Do You Need a Carbon Monoxide Detector?

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

When the answer is yes, and when it is genuinely no

The physics is simple. Carbon monoxide is produced whenever a carbon-based fuel burns with insufficient oxygen. That includes natural gas, propane, heating oil, wood, coal and petrol. A home with a gas boiler, a gas hob, a wood stove, an open fire or an attached garage therefore has at least one plausible CO source. A home with none of those, heated and cooked entirely by electricity, with a detached garage or no garage, has no combustion source inside the thermal envelope and the alarm has nothing to detect. That is the honest no. It is uncommon in the UK and Ireland, where most homes have a gas supply, but it exists.

The risk is not evenly distributed. The dangerous cases cluster around appliances that are badly installed, poorly serviced, or operated in spaces not designed for them. A boiler flue that has been boxed in, a gas fire with a blocked throat, a portable generator run in a garage with the door shut, a charcoal barbecue brought indoors, a wood burner with a cracked flue liner. In each case the CO is produced continuously and accumulates in still air. The alarm earns its place because CO gives no warning: you cannot smell it, see it or taste it, and the early symptoms, headache and tiredness, are indistinguishable from a hundred other things. By the time the exposure is heavy enough to be obvious, the person affected is often too confused to act.

Where people get this wrong is assuming the alarm covers them for every gas. It does not. A CO alarm responds to carbon monoxide and nothing else. It will not warn you about a natural gas leak, which is why a home with gas appliances needs the supply smell (the odorant added to the gas) and, in some cases, a separate combustible gas detector. The two instruments are different and neither substitutes for the other; the comparison is set out in gas leak detector versus carbon monoxide detector. A CO alarm also will not detect carbon dioxide, which is a different gas with a different hazard profile.

What the sensing cell inside a home alarm is actually doing

Almost every consumer CO alarm uses an electrochemical cell. Inside is a small sealed assembly with two or three electrodes, a liquid or gel electrolyte, and a porous membrane that lets gas reach the working electrode. Carbon monoxide arriving at that electrode is oxidised, which releases electrons, and the resulting current is proportional to the CO concentration in the air. The cell produces a tiny current, in the nanoamp to microamp range, which the electronics amplify and compare against threshold curves. The cell is selective for CO because the electrode potential is chosen so that other common gases either do not react or react weakly.

The alarm does not simply trip at one number. It applies a time-weighted exposure rule, because the health effect of CO depends on concentration multiplied by time. A low concentration breathed for hours can be as dangerous as a high concentration breathed for minutes. The exact curves differ between standards and between manufacturers, and the listing bullets for the products here do not publish the threshold table. What the bullets do state is the audible output: the Kidde COBD, the First Alert plug-in and the Kidde plug-in with battery backup all publish an 85 decibel alarm. That figure is a sound pressure level, not a detection threshold, and it is the number most often misread by buyers as an accuracy claim.

The digital display is a second, separate function. On the Kidde COBD the display shows status and alarm operation, and on the First Alert plug-in alarm a peak level feature displays the highest concentration measured. That peak memory is genuinely useful for a technician or a homeowner trying to work out whether an intermittent alarm was real, because it retains the highest reading since the last reset. But it is a peak of what the alarm sensed, not a calibrated measurement, and the listing does not publish the accuracy of that display. Treat it as a diagnostic hint, not a number you can put in a report.

The exposure arithmetic that explains why the alarm waits

Buyers often complain that an alarm did not sound during a small, persistent smell or a headache that turned out to be CO. The reason is that the alarm is tuned to protect, not to inform. The table below sets out the general shape of the relationship, using standard occupational exposure reference points rather than any product’s published curve. These are domain reference values, not figures from the listings.

The practical consequence is that a low-level chronic leak, the kind that produces a headache every morning and clears when you open a window, can sit below the alarm threshold indefinitely. That is not a fault. It is the design. The alarm is built to avoid nuisance trips from cooking and from the small amounts of CO produced by a gas hob, and the cost of that choice is that it will not tell you about a persistent 20 ppm background. If you suspect that kind of exposure, the instrument you want is a CO meter with a published accuracy and a logging function, not an alarm. The distinction is explained in carbon monoxide meter versus CO alarm.

Placement, power and the maintenance nobody mentions

Placement is where most installations go wrong. CO mixes with air at roughly the same density, so it does not stratify like natural gas or pool at the floor like propane. The old advice to mount alarms near the ceiling is not wrong, but the more important rule is proximity to the sleeping area. The alarm exists to wake people, so it belongs in or just outside each bedroom, and on every level of the home including the basement. A single alarm in the hallway downstairs does not reliably wake a heavy sleeper upstairs through a closed door. The listing for the Kidde COBD states that it can be wall-mounted or set on a table, which matters for renters and for travel, because a table placement beside a hotel bed is closer to the sleeper than a wall fixture across the room.

Power choice is a real decision. Battery-only units such as the COBD run on two AA batteries and keep working during a power cut, which is exactly when a generator or a back boiler might be running. Mains plug-in units such as the Kidde plug-in with 9V backup and the First Alert plug-in draw from a standard outlet and fall back to a battery when the supply fails. The plug-in form factor is convenient but has one failure mode nobody mentions: the outlet gets switched off, or the unit gets unplugged to free a socket for a vacuum cleaner and never goes back. A battery-only unit cannot be unplugged by accident. A plug-in unit with a sealed backup battery is a reasonable compromise, but check the backup battery is actually fitted and not still in its tab.

The maintenance nobody mentions is the sensor itself. Electrochemical cells age. Their output falls slowly and then faster, and the alarm’s end-of-life signal is the only warning you get. The listing for the Kidde smart detector and air quality monitor states that its voice alarm announces “Replace Alarm” when the unit has come to its end of life span. That is a useful feature because it removes the guesswork, but it does not remove the need to press the test button monthly and to note when the unit was installed. The test button does not prove the cell is still accurate; it proves the electronics, the sounder and the battery are working. The distinction is set out in best carbon monoxide detectors, and the button test itself is covered in best plug-in carbon monoxide detectors.

Where the alarm stops being the right instrument

The alarm is a life-safety device and it is good at that job. It is not a diagnostic tool, and there are four situations where buying one instead of the correct instrument is a mistake.

• You are investigating a suspected low-level leak. The alarm is deliberately insensitive to chronic low concentrations. A calibrated CO meter with a published accuracy and a data log is the correct tool. The options are compared in best carbon monoxide meters.
• You are a technician commissioning or servicing a combustion appliance. You need a flue gas analyser, which measures CO and CO2 in the flue and calculates combustion efficiency. A room alarm is not part of that procedure and its reading is not admissible for commissioning.
• You are worried about a gas leak rather than CO. A CO alarm will not respond to methane. You need a combustible gas detector, and the difference is explained in does a carbon monoxide detector detect natural gas.
• You want to monitor general air quality. CO is one pollutant among many. A unit such as the Kidde smart detector and air quality monitor combines CO detection with an indoor air quality monitor, which is a different measurement with a different sensor. That combination is useful, but it is not a substitute for a dedicated CO meter when you need a number you can defend.

There is one more case worth naming. A CO alarm is not a substitute for fixing the appliance. If an alarm sounds, the correct response is to ventilate, turn off the fuel, leave the building and call the emergency gas service. The alarm has done its job. It is not a licence to keep running a boiler that has tripped it, and it is not a reason to delay servicing. The instrument that tells you the appliance is safe is the technician’s analyser, not the alarm on the wall.

The short version for a buyer standing in a shop

If your home burns any fuel, buy a CO alarm for each sleeping level and press its test button monthly. If you have no combustion appliance and no attached garage, you do not need one, and the honest answer is that the alarm would have nothing to detect. If you are trying to measure CO rather than survive it, buy a meter. The alarm protects people; the meter informs decisions. They are different instruments and the best installations have both, in the hands of the people who need each one.

Frequently Asked Questions