The authors' comparison of site investigation predictions and actual tunnelling conditions encountered at 50 m depth in partly weathered granite make interesting reading. In particular, it is encouraging to note the increasing use of classification systems as an aid to tunnel support design decisions. The authors state that ‘the support system generally used in the tunnel varied from that actually recommended in the site investigation report but lay within the range of alternatives suggested by Bieniawski’ (Bieniawski 1973). The contractor in fact elected to use steel ribs as primary (temporary) support ‘in distinct preference to the rather more sophisticated roof bolting with sprayed concrete’.
Application of Bieniawski (1973) had in fact shown that suitable primary support for the 4.6 m span tunnel should be found among the following alternatives:
(i) B l-1.5m + S(mr)30mm
(ii) S 100 mm (crown), 50 mm (sides) (with occasional mesh or bolts as required)
(iii) light steel sets 1.5 to 2.0 m spacing where S = shotcrete, (mr) = mesh reinforcement, B = rock bolts of given spacing.
The final 300 mm thick concrete lining cast to a finished diameter of 4.0 m was presumably dictated by other requirements, since the authors state that ‘the rock together with its primary support system (should) provide the principal structural strength of the tunnel whether empty and at atmospheric pressure or full of water under tidal pressure’. With the above requirement in mind it is reasonable to apply the NGI Q-system of classification (Barton et al. 1974, 1975)