Repairing SLS anomalies in NZ seismic code to reduce earthquake losses

The 1992 advent of the Serviceability Limit State (SLS) was for the purpose of eliminating structural and non-structural damage to buildings subjected to small or moderate Earthquakes (EQs). This goal complimented the prior 1976 goal of minimising life-loss due to large Ultimate Limit State (ULS) EQs. However, moderate direct damage and large indirect losses occurred to many medium-rise pre-2004’ precast concrete-framed buildings in Christchurch and Wellington CBDs as a result of small or moderate EQ ground motions in 2010 [1-3], 2013 and 2016 [4-6.] A precedence for a proposed SLS level 1 upgrade was set when Christchurch upgraded to a 50 year recurrence SLS following the 2010-2011 earthquakes [7]. Many modern buildings have been engineered with little regard for SLS [8] nor the goal of eliminating disruption from moderate EQs [9, 10]. The proliferation of SLS building damage and large indirect losses [1] have arisen in NZ primarily because of the specification of a too-small SLS demand which corresponds to a ground motion with 25 year return period and because the Structural Performance factor (Sp) is specified in NZ as 0.7 for SLS, which results in a further 30% reduction of the SLS demand. There are also vulnerabilities in ‘pre-2004’ precast floor-to-beam connection detailing [3]. Cost-benefit analyses show that these building losses may be relieved by first correcting the precast vulnerabilities, then using a SLS limit of 50 year (rather than the current 25 year) return period and/or by specifying Sp = 1. The thus proposed ‘maxi-50 year SLS’ with a drift limit of 0.25%, has the same elastic seismic demand as the 100 year international SLS event [10, 11] (with Sp = 0.7) and will minimise non-structural and business disruption losses in small to moderate earthquakes.