A compensation model for an NDIR-based CO₂ sensor and its energy implication on demand control ventilation in a hot and humid climate

Non-dispersive infrared absorption spectroscopy (NDIR)-based CO₂ sensors currently dominate the market for building applications. Yet, the affordable ones are frequently found to have a reading that deviates from the actual CO₂ concentration, which hinders the proper operation of the demand control ventilation (DCV). This study aims to propose a compensation method to mitigate the CO₂ sensor deviation due to water vapor interference. Its energy implication is simulated afterward by creating a 464 m² office space model equipped with the DCV in EnergyPlus. Per the results of multiple standard gas experiments performed in a controlled environment, the proposed model has a coefficient of determination R²=0.899, and is able to reduce the normalized root mean square error from 8.1 % (before compensation) to 2.6 % (after compensation) for a selected NDIR CO₂ sensor, showing a 67.9 % reduction. Meanwhile, the tested sensor is originally prone to underestimate the CO2 concentration, leading to less ventilation rate in DCV. As a result, there will be a 6.97 % increment in the annual ventilation electricity load of the modeled building after compensation, but in exchange for a better controlled indoor CO₂ concentration that was underestimated by 3.0 to 13.1 % (on average 7.9 %) depending on zone conditions before compensation.