When we assess accidental chemical releases from plant derived from accident or natural hazard, it is necessary to construct reasonable scenario as non-stationary emissions and to perform realistic simulations. The goal of the study is to construct reasonable release scenario, however, actual accident data have not been summarized for such purpose. In this study, we used the data from "Information on Spills and Leakage Incidents of Poisonous or Deleterious Substances" database for 1999-2018, and summarized the characteristics of the major incidents, major chemical substances and the nature of the assumed effects of the incidents. Hydrogen chloride, ammonia, and chlorine were the most frequently reported substances for leakage into the air. On the other hand, sodium hydroxide, sulfuric acid, and hydrogen chloride were the most frequently reported substances for effluent into water. The cumulative frequency distributions of emission amount were close to log normal. The averages （50 percentile） and worst-case （95 percentile） of the release amount of these substances were calculated. As for hydrogen chloride to the air, 50 and 95 percentiles were 240 and 11,000 kg, respectively. It is expected that the results of this analysis can be used to develop a realistic scenario for non-stationary releases of chemical substances

The explosive impact experiment facility of the Institute of Pulse Power Science, Kumamoto niversity, which was selected as a research target for manufacturing safety, has taken safety measures to prevent accidents in impact molding experiments using highly dangerous explosives. Although there have been no accidents or disasters to date, it has been confirmed that there are residual risks centered on human error that remain even after taking risk reduction measures. With the aim of further continuing accident-free operations, we will re-examine the current safety management efforts and residual risks, conduct a new focused and multi-faceted human error risk assessment for investigation and analysis, and implement effective and practical measures such as visualization. We will report the results of a survey study to minimize the residual risk such as countermeasure study methods.

Adsorption and permeation of hexane, toluene, dichloromethane, and 2-propanol through disposable latex, nitrile, urethane, and vinyl gloves were studied. Disposable gloves stuffed by foamed polyethylene packing material were (1) soaked and floated in liquid organic solvents or (2) exposed to organic solvent vapors. Latex glove let pass through hexane, toluene, dichloromethane liquids and vapors. Nitrile and urethane gloves blocked liquid hexane, but passed toluene and dichloromethane solvents and vapors. Among the examined, vinyl gloves blocked all the solvents and vapors. Use of the Disposable gloves in organic solvent works would be hazardous by transdermal absorption.