Harm often occurs through several paths whereby system elements can have multi-states which can change irreversibly or reversibly. In order to derive effective control measures to prevent harm, hazards should be identified and analyzed systematically. This paper proposes a dynamic hazard analysis technique taking into account of system elements with multistates that can change irreversibly or reversibly. In the technique, first, a state of the system is defined by a combination of states of the system elements, and paths through which a system state can transition to other system states are shown in a state transition path diagram. Next, by tracing the path connecting the initial state and the harm state in this state transition path diagram, the procedure for systematically identifying and analyzing various hazards with different state transition paths without omissions is shown, schematizing by use of the S-A process chart. Furthermore, the technique is applied to two cases to demonstrate its validity.

In order to highlight the advantages of personal chemical exposure sampling methods for assessing exposure risks in manufacturing settings, an easy-to-operate layered passive sampler based on activated carbon fiber （ACF） is developed in this study. Experiments are conducted to determine the effectiveness of ACF for absorbing toluene organic solvent vapor, which is widely used in Japanese manufacturing, as well as the retention ability of the sampling device. Calculated sampling rate （SR） was 1.2 ml/min （4.5 ng / （ppm· min））. Limit of quantitation （LOQ） of toluene in the desorbed solvent was 2.00 µg/sample, lower than the value when 1/10 of TLV was sampled for 4 hours. The recovery rates of the spike and recovery test are 94％ or more, and the recovery rates of 8th day on storage tests with a sampled toluene of 20 ppm was 99.8％ stored at 4 ℃. The collected toluene amount was constant under the condition of 15-30 ℃ of temperature, 33-70％ of relative humidity, 0.1-0.3 m/s of wind velocity in the experiment of indoor environmental condition. The above results suggest that sampling is possible within a certain range of environmental conditions examined in this experiment. Furthermore, since there is little risk of this layered sampler coming in contact with the factory equipment, this sampler is suitable for sampling in narrow spaces and for various workspace. For future, it is important to experiment with or organic solvents other than toluene. These results indicate the effectiveness of the proposed personal exposure sampling method.