Back analysis of strut loads for deep excavation of NEL Dhoby Ghaut Station in Singapore

The new Dhoby Ghaut Mass Rapid Transit (MRT) station for the North East Line is the largest rail transit interchange in Singapore after the five-level underground basement station was put in operation in June 2003. The excavation was some 180 m in length from south to north and its width ranged from 24 to 150 m. The depths varied from 17 to 31 m in soft soils overlying weathered sedimentary rocks. Secant pile walls and solider pile walls were adopted with five to seven layers of laced struts and rock bolts as the retaining wall system. The design strut loads were estimated using Wallap program and the Terzaghi and Peck apparent pressure diagram for stiff clay. The larger forces computed by the two methods were used. The measured maximum strut loads in most of the excavation areas at the NEL Dhoby Ghaut station gave similar results, showing lower pressures at the upper and lower part of the excavation compared to that at the midheight of the excavation. The apparent pressure coefficients were back calculated to be 0.1-0.3 at the southern part of the site where the excavation was wide and the sedimentary rock strata were overlain by soft soils. The results appear to be similar to the Terzaghi and Peck diagram for stiff fissured clay. However, the measured strut loads were very high in Section 7 at the northern part of the site where the excavation was narrow. The apparent Terzaghi and Peck coefficient was back calculated to be as large as 0.65. Compared to the upper bound coefficient of 0.4 suggested for stiff fissured clay by Terzaghi and Peck, this is still larger by 60%. As a result, the strutting system had to be strengthened because of the high measured strut loads. This high strut loads only occurred in the narrow, deep excavation in the sedimentary rock condition. The measured wall deflections were very small. The high strut loads measured at Section 7 might be explained by considering the excavation geometry and the initial stress in the ground. Compared with other sections of the station excavation, only Section 7 involved a very deep, narrow excavation fully in weathered sedimentary rock. The effects of the excavation width B, depth H and at the rest earth pressure coefficient K0 on the strut loads were investigated by using a commercial FEM program PLAXIS. The ratios B/H of width to depth of excavation were 3.3 and 0.67, representing the values at the southern and northern part of the site, respectively. The field at rest earth pressure coefficient K0 was estimated based on the equation K0 1/4 (1 - sin/0)OCR0.5. OCR is the over-consolidation ratio and was determined from relationships between normalised undrained strength and OCR suggested by Ladd et al. (1977). The FEM analysis shows that a narrow excavation in the sedimentary rock tends to result in higher strut loads than a wide excavation. The computed strut loads in the narrow excavation are 2.5 times those in the wide excavation. The strut loads also increase significantly when the K0 increases from 0.5 to 1.4. The computed strut loads and the wall deflections based on K0 of 1.4 are found to be very close to the measured strut loads and wall deflections. The wall deflections also increase with K0. The calculated wall deflections based on K0 of 1.4 are close to the measured deflections of the walls. (A). "Reprinted with permission from Elsevier". For the covering abstract see ITRD E124500.