Catherine O’Leary, Benjamin Jones
Languages: English | Pages: 10 pp
Bibliographic info:
38th AIVC Conference "Ventilating healthy low-energy buildings", Nottingham, UK, 13-14 September 2017

Exposures to airborne fine particulate matter with a diameter of <2.5μm (PM2.5) are linked to multiple negative health effects, including cardiovascular and respiratory disease. Existing investigations of PM2.5 primarily focus on external sources and exposures, because outdoor air is easier to observe, and therefore, more widely monitored. However, as people spend up to 70% of their time in their own homes, exposures to indoor pollutants could have a greater impact on health. One method of investigating indoor exposures in a stock of houses is by modelling them. However, this process requires an understanding in the uncertainty of emission rates from internal sources.
Cooking has been identified as key source of PM2.5 in non-smoking households. Existing studies of emission rates use a range of techniques from small scale test chambers, which give control over all parameters in unrealistic conditions, to personal monitoring studies in real conditions where there is no control over influencing parameters. Reported emissions rates for single sources vary significantly indicating poor repeatability, and are generally presented without an indication of the uncertainty in them; for example, as a probability density function (PDF). Therefore, existing emission rates have limited use for stochastic indoor air quality modelling.
This paper seeks to develop a methodology to measure emissions rates of PM2.5s from the cooking of foods. A two-phase investigation measured the variation in emission rates when toasting bread in an electric toaster, a process that is simple and repeatable with fewer variables than many other cooking processes.
Phase one was conducted in a domestic kitchen. A TSI SidePak™ AM510 optical monitor measured temporal concentrations during and after toasting (n=40). A number of problems with the procedure were identified: (i) the recording time-step was too long, which lead to insufficient data points during the emission period; (ii) the ventilation rate and mixing conditions were unknown, although steps were taken to control them, which increased measurement uncertainty; (iii) the relative humidity was not monitored, which can affect the performance of the SidePak™ at high levels, and (iv) when the toaster and bread were not isolated at the end of the toasting period, increasing uncertainty in the total emission period.
The second phase used a test chamber, which offered greater control over the indoor conditions. An experimental procedure was followed which was similar to that use during the first phase, but with a one-second time-step, the toaster and toasted bread were sealed after the emission period, and the chamber was flushed with outside air between tests. Relative humidity and temperature were monitored during tests. Emission rates are estimated using an established model and reported as a histogram.