Predicting tunneling-induced ground movement

Cost-effective and permissible tunneling can occur only if ground movement prediction is refined to accommodate changes in both the urban environment and tunneling technology. As cities age, tunnels are being installed closer to existing structures and in increasingly complicated belowground conditions. The reality of stacked tunnels, abandoned facilities, and more extensive use of underground space raises the question of whether relationships derived for single open-shield tunnels in free-field conditions can adequately predict ground movement for modern tunneling techniques with more complicated site conditions. Traditional empirical methods to predict maximum surface settlements and the percentage of lost ground for paired tunnels of the new Austrian tunneling method (NATM) in noncohesive soils are evaluated. Predictive data are compared with field measurements for grouted and nongrouted sections. Results showed that the estimated maximum settlement values of an NATM tunnel were highly similar to those of an open shield tunnel for both the grouted and ungrouted sections, although in some cases the Gaussian shape significantly underestimated the depth of the settlement trough in the outer 30% to 40%. Grouting substantially altered the amount of settlement. The average percentage of volume of lost ground with grouting was 1.6%, whereas the value was 5.2 % where no grouting occurred. The empirical methods typically generated a fairly reasonable set of responses for an NATM tunnel.