MODELS DESCRIBING VARIATION OF SHORT-TERM EXPOSURE LEVELS OF TWO CHEMICALS

摘要

To examine inter-relationships between short-term exposure levels of two chemicals A and B, consecutive 7.5-min time-weighted average values collected over full shifts were analysed for six workers. In the cases of two chemicals mixed in a fixed ratio the correlation coefficients (r_(AB)s) between the two exposure concentrations were more than 0.95. In the case of two chemicals mixed in various ratios, _(AB) was 0.59. In the cases of two chemicals used in separate processes, r_(AB)s ranged from -0.32 to 0.27. The data were analysed using a previously proposed model describing the inter-relationships by classifying the causes of exposure level variation into a common factor for the two chemicals and a specific factor for each chemical. The model fitted four cases with positive r_(AB)s, but not three cases with negative r_(AB)s. Thus, the model was modified by introducing a timing factor for the two chemicals for application to the negative correlation cases. The original model was also developed further to make it possible to describe the autocorrelation in an exposure-time series. In some cases, this developed model could divide total autocorrelation between that related to the common factor and that related to the specific factor, although it could not explain a cyclical autocorrelation function which was observed in workplaces.