A case study of TBM performance prediction using a Chinese rock mass classification system – Hydropower Classification (HC) method

Abstract

Basic Quality (BQ) method is a basic national standard of rock mass classification suitable for different industries in geomechanics and geotechnical engineering in China. Referring to the relevant provisions of BQ method, Hydropower Classification (HC) method, a specialized engineering geological classification system widely used in China, was compiled for evaluation on overall stability of surrounding rock and guide of excavation and support design of underground engineering in water conservancy and hydropower industry. As the input parameters of BQ or HC method are quite different with those used in RMR or Q system, which indirectly limits the applicability of the foreign developed TBM performance prediction models for the China’s TBM tunnelling projects. In order to develop an empirical model for hard rock TBM performance prediction with more suitable applicability in China, 49 valid datasets were collected from a water conveyance tunnel mostly excavated in medium to hard igneous rocks, and the empirical relationships between TBM performance and each parameter in the database were studied. The results showed that the prediction accuracies of TBM performance based on HC or BQ are very limited as the effects of the input parameters of HC method on field penetration index (FPI) are different and their weights assigned are improper. TBM penetration rate (PR) reaches its maximum value in the HC range 40–60 and BQ range 350–450, respectively. Boreability of the rock mass in class III is higher than that in class II. Ridge regression, principal component regression and partial least-squares regression methods were employed to solve the multicollinearity between uniaxial compressive strength of intact saturated rock, intactness index of rock mass, angle between discontinuity plane and tunnel axis, and average overburden of tunnel section in the database. Comparisons between the measured FPI and predicted FPI showed good agreement. This highlights the powerful potential of multiple regression analysis model based on HC method in TBM performance prediction. However, it deserves to emphasize that the developed empirical relationships should be considered valid only for new projects with geological conditions similar to the studied tunnel in this study, and more field data from different projects need to be collected to develop a universal model in the future.