Assessment of plastic zones surrounding the power station cavern using numerical, fuzzy and statistical models

Abstract

Plastic zones evaluation around the powerhouse caverns is a very crucial issue in designing and constructing these structures and accurate determination of their related optimum support systems. Due to inherent difficulties during the field measurement of plastic zones around the powerhouse caverns and shortcomings of the available methods in this field, applying new predictive models is an attractive and helpful topic. Accordingly, plastic zones around the powerhouse caverns have been investigated in this research using numerical analysis (NA), fuzzy inference system (FIS) and multivariate regression (MVR) model. Based on the numerical simulations, a new predictive equation has been developed to determine the plastic zone at middle point of sidewall and induced key point around a cavern. The basic parameters including rock geomechanical properties and geometrical characteristics of cavern structures have been considered as input variables in plastic zones modeling at middle points of roof, floor, left sidewall and right sidewall as well as at key point. For FIS and MVR models construction, sufficient datasets were introduced based on the numerical simulations. Performance of established models has been assessed applying testing dataset and utilizing powerful statistical indices. Accordingly, it is proved that the derived results from FIS and NA models are more precise than MVR model and they are more satisfactory in plastic zone estimation. Finally, parametric study results revealed that lateral stress coefficient, depth of overburden and rock mass rating are the most effectual parameters and tensile strength is the least influencing parameter on the plastic zone around a cavern.