Cross-Sensitivity: Why Your Detector Reads Gas That Isn't There

An electrochemical sensor is not a lock that only one key opens. It is an electrode tuned to oxidize or reduce a target gas at a set potential, and other gases that react at a similar potential will also produce current. The instrument reports that current as a concentration of whatever gas the sensor is labeled for. This is cross-sensitivity, and it is one of the most common reasons a detector reads gas in an atmosphere where the target gas is not present.

Where it shows up

The pattern is predictable enough that most manufacturers publish a cross-interference table for each sensor. A few of the recurring offenders:

• H2S on a CO sensor. Hydrogen sulfide reads high on carbon monoxide sensors unless the sensor has a filter installed.
• Chlorine and chlorine dioxide. Cl2 and ClO2 sensors respond to each other, which matters in water treatment plants that use both.
• NO2 and ozone. Both are strong oxidizers and frequently show negative or positive interference on each other's sensors.
• Hydrogen on CO sensors. Battery charging rooms and electrolyzer areas routinely produce phantom CO alarms.

Interference can be positive (reads high) or negative (suppresses the reading). Negative interference is the dangerous one, because it makes a real hazard look smaller than it is.

How it corrupts calibration

Calibration assumes the sensor is seeing only the calibration gas. If your calibration is performed in an area with background contamination, or if the zero step is done in air that is not actually clean, the span factor you write to the instrument bakes that error in permanently until the next calibration.

Two practical consequences follow. First, zero air matters as much as span gas. Second, calibrating a chlorine sensor in a room where chlorine dioxide is in use, or an H2S sensor near a wet well, will produce a valid-looking calibration that is quietly wrong.

What to do about it

• Pull the cross-interference table for the specific sensor part number, not the instrument model. Sensors of the same gas from different suppliers behave differently.
• Calibrate in clean air, away from the process. If that is not practical, use a generator that delivers both a clean zero and a known span concentration at the sensor, so the baseline is controlled rather than assumed.
• Document known interferents on the entry permit for areas where two reactive gases coexist, so technicians reading an alarm know what else could be causing it.
• Test deliberately. If you suspect an interferent, apply it at a known concentration and record the response. That turns a suspicion into a documented sensor characteristic your team can rely on.

The short version

A reading is a current, not a gas. Knowing which other gases in your facility can produce that current is part of running a defensible detection program, and it starts with calibrating against a clean, known baseline every time.

Generators and sources for Cl2, ClO2, H2S, NO2 and O3 are listed under calibrators and sources, and ship the same day.