertised coverage is necessarily dishonest. It may assume one air change per hour. The hidden assumption is what makes it useless on its own.
- That my experience across two cities isolates pollution. Housing, season, pollen and my activities changed too. What remains is a clear, repeated fall in sneezing inside the filtered room.
« An excellent filter that sees the air once an hour remains slower than a good filter that sees it five times. »
What Helix would do
Air quality only becomes meaningful when connected to the rest: city, pollen season, cooking, symptoms, sleep and treatment. Helix would not turn a consumer sensor reading into a diagnosis. It would look for whether sneezing, dry coughing or sleep change when particles fall, within the same person over several weeks.
Connect your environment with what your body actually does, against your own baseline.
Join the beta