Reliability Analysis of Debris Basins in Los Angeles, California

摘要

Reliability model which can quantitatively calculate the probability of failure of debris basin was developed using MSDPM (Multi Sequence Debris Prediction Method). In reliability model, load and resistance of reliability function for debris basin were estimated by MSDPM and size of debris basin, respectively. Annual maximum 1-hr rainfall intensity of Los Angeles was statistically analyzed by 7 rain gages near Los Angeles area. It was found that statistical distribution for annual maximum 1-hr rainfall intensity of Los Angeles is matched well with Gumbel distribution. After the analysis of rainfall intensity as a random variable, reliability model was applied to the calculations of the probability of failure of Brand and Childs debris basins located in Los Angeles. Reliability analysis was performed to calculate the probability that debris basin is filled up with debris according to maximum 1 -hr rainfall intensity of 50-year return period. Probability of failure of two debris basins according to number of maximum rainfall events was calculated using 50-year return period with and without wildfire. Even if maximum rainfall events were happened consecutively several times, probability of failure of Childs debris basin was less than 1% without wildfire. It was found that wildfire must be very important variable which can generate the huge debris yields. Therefore, wildfire should be carefully considered for the design criteria and maintenance of debris basin.