Evaluation and prediction of tunnel boring machine performance in variable rock masses

Each Tunnel Boring Machine (TBM) excavation system is comprised of many subsystems, each with performance and reliability characteristics which may be dependent or independent of the rest of the subsystems. Each rock mass to be tunneled varies in intact rock and rock mass properties, and our knowledge of what lies ahead along a tunnel drive is necessarily imperfect. It is important to be able to describe both equipment and property variability and knowledge uncertainty in order to choose among different excavation systems for a given project, and to establish realistic expectations for excavation completion date and cost. This study summarizes the assembly and application of a TBM-drive data base to estimate TBM mining schedule and cost. The data base contains performance, site, machine and rock mass parameters on over 600 TBM case histories involving the use of full face, single disc cutter equipment. The data was used to develop predictive methodologies that allow for simulation of TBM performance in a geologic framework. The adoption of the data base and predictive methodologies improves the ability of the geologist and engineer to identify and quantify the major impacts of rock mass conditions on TBM performance, aids in the preparation of mitigation measures, and, most importantly, provides a mechanism for quantifying excavation risk within the context of the TBM project plan. The probabilistic predictions of time and cost to complete tunnel generated from these studies can better reflect the intrinsic uncertainties of tunneling work than the "unique number philosophies" currently employed in the civil engineering community. The use of planning methodologies based on probable outcomes allow for project decisions to be based on a realistic prediction that takes account of the high level of uncertainty that is inherent in the planning of most tunnel projects.