Realtime analysis of carbon dioxide concentration curves for intelligent ventilation control

Abstract

This thesis is an investigation into the intelligent control of on-demand ventilation systems using carbon dioxide concentrations to determine occupancy patterns, the building’s natural infiltration rate and as a proxy for indoor air quality (IAQ). The suitability of using CO₂ sensors to continuously monitor concentration changes in a ventilation zone in order to determine occupancy and natural infiltration rate is investigated. A heat recovery ventilator (HRV) was modified by replacing the factory electronics with a designed controller that enabled variable air flow rates and direct control of recirculation via direct computer control. Apparatus was built to centrally sample the incoming fresh air and outgoing exhaust flows for CO₂ which was measured using a non-dispersive infrared (NDIR) sensor. An additional controller was designed to select air streams for sampling, command the HRV and send data to a data acquisition system. Data was logged for several months into a database and analysed. CO₂ concentration decays that highly correlated with a theoretical decay curve using a moving window technique were studied as an indicator of vacancy while the decay constants of the decay curves were used to calculate the natural infiltration rate. The results demonstrate that building ventilation zone occupancy schedules and natural infiltration rates can be accurately determined via the analysis of CO₂ concentration decay curves. These techniques allow a control system to automatically adjust to changing usage patterns without manual reprogramming. The natural infiltration rate of a ventilation zone would normally be unknown, making prediction of future CO₂ concentrations difficult. These calculated parameters allow intelligent, predictive ventilation control schemes which have the potential to minimize ventilation energy losses to their theoretical limits.