Methodologies for evaluating demand-controlled ventilation retrofits in HVAC applications

The retrofit of a building or HVAC system with energy savings technologies requires upfront analysis to determine the potential for savings and post-retrofit analysis to verify the savings. The focus of this research project was in the development and evaluation of methods for analyzing demand-controlled ventilation (DCV) retrofits on smaller scale commercial buildings. The analysis of a DCV control retrofit requires predictions of CO2 levels in the occupied zones, building thermal characteristics and HVAC equipment performance. Methods for determining CO2 levels in occupied zones were developed and evaluated for forward simulation and calibrated simulation modeling approaches. A quasi-equilibrium model was found to be sufficient for analysis of the energy savings potential with a DCV retrofit. A methodology was developed for estimating schedules for CO2 generation rates and flow parameters using short-term testing. The method determines representative average values of occupancy for each hour and statistics that characterizes day-to-day variations. Benefits are shown with the use of tuned models for the evaluation of DCV retrofit potential. Training periods of only two days to one week are sufficient for estimating energy cost differences between pre- and post-retrofits for economic evaluation of DCV. However, longer training data periods of up to eight weeks might be necessary if the goal is to make accurate absolute predictions for a given site for detailed engineering evaluations or performance measurement and verification of a retrofit. A site with higher variability in occupancy levels, such as a restaurant, might require a minimum of eight weeks of training data to accurately model the CO2 levels in the zone. Expressed in terms of the coefficient of variation, errors in predicted CO2 concentrations ranged from 4% to 35% depending on the sites and length of training data period. The predicted frequency of time that CO2 concentrations were within a given range agreed well with field measured data. Improvements in the coefficient of variance for overall HVAC cooling power usage averaged 20.0%, and ranged from 2.3% to 79.2%. No advantage was found in using day-to-day variations in occupancy compared to using average hourly values when analyzing energy savings potential