A fast response radiation detector (survey meter) based on a response prediction method was developed and implemented to determine contamination levels and to decontaminate areas contaminated by cesium-137 (137Cs) released during the nuclear reactor accident. Traditionally, 30 s is required to measure the count rate (1/min), which is a criterion for the level of contamination, when the time constant of the equipment is 10 s. A detector with faster response would eliminate the excessive time necessary to inspect for contamination. A previously proposed response prediction method was applied to overcome this problem; however, increase in errors is possible at short measurement times. The effect was determined through an experiment. Relative error (standard deviation/count rate) for a static response was approximately 0.5% in the vicinity of the maximum value. The predicted value suggests that predictions close to real time (requiring 0.1 s) are possible. However, the relative error for the almost maximum value was about 4.3%. The relative error for a three-point moving average of the predicted values (requiring 0.3 s) was about 3.8%, and the relative error for a five-point moving average of the predicted values (requiring 0.5 s) was about 3.1%. The experiment shows that while the response prediction method has larger relative error than the static response, the prediction will take less than 1 s, enabling significantly faster processing.
ASJC Scopus subject areas