Do Radon Detectors Work? What They Actually Measure

As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.

What a radon detector is actually counting

Radon is a gas produced by the decay of uranium in soil and rock. It is chemically inert, so it moves through foundations and floors and accumulates indoors. The radon atom itself decays into a series of short lived progeny, and that decay emits alpha particles. A radon detector does not smell the gas or weigh it. It counts alpha particles, either in a chamber where the gas diffuses, or on a filter where progeny collect.

That counting is the whole mechanism. Each detected alpha is one decay event. The detector divides the count by the volume of its chamber and the time it collected, and reports a concentration. The arithmetic is simple; the statistics are not. Radioactive decay is a Poisson process, so the uncertainty of a count falls with the square root of the number of counts. A ten minute window might collect a few dozen events, which is enough to show that radon is present but not enough to pin the level to a tight band. A week of counting collects thousands, and the estimate tightens.

This is why two detectors in the same room can disagree for an hour and agree over a month. Neither is broken. They are sampling the same random process with different patience. A listing that quotes a percentage accuracy after a fixed period, such as the X-Sense XR0B-SR claim of 95 percent after seven days or more, is describing how the estimate converges, not a lab grade calibration certificate.

Why a fast result is not a wrong result, but it is a preliminary one

Several consumer units advertise a first reading in minutes. The GZAIR radon detector states a first data point in ten minutes and reliable data after one hour. That is plausible for a pulse ionisation chamber with high sensitivity, because a sensitive chamber collects more counts per minute and reaches a usable estimate sooner. The listing quotes 30 counts per hour per picocurie per litre, which is a sensitivity figure, and says it is fifteen times the minimum professional standard. That number describes how many counts the chamber yields for a given concentration, and more counts per unit time means less statistical noise per unit time.

But sensitivity does not repeal the physics. A one hour average at a low concentration still rests on a small number of counts. It is a valid short term indicator, useful for finding a suspect room or checking whether a mitigation system is running, and it is not the same as a long term average. The same listing also states that it displays one day and long term averages, which is the number a homeowner should actually act on. The fast number is a convenience, not a substitute.

The practical rule: use the fast reading to decide where to keep monitoring, and use the long term average to decide whether to mitigate. A single elevated morning reading after a pressure change is not a mitigation trigger.

The sensor types and what each one can and cannot do

There are three sensor families in consumer and professional radon instruments, and they set the accuracy floor before any percentage is printed.

• Ionisation chamber. A chamber with an electric field collects charge from alpha decays. It is sensitive and responds quickly, which is why fast reading claims usually come from this type. The GZAIR unit names a pulse ionisation chamber. The trade off is that the chamber must be kept clean and the calibration must be maintained, and the listing says the sensors are calibrated but does not publish a calibration interval.
• Alpha track or electret. A plastic film or electret is exposed for weeks or months and etched or read afterwards. This is the classic long term integrating method. It is slow by design and very stable, and it is the reference method behind many building surveys.
• Scintillation or photodiode based continuous monitors. A phosphor or photodiode counts light flashes from alpha interactions. These are common in continuous monitors that log data and produce averages. The Airthings Corentium Pro is a continuous monitor of this class, and the listing states it is fully AARST/NRPP certified, which is a statement about the device meeting a recognised proficiency programme rather than a raw accuracy number.

None of these sensors measures the radon concentration instantaneously. Each integrates over a window. The difference is how long the window is and how the device reports it.

What the published numbers let you compare, and what they do not

Buyers want a single accuracy figure. The listings rarely give one that means what the buyer thinks. The table below collects only what the supplied listings state, with gaps marked.

The gaps are the point. A certification such as AARST/NRPP is a process statement: the device has been through a proficiency programme. It is not the same as a published accuracy band. A 95 percent figure after seven days is a convergence claim, not a calibration tolerance. A sensitivity figure in counts per hour per unit concentration is a physical property of the chamber, and it tells you how quickly the estimate settles, not how close it is to truth. Where a listing is silent, the honest answer is that the figure is not stated in the listing bullets and cannot be compared.

One further caution: a consumer humidity style accuracy figure of half a percent, if it appeared on a radon listing, would more likely be display resolution than measurement accuracy. Radon counting statistics do not support that kind of precision in a short window.

Placement, averaging and the maintenance nobody mentions

A detector measures the air that reaches its chamber. Radon concentration varies by floor, by room, and by the pressure difference between the soil and the house. A basement reading is usually higher than a first floor reading because the gas enters low and dilutes upward. A monitor placed next to a draughty window reads lower than one in still air. The instrument is not wrong; it is sampling a different micro environment. This is why the guidance for a home test is to place the detector in the lowest lived in room, away from external walls, drafts and heat sources, and to leave it there.

Averaging is the second half of placement. A long term average smooths the daily swing caused by weather and heating. The Safety Siren Pro4 listing states that it provides real time readings plus short term and long term averages, and that it has three selectable audible alarm modes plus a visual alert for the hearing impaired. That alarm structure is useful because a radon level can rise without anyone watching a screen. But an alarm is not a measurement; it is a threshold crossing. The long term average remains the decision number.

Maintenance is where consumer units differ most. A battery unit such as the X-Sense XR0B-SR claims two year battery life and low power notifications, which removes the most common failure mode, a monitor that quietly stops. A plug in unit depends on mains power and on the socket not being switched off. A professional unit such as the Airthings Corentium Pro is described as portable and tamper resistant, which matters when a reading may be used in a transaction. None of the listings publish a calibration interval, and that is the maintenance item buyers should ask about before purchase.

Where a radon detector stops being the right instrument

A continuous monitor is a screening and trending tool. It is not a substitute for a certified laboratory test when a result has to stand up in a property transaction or a regulatory context. The reference method is a passive detector exposed for a defined period and read by an accredited lab. A consumer monitor can tell you whether to order that test; it cannot replace it.

A radon detector also does not measure the things people sometimes expect it to measure. It does not detect carbon monoxide, it does not measure air quality in the general sense, and it does not tell you whether a mitigation system is working unless you compare before and after averages over comparable periods. If the question is whether a sump fan is running, a pressure gauge or a smoke test answers it faster.

Finally, a detector is not an alarm in the safety sense. A threshold beep is a prompt to look at the average, not an evacuation signal. Radon risk is cumulative over years, so the correct response to a high reading is to confirm it with a long term measurement and then act on the building, not to leave the room.

For buyers comparing specific models, the selection logic is set out in how to choose a radon monitor and the room by room placement question in radon monitors for basements. If a lab grade confirmation is the next step, see radon test kits.

The honest answer to whether radon detectors work

They work, in the same way a rain gauge works: they measure a real quantity with a known uncertainty, and the uncertainty shrinks with time and care. A radon detector that reports a long term average, sits in the right room, and keeps running is a genuinely useful instrument. A radon detector bought for a ten minute number is a novelty. The listings that publish certification, sensitivity or averaging behaviour give you something to reason about; the ones that publish only a percentage do not.

Frequently Asked Questions