Ventilation rates are inferred from carbon dioxide measurements. Occupants generate carbon dioxide, causing indoor carbon dioxide concentrations to exceed outdoor concentrations. The ventilation rate (rate of outside air flow into the building) can be estimated if the indoor carbon dioxide source strength and the concentrations of supply air and room air are known (ventilation is the only significant process for carbon dioxide removal). Indoor and outdoor CO₂ concentrations are measured and the indoor CO₂ source strength is estimated based on the number of occupants in a building and an estimate of their CO₂ production. However, this method is subject to several sources of error which are described in detail elsewhere (Persily 1997, Mudarri 1997, ASTM D 6245-98) and summarized below:

• Carbon dioxide concentrations have often not stabilized when the measurements are performed, and the use of non-steady-state values of carbon dioxide concentration in a steady-state mass balance equation usually leads to overestimation of the ventilation rate.
• Carbon dioxide concentrations are often measured using instruments, such as indicator tubes, with large potential errors.
• Concentrations of carbon dioxide in outdoor air vary with location and time, and significant error may result if assumed outdoor concentrations are used in calculations.
• The number, weight, activity and diet of the occupants affect the indoor carbon dioxide generation rate and each of these parameters can only be estimated.
• Indoor carbon dioxide concentrations may be spatially non-uniform and measurements at a few locations may not accurately represent the average concentration in the exhaust air.
• Use of the peak CO₂ instead of actual steady state values may produce erroneous ventilation rate estimates, off by a factor of 2 at low ventilation rates, and less at higher ventilation rates (Persily and Dols 1990).

Figure 1. Typical Carbon Dioxide values measured in US Office Buildings, Statistical distributions of average workday indoor minus outdoor CO₂ concentrations (dCO₂) and peak one-hour minus average outdoor workday CO₂ concentrations (dCO₂MAX) in 41 1994-1996 BASE Study office buildings. (LBNL-44385)

Figure 2. Adjusted analyses of trend for the relationship between workday average indoor minus outdoor CO₂ concentrations (dCO₂) and combined and individual mucous membrane and lower respiratory SBS symptoms in the 1994-1996 BASE Study office buildings with relative humidity > 20%. Odds ratios and 95% confidence intervals, sample size (N) and WML test statistical significance of the dose-response trend are shown. The models included covariates to control for age, gender, smoking status, carpet, thermal exposure, RH, and VOC exposure. (LBNL-44385)

Adjusted analyses of trend for the relationship between workday average indoor minus outdoor CO₂ concentrations (dCO₂) and combined and individual mucous membrane and lower respiratory SBS symptoms in the 1994-1996 BASE Study office buildings with relative humidity 20%. Odds ratios and 95% confidence intervals, sample size (N) and WML test statistical significance of the dose-response trend are shown. The models included covariates to control for age, gender, smoking status, carpet, thermal exposure, RH, and VOC exposure. (LBNL-44385)