THE (MIS)BEHAVIOR OF BEHAVIOR ANALYSIS STORAGE ESTIMATES

This paper investigates the dependence of estimates of reservoir stora ge capacity derived using behavior analysis on the length of inflow se quence used for overyear reservoir simulation, It has generally been a ssumed that simulation using behavior analysis, which incorporates a g iven probability of failure, will always give steady state estimates o f the storage capacity (apart from the effects of the initial reservoi r condition). The results reported here show that it may take sequence lengths as much as 1000 years or more for the mean of the distributio n of storage capacity estimates to approach a stationary value. For so me cases with high draft and high inflow variability, we show that a s ignificant swing of the mean storage estimates from an initial downwar d bias into an upward bias occurs before their convergence to a statio nary level. However, the median storage estimates always showed downwa rd bias which sometimes decreased very slowly with increasing sequence length. We provide explanations for these observations and discuss so me of the implications on the choice of inflow sequence length in dete rmining reservoir storage capacities using behavior analysis.