Tunnel boring machines (TBMs) have grown in popularity as a method of constructing lengthy tunnels with a medium to wide span. Its use has been documented for terrain conditions ranging from hard to soft strata, thanks to its cutting-edge features that allow for fast advance rates. However, severe geological conditions such as mixed-face and squeezing ground conditions, as well as discovering undetected aquifers, pose significant tunneling obstacles and may result in risks if proper precautions are not taken. This document compiles case studies on the challenges that different types of TBMs face while excavating in various challenging ground conditions, as well as the solutions that have been developed to address the issues. This brief review's goal is to simplify existing knowledge and strengthen our grasp of this tunneling approach for usage in a variety of ground situations. The study then goes over the TBM performance prediction models that have been developed. The TBM's progress rate and performance are predicted using performance prediction models, which are a significant component of project management. They are based on geological constraints, machine parameters, and site management. Tunnel engineers can avoid surprises during tunneling by improving the prediction model's performance and identifying the main parameter in machine performance. The difficulty to ascertain specific geological conditions prior to project start-up, on the other hand, appears to be a disincentive to these models. This study seeks to analyze the effectiveness of performance prediction models for severe ground conditions through this brief review.