Proceedings of the Ninth Pacific Conference on Earthquake Engineering Building an Earthquake-Resilient Society

14-16 April, 2011, Auckland, New Zealand

Paper Number 007

NZS 1170.5:2004 site subsoil classification of Wellington City

S. Semmens1, N.D. Perrin2, G. Dellow2 & R. Van Dissen2 1AECOM, Auckland, New Zealand. 2GNS Science, Lower Hutt, New Zealand.

ABSTRACT: Wellington has an appreciable seismic risk due to the proximity of its concentrated population and infrastructure to several major earthquake sources. Geotechnical data mainly from 1025 drill holes, along with shear-wave velocity (Vs) determinations specific to this project were used to construct a 3D engineering geological model for Wellington City centre. From this model, the following maps were derived, and are presented in this paper: surficial geology, depth to bedrock, low amplitude site period, and NZS 1170.5:2004 site subsoil classes. The results show that a significant ground shaking amplification hazard is posed to the city, with the waterfront, Te Aro and Thorndon areas having a poorer site subsoil class in terms of NZS 1170.5:2004 than previous studies had estimated.

1 INTRODUCTION

Wellington City lies within a geologically active landscape, and has a relatively high seismic hazard (e.g. Stirling et al. 2002), and a variety of ground conditions are present (e.g. Begg & Mazengarb 1996). The objective of the study presented in this paper was to better define ground response models for the Wellington central business district (CBD) using compiled geological, geotechnical and geophysical data in conjunction with 3-dimensional (3D) modelling and limited 1:5,000 scale mapping (Semmens 2010; Semmens et al. 2010a, 2010b). A major output of this research was derivation of maps of surficial geology, depth to bedrock, low amplitude natural period (referred to hereafter as site period) and NZS 1170.5:2004 site subsoil class. These maps are described and presented below.

1.1 Study Area

The study area encompasses the Wellington CBD, extending from the Thorndon overbridge in the north, through to Wellington Hospital in the south and from Kelburn in the west, through to Oriental Bay in the east (Fig. 1). The study area was chosen for two reasons: many of the major building and infrastructural elements vital to Wellington and the surrounding region are located within the CBD; and the area also contains the widest range of ground conditions likely to be found in the city.

Basement rocks (greywacke bedrock) within the study area comprise Permian to early Jurassic quartzo-feldspathic sandstone and mudstone sequences of the Rakaia terrane (e.g. Begg & Mazengarb 1996). Younger Pleistocene deposits lie unconformably on top of greywacke bedrock with the deepest recorded sequence of 137 m occurring at Te Papa Tongarewa (Museum of New Zealand). These deposits are frequently targeted as bearing stratum for pile foundations.

The Pleistocene deposits generally encompass weathered alluvium, colluvium and shallow marine deposits, typically consisting of dense silty sandy gravels with interbedded stiff silts and organic clays. Holocene sediments overlie Pleistocene deposits and generally consist of weathered alluvium and colluvium with minor beach, estuarine and swamp deposits. Waterfront reclamation in Wellington has added more than 155 hectares to the city (Kelly 2005), including large areas of the CBD. Two main types of fill were used, hydraulic fill (pumped sand and mud from the seabed) and locally sourced, end-tipped fill (boulders, domestic building and construction waste, spoil etc). However, compaction was coincidental rather than to a standard (engineered fill). Engineered fill was used in the construction of the Thorndon container terminal and in recent developments.

2

1.2 NZS1170.5:2004 site subsoil classification

The influence of local geologic (site) conditions on the intensity of ground shaking and earthquake damage is well documented. Structural design in New Zealand takes into consideration site conditions largely through the incorporation of the New Zealand loadings standard (NZS 1170.5:2004). The loading standard prescribes structural design actions on the basis of site subsoil class to accommodate likely increased earthquake loadings due to shaking modification. The standard sets out five site subsoil class categories, based on geological and geotechnical properties, which must be used in the calculation of horizontal and vertical loading. The five site class categories are outlined in Tables 1 & 2. Site subsoil classes D and E (soft or deep soil, and very soft soil respectively) require increased loadings to be considered, resulting in increased design and construction costs.

The preferred method for site classification in New Zealand uses the “site period” parameter. Site period (when not measured directly) is defined as (approximately) four times the shear wave travel time from the surface to bedrock. This approach addresses the effects of deeper softer soils which exhibit longer period site response characteristics. However, the ability to classify sites using shear-wave velocities is limited by a lack of data. The NZS 1170.5:2004 site subsoil class map derived as part of this study uses the preferred method (site period) to assign site subsoil class.

Sites underlain by multiple layers are evaluated by estimating and summing the

contribution to natural period for each layer whereby all material above bedrock is included in the evaluation. An unconfined compressive strength of 1 MPa delineates the rock-soil boundary. In general, the classification of a site will be dependent on the surficial sediments present even when piles extend below to stiffer stratum (Standards New Zealand 2004).

Figure 1 (left). Study area map showing locations of 1025 drill holes and ~30 microtremor sites used in this study.

3

Table 1: NZS1170.5:2004 Site Subsoil Classes.

Class Description Definition

A Strong Rock UCS > 50 MPa & Vs30 > 1500 m/s & not underlain by < 18 MPa or Vs 600 m/s materials.

B Rock 1 < UCS < 50 MPa & Vs30 > 360 m/s & not underlain by < 0.8 MPa or Vs 300 m/s materials, a surface layer no more than 3 m depth (HW-CW rock/soil).

C Shallow Soil not class A, B or E, low amplitude natural period ≤ 0.6s, or depths of soils not exceeding those in Table 2.

D Deep or Soft Soil not class A, B or E, low amplitude natural period > 0.6s, or depths of soils exceeding those in Table 2, or underlain by < 10 m soils with undrained shear strength < 12.5 KPa, or < 10 m soils SPT N < 6.

E Very Soft Soil > 10m soils with undrained shear strength < 12.5 KPa, or > 10m soils with SPT N < 6, or > 10m soils with Vs ≤ 150m/s, or > 10m combined depth of previous properties.

Table 2: Maximum depth limits for site subsoil class C.

Soil type and description Maximum depth of soil (m)

Cohesive Soil Representative undrained shear strengths (KPa) Very soft < 12.5 0 Soft 12.5-25 20 Firm 25-50 25 Stiff 50-100 40 Very stiff or hard 100-200 60

Cohesionless Soil Representative SPT N values Very loose < 6 0 Loose dry 6-10 40 Medium dense 10-30 45 Dense 30-50 55 Very dense > 50 60

Gravels >30 100

2 DATABASE

A database containing lithological and geotechnical information from 1025 bore logs, test pits and site observations was compiled (Semmens 2010). Data used to characterise the geological and geotechnical conditions in Wellington CBD came from files held by Tonkin & Taylor and GNS Science. In addition to the 1025 boreholes compiled for this study, the non-invasive Spatial Autocorrelation (SPAC) microtremor technique (Beetham et al. 2010; Fry et al. 2010) was used successfully at 12 sites to obtain site response information including site period, shear-wave velocity (Vs) and unit thickness (Perrin et al. 2010). The database information has been entered into GNS Science’s web-accessible PETLAB database (http://pet.gns.cri.nz).

3 3D ENGINEERING GEOLOGICAL MODEL

A simplified 3D engineering geological model was created for Wellington CBD using Earth Research software (ARANZ). Each individual lithological unit from each borelog was assigned one of seventeen lithotechnical unit codes based upon their lithological description, grain size, SPT N count, weathering condition and depositional setting as recorded in the geotechnical borehole database. The interbedded relationship between many of the seventeen units was complex and required simplification for modelling. The simplification process involved grouping “like” units together, resulting in four different engineering geological units which could then be modelled. The four

4

engineering geological units are: Hydraulic Fill, Soft/Loose Deposits, Stiff/Dense Deposits and Greywacke Bedrock (Table 3).

Table 3: Model lithologies.

Engineering Geological Model Unit

Description Approx. Vs Range

Hydraulic Fill Dredged mud, sand and silt from the seabed. 100-150m/s Soft/Loose Deposits Artificial fill, landfill, reclamation (excluding hydraulic

fill), unconsolidated, granular sediments (typically Holocene age) including alluvium and marginal marine (e.g. beach) deposits and swamp deposits.

105-385m/s

Stiff/Dense Deposits Pleistocene dense to very dense sand/gravel alluvial and colluvial deposits, commonly weathered in-situ to form complex mixtures and discontinuous lenses of silt/sand/gravel with minor clay, and thin (< 1 m), stiff paleosol layers

220-730m/s

Greywacke Sandstone/ Mudstone

Grey (becoming yellow brown with increased weathering) interbedded quartzo-feldspathic sandstone and mudstone sequences

365-1760m/s

4 Vs CHARACTERISATION OF ENGINEERING GEOLOGICAL UNITS

Vs characterisation of the four engineering geological units (Table 3) was required to derive site period and site subsoil class maps. Vs data from Boon et al. (2010), Fry et al. (2010), Louie (unpublished), Ingham (1971) and various commercial studies have been used to characterise each engineering geological unit, the first time this has been attempted for Wellington City.

Vs is dependent on the stiffness and density of a soil. Generally, as soil depth increases so too does stiffness, thus Vs also increases. This relationship has been highlighted by the comparison of borehole data with measured Vs (e.g. from Seismic Cone Penetrometer testing) in Wellington CBD. Based on this principle the engineering geological units from the engineering geological model were assigned Vs consistent with real measured values for each unit. This was done by assigning each layer velocity from all existing Vs profiles in Wellington CBD to one of the four engineering geological units. This assignment of Vs to the four engineering geological units allows site period and Vs30 to be estimated anywhere within the study area. The loosest and softest soils in Wellington CBD have Vs as low as 50 m/s while very dense/hard soil velocities are ≥ 700 m/s (Semmens et al. 2010a, 2010b).

5 SURFICIAL DEPOSITS MAP

The Wellington CBD surficial deposit map (Fig. 2A) was constructed to show different Quaternary sediment zones and greywacke bedrock at a scale (1:5000) useful for initial site investigation. The surficial deposit map was constructed from drill hole investigations, test pits, site observations, aerial photographs, limited field mapping, and utilises information from existing maps in conjunction with specialist opinion (see Semmens 2010; Semmens et al. 2010a, 2010b for more detail). The map can be used to assist in site investigations by illustrating the most likely sediments found at or just below the surface of a site.

6 DEPTH TO BEDROCK

The depth to bedrock contour map (Fig. 2B) was constructed by extrapolation between the few holes that reached bedrock in the deepest parts of the Te Aro and Thorndon basins, mapped bedrock at the surface, and drill holes near the basin edges. In the absence of any deep holes to bedrock in the Thorndon area, the contours have been guided by estimates from microtremor methods. While this map can be used to estimate the likely depth to greywacke basement at a location, there are large

Harbour

5

uncertainties, particularly where the borehole density is low.

Figure 2: A) Wellington CBD surficial deposits. Map shows the nature of sediments at or just below the ground surface (< 5 m) in Wellington CBD. B) Depth to bedrock. Map shows “best fit” contours at 20 m intervals on the modelled greywacke bedrock surface with the projection of surface topography used where data is insufficient.

7 LOW AMPLITUDE NATURAL PERIOD (SITE PERIOD) MAP

Measured and pseudo site periods were used to construct the site period map (Fig. 3A). Due to the limited number of measured site periods in Wellington City, additional pseudo points were added by constructing Vs profiles to Greywacke basement. A site period was then calculated using the methods prescribed by NZS 1170.5:2004. These pseudo points were then used to refine and constrain the site period contour lines.

(B) (A)

6

Site period contours in Te Aro range from 0.2s (generally bedrock is < 20 m from the surface) to > 1.4s (along the waterfront at the Museum of New Zealand). In the Queens Wharf area site period contours reach a maximum value of 0.6s, but generally ≤ 0.4s. In Thorndon the site period contours range from 0.4s to > 1.8s, but are generally > 0.6s. The 0.6s site period contour is significant because it is used to differentiate between NZS 1170.5:2004 site subsoil class C and D sites.

8 NZS 1170.5:2004 SITE SUBSOIL CLASS MAP

A site subsoil class was assigned (NZS 1170.5:2004) to each of the 1025 boreholes from the geotechnical drill hole database (Fig. 3B). Four site subsoil classes (B, C, D & E) were then assigned to areas on the map using the drill hole site class designation (mentioned previously). In the absence of direct measurements of Vs in a particular area, the methods in Borcherdt (1994) and NZS 1170.5:2004 were used to estimate Vs, guided by the ranges of measured Vs established for the same geological units nearby.

The NZS 1170.5:2004 site subsoil class map (Fig. 3B) shows the distribution of the four different site subsoil classes present in Wellington CBD. Site subsoil class B occurs in areas where greywacke bedrock is at or just below the ground surface. Site class C areas occur around the basin edge in Te Aro, Wellington South, the Aro Valley and western Thorndon areas. Site class D occurs in a significant portion of Thorndon and Te Aro areas. Site class E has been conservatively assigned to areas underlain by hydraulic fill in the vicinity of Aotea Quay. It is recommended that site specific investigations are used to prove/disprove the assigned site class especially where a site subsoil class cannot be assigned with confidence because subsoil conditions may be close to the boundaries of an NZS 1170.5:2004 site subsoil class. If a ground class still cannot be applied with confidence, direct site-specific Vs determinations are required.

9 DISCUSSION & CONCLUSIONS

This study was the first in Wellington City to use 3D modelling and high quality mapping in conjunction with shear-wave velocity determinations to characterise subsurface conditions in the central city. The methodology used has application in other locations, provided the distribution, geotechnical properties and Vs of the subsurface materials can be characterised. A better understanding of the controls on site effects in the Wellington CBD has resulted from improved access to the available data. Improvements in the treatment of the data have enabled local site response parameters to be established more accurately and with greater consistency than in previous work. The new knowledge and understanding gained from this project will enable more accurate hazard and risk estimates for Wellington City. This will help increase community resilience and preparedness within the central city and provided engineering measures are properly implemented would enable faster recovery after an earthquake event.

Previous research (Grant-Taylor et al. 1974; Perrin & Campbell 1992; Van Dissen et al. 1993) has concluded that Wellington City is likely to experience varying levels of ground shaking amplification during a large damaging earthquake, with the reclaimed areas surrounding Wellington’s waterfront, Thorndon and Te Aro valley the most susceptible to ground shaking amplification and related phenomena. A limitation of previous seismic zonation studies of the Wellington CBD is that results were not presented in a form directly usable in engineering prescribed design (i.e. no prescribed design requirements). A major objective of this project was to present a seismic zonation for Wellington CBD based on site subsoil class currently used in the New Zealand loading standard (NZS 1170.5:2004). Thus, the first fit for purpose NZS 1170.5:2004 site subsoil class map for Wellington CBD was created.

Further work is needed, particularly in the Thorndon area, to better constrain the depth to bedrock, site period, Vs30 and site subsoil class maps, where the current site subsoil class could be revised if more detailed data becomes available.

7

Figure 3: A) Site period map shows contour lines representing site period at 0.2 second intervals. B) NZS 1170.5:2004 site subsoil class. Map shows the distribution of the four different site subsoil classes which are present in Wellington CBD.

10 REFERENCES

Beetham, R.D., Stephenson, W.R., Barker, P.R. & Perrin, N.D. 2010. Site classification from microtremor records, HVSR/SPAC: an effective, non-invasive site investigation method. p. 1695-1702 (paper 200) In: Williams, A.L., Pinches, G.M., Chin, C.Y., McMorran, T.J. & Massey, C.I. (eds.), Geologically active: delegate papers 11th Congress of the International Association for Engineering Geology and the Environment, 5-10 September 2010, Auckland, New Zealand: CRC Press.

Begg, J.G. & Mazengarb, C. 1996. Geology of the Wellington area, scale 1:50 000. Institute of Geological & Nuclear Sciences geological map 22, Lower Hutt, New Zealand: Institute of Geological & Nuclear Sciences

(A) (B)

8

Limited. 1 Sheet+128 p. Boon, D., Perrin, N.D., Dellow, G. & Lukovic, B. 2010. It’s Our Fault – Geological and Geotechnical

Characterisation and Site Class Revision of the Lower Hutt Valley. GNS Science Consultancy Report 2010/163. 68 p.

Borcherdt, R.D. 1994. Estimates of site-dependent response spectra for design (methodology and justification). Earthquake Spectra Vol 10(4) 617-653.

Fry, B. Stephenson, W.R. & Benites, R. 2010. It’s Our Fault - Seismic instrumentation and inversion for velocity structure of the Wellington region. GNS Science Consultancy Report 2010/18.

Grant-Taylor, T.L., Adams, R.D., Harherton, T., Milne, J.D.G., Northby, R.D. & Stephenson, W.R. 1974. Microzoning for earthquake effects in Wellington, NZ; New Zealand Department of Scientific and Industrial Research Bulletin 213. Ruscoe, Q.W. (ed.) Wellington, New Zealand Department of Scientific and Industrial Research. 7 Sheets+62 p.

Ingham, C.E. 1971. Elastic properties of rock, Wellington urban motorway tunnel. Department of Scientific and Industrial Research, Wellington, Geophysics Division report 71.

Kelly, M. 2005. Heritage trail: old shoreline, Wellington City, Wellington City Council. 69 p. Louie, J.N. 2001. Faster, better: Shear-wave velocity to 100 meters depth from refraction microtremor arrays.

Bulletin of the Seismological Society of America Vol 91(2): 347-364.

McVerry, G.H. 2011. NZS1170.5:2004 site-effect terms as a continuous functions of site period and Vs30. Proceedings of the Ninth Pacific Conference. Building an Earthquake-Resilient Society, paper number 010, Auckland, New Zealand.

Perrin, N.D & Campbell, H.J. 1992. Compilation of geological data, Wellington area. DSIR Geology & Geophysics Contract Report No. 1992/24.

Perrin, N.D. Stephenson, W.R. & Semmens, S. 2010. Site class determinations (NZS 1170.5) in Wellington using borehole data and microtremor techniques. 8p (paper No. 22) in: Earthquake prone buildings: how ready are we? proceedings New Zealand Society for Earthquake Engineering Technical Conference, Wellington, New Zealand, 26-28 March, 2010.

Semmens, S. 2010. An engineering geological investigation of the seismic subsoil classes in the central Wellington area. Unpublished thesis, University of Canterbury, Christchurch, New Zealand.

Semmens, S., Perrin, N.D. & Barker, P.R. 2010a. What lies beneath: geological and geotechnical characterisation of the Wellington central commercial area. p. 659-666 (paper 078) In: Williams, A.L., Pinches, G.M., Chin, C.Y., McMorran, T.J. & Massey, C.I. (eds.) Geologically active: delegate papers 11th Congress of the International Association for Engineering Geology and the Environment, 5-10 September 2010, Auckland, New Zealand: CRC Press.

Semmens, S., Perrin, N.D. & Dellow, G. 2010b. It’s Our Fault – geological and geotechnical characterisation of Wellington City. GNS Science Consultancy Report 2010/176. 48p.

Standards New Zealand 2004. NZS 1170.5:2004 Structural Design Actions - Earthquake Actions. Section 3 - Site Hazard Spectra, Standards New Zealand. Pp. 81.

Stirling, M.W., McVerry, G.H. & Berryman, K.R. 2002. A new seismic hazard model for New Zealand. Bulletin of the Seismological Society of America Vol 92(5): 1878-1903.

Van Dissen, R.J., Taber, J.J., Stephenson, W.R., Sritharan, S., Perrin, N.D., McVerry, G.H., Campbell, H.J., & Barker, P.R. 1993. Ground shaking hazard zonation for Wellington city and suburbs, New Zealand. p. 134-141 In: Proceedings, New Zealand Society for Earthquake Engineering technical conference, Taupo, New Zealand.

Van Dissen, R.J., & 37 others. 2010. It's our fault: better defining earthquake risk in Wellington. p. 737-746 (paper 088) In: Williams, A.L.; Pinches, G.M.; Chin, C.Y.; McMorran, T.J.; Massey, C.I. (eds.) Geologically active: delegate papers 11th Congress of the International Association for Engineering Geology and the Environment, 5-10 September 2010, Auckland, New Zealand: CRC Press.

Acknowledgements: The authors would like to thank Mark Rattenbury (GNS Science) for his expertise in model and map design and preparation. Tonkin & Taylor Ltd are thanked for support and allowing access to their archived information. The research presented here has been conducted as part of the “It’s Our Fault” project (e.g. Van Dissen et al. 2010).