New Zealand Geotechnical Society Special Meeting: Geotechnical Earthquake Engineering in Canterbury

6.00pm – 8.15pm, Tuesday 28th September 2010
Lecture Theatre E1, College of Engineering, University of Canterbury

This Canterbury Technical Clearinghouse meeting summary was prepared by Bruce Deam on behalf of the NZ Society for Earthquake Engineering (NZSEE). This final version (13 Oct 2010) contains corrections from most of the contributors. Click on contributor's names to send them an email message.

The meeting was introduced and chaired by the NZGS Canterbury Branch Coordinator, Nick Harwood (Coffey Geotechnics). He outlined how many of the failures in the infrastructure were principally due to soil and foundation deformations, namely lateral spreading, subsidence and liquefaction.

Bruce Deam (University of Canterbury and NZSEE) introduced this meeting as a successor of the Canterbury Structural Group Forum the previous week. He indicated that he was making a record of the proceedings and that this would be available on the Canterbury Technical Clearinghouse website so all those involved with the reconstruction could be well informed about the issues each group is addressing.

Mark Stirling (GNS Science) presented an overview of the National Seismic Hazard Model, as embodied in the NZS 1170.5 Earthquake Loadings Standard, outlining the background to the model and how the Sep 4 earthquake from a previously unknown fault was catered for in the "background seismicity" for the region. Graeme McVerry then outlined significant features of the ground motion recorded by the Canterbuty Seismic Instruments devices developed by John Berrill. Mark and Graeme kindly agreed to their Powerpoint Presentation being posted for later review.

Misko Cubrinovski provided a geotechnical overview as that "The ground motion due to the Greendale fault rupture indicates cyclic mobility (excess porewater pressures in dense soils)". He summarised the event as producing:

No evidence of acceleration directionality within the city.
Similar overall shapes of the acceleration history throughout the city;
Only moderate to strong ground shaking, even in Kaiapoi. The accelerations were much smaller than Kobe though;
Liquefaction in the same most heavily hit areas of Kaiapoi as in 1901, consistent with the anticipated minor 3-4% soil improvement provided by an earthquake of that magnitude; and
A 400-500 mm thick layer of ejected sand in parts of Kaiapoi. Less in Avonside and pockets in areas such as Bishopdale.

He noted that mapping the extent of liquefaction is critically important. Research teams (including local and VisitingScientists) have been busy with the following investigations that will be published in a (70-100 page) report in early October:

Using Sweedish weight sounding tests, which are well correlated with SPT blow counts and only require a simple (and reasonably lightweight) apparatus to record to depths of 9 m. He indicated that water is still moving around and that the sites will have one or more repeat tests to study the evolution of the soil strength.
Mapping transects of the significant lateral spreading in Kaiapoi. Courtney drive had up to 700 m of land spreading (by 2 to 4 m, which is large movement), which was still moving several days later. They are aiming to identify spreading areas and quantify the deformations.
Residential house surveys recording the type and amount of ground movement and its effects on the structures.

Aaron Haymes (CCC, Building Consent team leader) indicated that there were many projects at various stages of construction at the time of the earthquake. They were evaluating these, in addition to the post-earthquake reconstruction. Those in high risk areas are being asked to get a geotechnical engineering report before proceeding. He concluded that it will take some time to provide guidance.

Another TA representative indicated that they were legally required to accept engineering opinions and current standards (e.g. NZS 3604), even though the foundation requirements were clearly inadequately specified.

Someone noted that house floor slabs needed better reinforcing (and that mesh reinforcing has been demonstrated as inadequate) and that stronger and better detailed perimeter ties were required.

Nick Harwood (Coffey Geotechnics) noted that the NZGS have published guidelines that indicate that construction must comply with codes but that unlike the AS/NZS 1170 suite, they don’t give serviceability limits.

Colin Ashby (Ashby Consulting Engineering, Warkworth) suggested that the unreinforced slabs currently permitted by NZS 3640 should not be used in Canterbury or elsewhere in New Zealand. He suggested that all slabs on grade should have, as a minimum, 665 mesh reinforcing. He also gave examples of ribraft foundations that had performed well on poor soils during other extreme events.

Mike Jacka (Tonkin and Taylor) outlined the work he had beeng doing for EQC over last few weeks:

Initially using regional mapping (for 1 week) to identify the worst areas around town to allow EQC teams to focus on those areas. Habitability was the primary issue.
The second phase was conducting detailed surveys of damage (from week 2).
The next phase is teams doing a comprehensive site-by-site assessments looking briefly at structure but mostly assessing land damage (i.e. focusing on claims on land rather than the house).

Mike reiterated an earlier need to define an expectation of damage at the Ultimate Limit State (ULS) for the soil. However, it needs to maintain parity with other hazards and balance the requirements against the risks for other hazards in other parts of country. Also, this is a sample of only one earthquake.

He also stressed that technical information is prone to misinterpretation and therefore needs to be publicised carefully. Some are forming the opinion that this is ULS damage and that the damage is not so bad that the areas need to be abandoned now (and probably repopulated in 15 years when perception of risk has diminished).

Mike concluded that EQC is bringing a community focus to the technical details. They need to convince other insurers that the land (and building) is reinsurable. Also, building on repaired land is no different to building on new land so how about introducing a 150 year habitability limit state?

Ian McCahon (Geotech Consulting) commented that he was surprised by the large areas where liquefaction hasn't occurred. Therefore he saw the issue of how to deal with liquefaction applies to a much greater area than just the areas affected in this earthquake. He cautioned that there needs to be a measured approach, rather than just reacting to this event, as improving liquefaction resilience will come at a cost.

He noted that the NZ Building Code has no reference to liquefaction - the closest is subsidence - so there is no legal grounding. There is nothing in NZS 3604 either, and AS/NZS 1170 explicitly excludes structures that are subject to liquefaction. He concluded that liquefaction has to be addressed on a site-by-site basis; there is no one fix for all situations.

The final discussion in the meeting centred on whether grout injection could be a viable method of releveling buildings or improving the soil. Trials of these and other methods are being discussed as a means of verifying the viability of the treatments.

Use of Information Disclaimer from the Canterbury Technical Clearinghouse