Innovative design modification during construction of a twin tunnel using real-time field data

Tunnels are vital substructures that are strongly influenced by the surrounding soil, especially when using the New Austrian Tunneling Method (NATM) in which the soil surrounding the tunnel displaces under part of the applied load and the remaining load is borne by the tunnel initial lining. Displacement monitoring is of great importance in this method and designers should monitor data during construction in addition to a choosing an initial design based on primary soil parameters. The method should be a performance-based approach that is required to meet a target. This paper reports on a case study on an urban twin tunnel constructed using NATM. The construction method is compatible with existing geological conditions and space constraints caused by the high traffic congestion on Valiasr Street in Tehran. The twin tunnel is composed of asymmetric sections and has an overburden of about 12.5 m, height of 8 m and width of 18.3 m with three lanes. In other words, it is composed of a two-lane and a one-lane tunnel with connected initial linings and separate final linings. The numerical modeling was based on a 2D finite element program (PLAXIS version 8). Performance was evaluated by inducing settlement during and after excavation and initial and final stabilization. The model was analyzed statically. During construction, limiting settlement to the regulation values required strategies such as reducing the length of excavation, partial construction of the final lining on one side and reinforcement with micropiles and compensation grouting. The numerical results show that the length of excavation resulted in low efficiency of about 13%, but the micropile elements reduced the resultant surface settlement 30% and the induced total strain 46% over the critical value. In other words, the displacement trend line decreased with a dip of 30%. In comparison with field observations and field data, the numerically estimated settlement at the ground surface was approximately twice that of the measured value and, interestingly, the recorded convergence was fixed during tunneling, suggesting that unknown conditions like a cavity or deep soil with filling material exists above the tunnel. Comparison of the back-analyzed results with the field data showed a decrease in the real values of the geotechnical parameters for the soil layer of more than 20%. This confirms that the existence of local unknown conditions can cause a difference of up to 50% in surface settlement. It was found that the loose soil properties decreased the efficiency of soil improvement to less than 15%.