Evaluation of risk assessment approaches of occupational chemical exposures based on models in comparison with measurements

Research output: Contribution to journalArticle


Risk assessments of chemicals in work places are needed to protect workers’ health and safety. Several different strategies can be used for conducting risk assessments. The aim of this study was to investigate approaches to risk assessment of chemicals based on exposure assessment models relative to occupational exposure limits values (OELs) and derived no-effect levels (DNELs) and in comparison with measurements relative to OELs. A second aim was to evaluate the modelled recommended outcome and compare it with measurements of exposure. In this study, 29 situations were assessed with ECETOC TRA, Stoffenmanager® 5.1 and ART. Personal exposure measurements were also performed. The percentage of measured exposure exceeding the recommended output was calculated to investigate the level of conservatism. All the modelled exposures were compared with OELs and DNELs where possible, and the GM of the measured exposure was compared with OELs (risk quotas). For ECETOC TRA, 31% of measured exposure exceeded modelled exposure. For Stoffenmanager® it was 17% and for ART and ART B it was 3% and 0% respectively. Hence, according to our data, ECETOC TRA is the least conservative. An investigation of the risk quotas showed that ECETOC TRA had 4 false safe situations, meaning the risk was low when the model was used but was high when measurements were used. This may lead to underestimating risks. All models had an elevated proportion, ECETOC TRA and ART the highest, of false unsafe situations meaning the risk was low when measurements were used but high when models were used.


External organisations
  • Örebro University Hospital
Research areas and keywords

Subject classification (UKÄ) – MANDATORY

  • Environmental Health and Occupational Health
Original languageEnglish
Pages (from-to)412-420
Number of pages9
JournalSafety Science
Publication statusPublished - 2018 Nov 1
Publication categoryResearch