Terror threat to civilization has been increasing all over the world. Thus it is important to detect explosive objects in gathering place, in order to assure safety and security. However sufficient explosive-detection system has not existed. Part of the reason for this fact is the limitation of detection accuracy using only one detector. From a practical point of view, we also detect explosives not to prevent the flow of people. In order to accomplish these two objectives, we propose integrated model for explosive detection. Using this model, we could achieve high detection accuracy without preventing the flow of crowd. We examined three detectors as a case study: millimeter-wave imaging detector, VUV-SPI-TOFMS detector, and pulsed neutron detector. We have established the integrated model using these detectors, and assured the improvement of detection accuracy.

Estimation of the heat radiation from tank fire is important for the safety and reliability of chemical and oil plants, however, it is very difficult to estimate the heat effects because of complex phenomena such as wind conditions, the location and the shape of tanks, etc. In this study, we have developed a simulation system to clarify the effects of tank fire based on Monte Carlo method. In the evaluation of the heat, there are various parameters such as the tank diameter, height, wind velocity, distance from the tank, and so on. We proposed the development of the database of the heat radiation by using three parameters of 'wind velocity', 'Height / Diameter of a tank', and 'Distance from a tank / Diameter'. Furthermore, the simulation system for heat radiation from tank fire under the wind condition has been developed based on Monte Carlo method, the database and GUI（ Graphic User Interface）. From the numerical examples, the dangerous area for a human body can be estimated conveniently, and the proposed system is effective for the defense activity plan These results indicates that the proposed system can be applied for the safety and reliability of plants.

Conventionally, such techniques as FMEA, FMECA and FTA have been widely applied to both analyses of reliability and safety of products from the viewpoint of design defect prevention. The present paper proposes a new technique, S-N-FTA, considering the sequences of the events in non-coherent systems. Next, the S-N-FTA is applied to the accident causation analysis for forced- draft-balanced flue stoves using oil as fuel. It is demonstrated that the common failure, namely a single failure that brings about two or more faulty causes resulting in a serious accident, can be analyzed efficiently by use of the S-N-FTA.

Chemical plant, gas, and various other industries use large-scale buried pipelines to transport inflammable gases and other products. To ensure the safe design of pipeline systems and their efficient operation and maintenance, it is crucially important to evaluate the long-term reliability of piping materials and to estimate how and to what extent gas will diffuse if it leaks in the ground. Under these circumstances, there is growing need for a numerical simulation tool that can efficiently analyze the diffusion behavior of gas leaks in ground with specific soil properties. Though many papers have discussed many cases of gas diffusion in both indoor and outdoor atmospheres, few papers have focused on the diffusion of gas leaks in the ground. This paper is the first one to report results verifying the applicability of a numerical simulation model we developed on the basis of a series of full-scale underground gas leak experiments. This simulation model is a practical analysis tool, and is expected to enhance the safety management of underground gas pipelines in related industries.