Exponential pattern recognition for deriving air change rates from CO2 data

A novel procedure for automated determination of air change rates from measured indoor CO2 concentrations is proposed. The suggested approach builds upon a new algorithm to detect exponential build-up and decay patterns in CO2 concentration time series. The feasibility of the concept is proved with a test run on synthetic data that shows a good reproduction of the previously defined air change distribution. The demonstration continues with test runs on CO2 datasets measured in the kitchen and the sleeping room of two residential buildings. The derived air change rates were within the expected distributions and ranges in both cases when natural or mechanical ventilation was used.