Probit functions for selected chemicals based on AEGL-3 values

Authors:

  • Ulrika Bergström
  • Sofia Jonasson
  • Åsa Gustafsson

Publish date: 2019-07-05

Report number: FOI-R--4720--SE

Pages: 23

Written in: English

Keywords:

  • Risk area assessments
  • probit function
  • toxic industrial chemicals
  • gases from smoke and fire
  • chemical warfare agents

Abstract

Leaks of volatile and highly toxic chemicals can form gas clouds, which may be transported long distances by the wind before the turbulence dilutes the gas cloud to harmless concentrations. Dispersion models obtained at FOI describe the distribution of chemicals in the environment after, for example, an accident or fire. To interpret how various emission scenarios / events described in these dispersion models may affect the population there is a need to estimate response rates (percental individuals with detrimental health effects), preferably at different levels of injury. In order to estimate the response rate, good quality data from chemical exposure is required at different exposure periods and with relevant endpoints. In this report, we have used AEGL (Acute Exposure Guideline Levels) guideline values because they are scientifically based, open access, and describe effects and response rates at multiple exposure durations up to eight hours of exposure. The AEGL values are based on an average population and include both healthy and sensitive individuals. The probit function describes the relationship between the exposure period, the chemical concentrations in air, and the response rate. Data from dispersion models concerning chemical concentration in air and exposure period can be used in chemical specific probit functions to estimate the response rate in the population. Here we have calculated regression coefficients for the probit function based on AEGL-3 values. We have included several common and / or toxic chemicals found in industry and in smoke from fires. It also includes chemicals that are classified as weapons of mass destruction/chemical warfare agents. To validate the regression coefficients, we compare the AEGL-3 values for the chemicals with our calculated values based on probit functions derived via AEGL-3 values. For most chemicals there is a good fit between our calculated values and the corresponding AEGL-3 values. We have calculated the chemical concentration in air with a defined exposure time and response rate using the probit functions derived by RIVM and via AEGL-3, respectively. In general, the calculated chemical concentrations derived via the probit functions from RIVM were higher than the chemical concentrations calculated via the probit functions based on AEGL-3, i.e. AEGL-3 estimates that life-threatening effects may occur at lower chemical concentration of exposure. This can be explained by the use of different scientific studies, differences in the interpretation of toxicity data, and the use of different assessment (safety) factors to extrapolate / translate results from animal to human, and if sensitive individuals are considered or not. We have not analysed the contribution of these factors. In conclusion, we suggest that the probit functions derived by RIVM are used to calculate response rates of the general population where there is a risk of lethal exposure after a chemical emission. Furthermore, probit functions based on AEGL-3 are recommended by us to be used for chemicals where probit functions have not been published by RIVM.