Electrochemical CO Detectors and the UL 2034 Time-Delay Curve
A three-electrode fuel cell oxidises CO at a platinum working electrode (CO + H2O → CO2 + 2H+ + 2e−), producing a current proportional to concentration. The chip integrates concentration × time before alarming because that models carboxyhaemoglobin accumulation — which is why UL 2034 mandates 70 ppm within 60–240 minutes and deliberate silence below ~30 ppm. That last provision leaves a real gap at 20–40 ppm chronic exposure, which low-level monitors cover.
A household carbon monoxide detector solves the same problem biology does — detecting an inert gas — with completely different chemistry. Biology binds CO to iron and reads a shape change. The detector binds it to **platinum** and reads electrons. ## The electrochemical cell The standard sensor is a three-electrode **electrochemical fuel cell**. Behind a gas-permeable membrane sit electrodes in an acid electrolyte. CO diffuses in and is catalytically oxidised at the platinum working electrode: CO + H₂O → CO₂ + 2H⁺ + 2e⁻ Oxygen is reduced at the counter electrode to complete the circuit. The liberated electrons constitute a current, and that current is **directly proportional to CO concentration**. It is a tiny CO-powered battery whose output is the reading. Two consequences worth knowing: the cell is consumed slowly by its own operation, so **sensors expire in roughly 5–7 years** regardless of battery state; and the electrolyte is affected by humidity extremes and by some solvent vapours, which is a source of false alarms. ## The time-delay curve, and why it exists **UL 2034**, the US standard, requires an alarm at: - **70 ppm** within 60–240 minutes - **150 ppm** within 10–50 minutes - **400 ppm** within 4–15 minutes The delays look like a design flaw and are the opposite. The chip integrates **concentration × time** before alarming, because that is how carboxyhaemoglobin accumulates in blood. The alarm is modelling your bloodstream rather than the air — a brief 70 ppm spike from a gas oven is harmless, and four hours at 70 ppm is not. See Carboxyhaemoglobin: Binding, Half-Life, and Reference Levels. The standard also deliberately requires alarms **not** to sound below about 30 ppm, to prevent nuisance activations. ## The low-level monitor gap That last provision creates a real gap most people are unaware of. Chronic exposure at **20–40 ppm** produces fatigue, headache, palpitations and cognitive fog — and a compliant UL 2034 alarm will never make a sound. Infants, elderly people, pregnant women and anyone with cardiac disease are affected at levels the standard alarm ignores by design. **Low-level CO monitors** alarm from around 9–25 ppm and display a live reading. They are the upgrade for anyone with a combustion appliance and unexplained chronic symptoms, and they are not a replacement for a UL-listed alarm — they are an addition, with more false alarms as the trade-off. ## The practical failure mode The commonest failure is not the absence of a detector. It is a detector with a dead battery that someone intended to replace, or a unit past its sensor life still sitting on the wall looking functional. Sealed **ten-year-battery** units on every level address both, because they eliminate the maintenance step entirely and expire as a unit. Placement should follow the manufacturer's instructions — CO mixes fairly evenly with air rather than reliably rising or sinking, so the folk rules about mounting height are less useful than the specific guidance.