Leak detection of CO2 pipelines with simple atmospheric CO2 sensors for carbon capture and storage

This paper presents a field test performed with five relatively simple CO2 sensors (Vaisala Carbocap GMP343) that were placed for more than one year in a field in Ten Post, Groningen, The Netherlands. Aim was to investigate their potential use in monitoring pipelines transporting CO2 for carbon capture and storage. All sensors react differently on temperature changes which decreased signal to noise ratio for this application. The detection limit for leak detection was improved by a factor of 2.5 (the standard deviation of the average difference between two sensors decreased from 10 ppm to 4 ppm) with either laboratory or field data correction. Using laboratory calibration, the sensor drift requires that temperature response and calibration have to be re-established about every three to six months. Using field data for calibration is less expensive and labor-intensive, but has also several drawbacks, like the requirement of sufficient variability of the atmospheric parameters during the calibration period and the risk of “correcting” an actual leak during cross calibration. A release test of CO2 showed that a leak of >3 g/s would be easily detectable with sensors placed in a 70 m grid.