Rock Mech Rock Eng DOI 10.1007/s00603-016-0936-x
ORIGINAL PAPER
Block Slides on Extremely Weak Tectonic Clay Seams in Openly Folded Tertiary Mud-Rocks at Auckland and the Rangitikei Valley, North Island, New Zealand Warwick M. Prebble1 • Ann L. Williams1
Received: 14 March 2015 / Accepted: 21 February 2016 Springer-Verlag Wien 2016
Abstract Block slides have developed on extremely weak, thin clay seams of tectonic origin, parallel to bedding in gently dipping sandstones and mudstones of Tertiary age. Two areas of noted instability are investigated at Auckland and the Rangitikei valley. Dimensions range from 100 m across 9 100 m long for short displacement block slides up to 4 km across 9 3 km long for large landslide complexes in which block slides are a major component. Displacements of blocks range from incipient (cm) through short (30 m) to 2 or 3 km for large slides. Many of the Auckland slides are dormant but likely to move in a 2000 year return period earthquake or 100 year high intensity rain storm. At Rangitikei there are many active, younger slides. Sliding rates for active failures vary from a few cm/year to 50 m in 30 min. Host rocks are weak to very weak clayey sandstones and sandy mudstones. The seams are rich in smectite. They have polished and crushed walls, may have slickensides and some contain rounded rock fragments. Laboratory shear strength of the seams is 13 kPa cohesion and 13 friction, with a lower bound of 8 at zero cohesion. Strength is increased at the field scale by waviness, steps and splays. Continuity can be demonstrated over distances of hundreds of metres. Key investigation methods were mapping, shafts and trenches. Tectonic uplift, folding and faulting of the weak Tertiary strata and river down-cutting are perpetuating block slide development. Keywords Field investigations Geomorphology Shafts Clay seams Block slide Geological models & Ann L. Williams
[email protected] 1
Beca Ltd, PO Box 6345, Wellesley Street, Auckland 1141, New Zealand
1 Introduction Weak mud-rocks of Tertiary age form an extensive carapace over much of the North Island of New Zealand but are restricted to coastal belts and a few relatively small inland basins in the South Island (Prebble 1995, 2001), Fig. 1. These rocks occupy approximately one-third of the area of the North Island and are noted for their instability (Brown 1974; Prebble 1995). They are the younger strata in a sedimentary sequence which spans the whole of the Tertiary and in some basins also the upper Cretaceous. In other areas the weak mudrocks rest unconformably on older rock masses or are tectonically displaced over a complex undermass. The weak mudrocks are less deformed than older rocks and are openly folded, with low dips over considerable distances. They are porous, weak to very weak (1–7 MPa), wet and clay-rich. Sandstones and mudstones dominate. Mineralogy is dominated by quartz, feldspar, clay-rich weathered rock fragments and a clayey matrix. Smectite is the most abundant clay mineral, with subordinate chlorite, illite and kaolinite. The clays are mainly detrital, in the matrix and in weathered clasts. Landslides in general and block slides in particular have been noted in these rocks (Brown 1974; Hawley and Riddolls 1975; Stout 1977; Prebble 1995; Williams and Prebble 2010). Large areas of complex slides and flows can be seen at Auckland, Taranaki to Rangitikei, the East Coast Deformed Belt (Hawkes Bay region), coastal Marlborough, Nelson and Dunedin (Fig. 1). In the last 100 years, seismic triggering of block slides in Tertiary weak rock has been observed within 10 km of the epicentral zone for earthquakes of magnitude 7 and greater (Lensen and Suggate 1968; Pettinga 1987). High intensity cyclonic rain storms and exceptionally wet winters have also triggered new
123
W. M. Prebble, A. L. Williams Fig. 1 Major tectonic features, lithologic groupings and geotechnical terrains of New Zealand; soft weak rock highlighted
slides in Tertiary weak rock and renewed movement on existing ones (Ker 1970; Stout 1977; Thompson 1981; Coombs and Norris 1981; Pettinga 1987). This paper focuses on examples of block slides in two landslide areas where there is a noticeable abundance of failures. These are in the very weak sandstones of the southern landslide zone of Auckland and in the very weak sandstone of the Rangitikei River gorge in the southern central North Island. Recognition and assessment of deepseated block slides in these areas was critical to the location and design of engineering infrastructure. In the southern landslide zone of the Auckland region, construction of a sanitary landfill was proposed within a range of hills which extend from the coast near Howick
123
south to Alfriston, southeast of Auckland city, New Zealand (Figs. 1, 2). The landslide zone is also crossed by arterial roads, water pipelines and electricity transmission lines. Residential subdivision is advancing in the zone. The main state highway and the main trunk railway connecting Auckland and Wellington pass through the Rangitikei area, where they cross a number of landslides. One of these reactivated in the 1960s, moving at several cm/yr and required considerable on-going ballast build-up of the railway embankment and realignment of the state highway (Stout 1977). In the Rangitikei valley the railway follows a direct route across the terraces 70–80 m above the deeply incised river, which flows in a very steep-sided gorge. This required the building of three large viaducts,
Block Slides on Extremely Weak Clay Tectonic Seams in Openly Folded Tertiary Mud-Rocks at…
Fig. 2 The landfill site was located in the south-eastern part of Auckland between Howick and Alfriston
the northernmost of which, North Rangitikei Railway Bridge, is the second example presented in this paper. In particular, discussion focuses on the North bank abutment of this viaduct where there is evidence of block sliding.
2 Tectonic Setting and Formation of Clay Seams The clay seams are found in tabular sandstone and mudstone rock masses of Late Tertiary age, which formed in marine basins on the active convergent plate boundary through New Zealand. The seams in Auckland are in wellbedded Miocene flysch, consisting of turbidite sandstones and pelagic mudstones deposited in a subsiding inter-arc basin and derived from volcanic arcs and undermass highs of argillaceous greywacke and igneous rocks (Kermode 1986; Edbrooke 2001). In the early Miocene, an active convergent margin propagated through the Auckland region. Regional uplift, tilting and open folding followed. The now exposed Miocene flysch is dissected by active coastal and stream erosion with block slides common in some of the openly folded, gently dipping structural domains.
The seams at Rangitikei are in Pliocene massive marine terrigenous sandstones and mudstones, which accumulated in shelf environments of the Whanganui Basin (Journeaux et al. 1996). Much of these are quartz and feldspar rich but mafic and lithic clasts are dominant over some intervals. The Rangitikei region shows a simple post-depositional structure. Up-section decrease in the dip is attributed to southwards tilting and syn-sedimentary subsidence of the basin (Journeax et al.). A major NE–SW striking high angle reverse fault and an active anticline (Milne 1975; Thompson 1981) indicate on-going compression, uplift and folding. Both areas accumulated poorly sorted sediment with a clay-rich matrix and a variable component of weathered clayey lithic grains. Hence even at depths of 40 m or more the un-weathered rock is clay-rich. The mudstones are stronger and more massive. The sandstones are weaker, clayey and silty, with thin mudstone beds. In south Auckland, as part of a regional mapping study, Kermode (1992) recognised a concentration of slope failures in weathered Miocene sandstone and mudstone and mapped them as unstable ground. He did not specify the type of failure or geological controls on the slip surface. At Rangitikei, Ker (1970) identified the boundary between mudstone and the overlying sandstone as a zone of potentially high pore-water pressures and sliding failure parallel to bedding but he did not recognise any clay seams. Later, clay seams and clay-coated fractures, with splays, wall crushing, polished surfaces, slickensides and sheared fabric were found parallel to bedding in these rock masses at both Rangitikei and Auckland (Stout 1977; Prebble 1979; Thompson 1981; Williams and Prebble 2004). They have since been attributed to a tectonic origin (Williams and Prebble 2010). This is considered to be flexural slip parallel to bedding. There is further reworking during slope failure. The seams are 1–20 mm thick and the coatings \1 mm. The open folding and simple structural domains produce tabular rock masses with low dips (1–10) of bedding. Now uplifted and exposed by the compressive tectonic regime, these rock masses are particularly susceptible to block sliding. At both Auckland and Rangitikei, the clay seams are most common in the sandstone units and are parallel to bedding, usually at the boundary with thin mudstone beds or close to them. Seams exist at all depths explored from the ground surface to at least 40 m deep and throughout all weathering grades, including un-weathered rock. Even where they are encountered in shafts at depths where no slope failure is possible the seams are polished and/or striated. Clay mineralogy of the seams is the same as the clay matrix of the enclosing rock.
123
W. M. Prebble, A. L. Williams
Therefore the seams are considered to have formed by flexural slip of the beds during open folding as a result of regional compression and uplift.
3 Successful Methods of Field Investigation of Clay Seams Investigation of the seams and assessment of their significance as the basal ruptures to block slides commenced with engineering geological mapping and a geomorphological assessment (Fig. 3). The size and the variety of landforms in the landslide areas necessitated a regional approach in order to set the Auckland landfill site and the North Rangitikei Bridge into meaningful local contexts. Scale and complexity of the block slide areas in Tertiary weak rock were not recognised until the regional studies by Stout (1977), Thompson (1981), Prebble (1979), (1987), (1995), Pettinga (1987) and Tilsley (1993). These demonstrated that the sliding did not change the stratigraphy but severely disrupted and dislocated the geomorphology. Stout found that at Rangitikei, 3 km north of the bridge site, a large,
mainly translational slide had taken place on thin montmorillonite clay seams, and that movement was almost parallel to the strike. In Auckland block slides on bedding-parallel clay seams in Miocene weak rocks were recognised in the ‘‘Southern Landslide Zone’’ (Prebble 1995, 2001), which includes the landfill site (Fig. 4). Comparison with Rangitikei (Stout 1977; Thompson 1981) and with similar Tertiary rock masses elsewhere in New Zealand (Coombs and Norris 1981; Pettinga 1987) established the view that this model was probably widespread (Fig. 1). Initial investigations comprised regional photogeology and mapping of surface exposures and landforms for several kilometres around the project site areas. Conventional cored boreholes investigated Quaternary alluvial deposits, slope debris, the soil/rock profile, rock and soil properties, rock defects and were used to assess slope movement types. At each site a preliminary model of block sliding on clay seams was established. Using this tentative model, more detailed investigations followed at each site: geological and geomorphological mapping, trenches and pits and inspection shafts.
Fig. 3 Geomorphological map of the landfill site (north is to the left of the page)
123
Block Slides on Extremely Weak Clay Tectonic Seams in Openly Folded Tertiary Mud-Rocks at…
Fig. 4 Scarps and benches of block slides in the south-eastern landslide zone looking across the Central Stream to the south-western ridge
3.1 Geological Mapping Mapping on a regional scale was done to establish the significance of rock types, structures, defects and geomorphic features at the sites, in the context of those in the region. Coastal cliffs at Auckland, river banks at Rangitikei and road cuts provided exposures of considerably greater extent than any at the sites. Detailed mapping of both natural and man-made exposures provided further information on defects, in particular clay seams (Fig. 5). 3.2 Inspection Trenches and Pits Excavation of additional deep inspection trenches and pits at key locations were used to locate the clay seams and to provide detailed information on the orientation of bedding and defects (e.g. Fig. 6). Selected investigation sites were those likely to expose an extension of the clay seams found in the investigation shafts. At Auckland a series of trenches excavated across topographic lineaments were examined for Quaternary deposits, soil horizons including paleosols, datable materials, fault offsets and crush zones, shear zones or closely fractured zones attributable to faulting. At Rangitikei, following the confirmation of vertical fractured zones in the pile shafts, four trenches were excavated through the alluvial terrace deposits across the full width of the North abutment promontory of the bridge site to establish the presence of any additional fractured zones and the likely extent of all such zones, including one located in the pile shaft. 3.3 Drilling of Inspection Shafts Twelve 0.9 m diameter inspection shafts were drilled to depths in excess of 40 m at the Auckland site and provided
Fig. 5 Clay seam (a) found in the Redoubt Road tunnel excavations of the south-eastern landslide zone described by Wylie (1989); b encountered in excavations in Grafton Gully, central Auckland
continuous exposure of the surficial deposits and bedrock as well as allowing direct, in situ inspection and sampling of rock and defects in particular bedding parallel clay seams (Fig. 7). Boreholes had proven unreliable for recovery of defects. Shafts were drilled sequentially with information obtained from drilled shafts being used to determine the preferred location and depth of subsequent shafts. Logging and sampling was done from a specially designed cage. High water inflows into shafts 54 and 55 prevented detailed inspection of the lower portion of these shafts. The measured strike of planar features was estimated to be accurate to ±3. Collection of undisturbed samples of the clay seam for shear strength testing was achieved by placing a metal canister (approximately 120–150 mm square) end on over the seam and adjacent rock and hammering the tin into the face of the shaft. Rock surrounding the tin was progressively chipped away to release the tin and intact clay seam sample. At Rangitikei, two large diameter pile shafts confirmed rock conditions indicated by boreholes. Boreholes were not completely reliable for defect recovery. Successive
123
W. M. Prebble, A. L. Williams
Fig. 6 Elevation sketch (part) of the western face of trench excavation 7
attempts with progressively larger and more sophisticated drilling bits and barrels yielded better recovery, including samples of clay seams from some boreholes. Similar methods to those at Auckland were used for logging and sampling from shafts.
4 Results of Field Investigations 4.1 Auckland Landfill Site The proposed landfill was to fill a natural valley bound to the east and south by prominent ridges and in the northwest by a broader ridge of lower elevation (Fig. 3). The landfill was designed to cover the headwaters of the ‘‘Central Stream’’ (Fig. 3) including one large and several smaller tributary gullies. Slopes between these gullies show widespread indications of landslides, as do the steep headwalls above them. There are numerous head-scarps, landslide blocks, debris mounds, areas of hummocky ground, irregular mid-slope benches and debris flow lobes (Fig. 3). Beyond the three enclosing prominent ridges,
123
slopes in adjacent streams show similar indications of landslides. The bedrock comprises relatively weak interbedded sandstone and mudstone dipping typically at 0–7 in various directions. Steeply dipping tectonic joints and faults with small vertical displacements are common. The absence of surface fault scarps and a general lack of shearing and crushing, indicate that active faulting is unlikely at the site. No datable deposits were found in the trenches, no offsets of soils observed and very few crush zones were located. Slope instability is characterised by earth and debris flows and debris slides in the top few to several metres and deeper seated block slides within weak rock (Fig. 3). Geomorphic mapping of the site indicated that disturbance due to slope movement was likely to extend down to 10–20 m below the ground surface. Semi-intact blocks of rock and soil inferred to have moved downslope on gently inclined planar surfaces, are estimated to have moved at least 30–100 m, at some time in the last few 1000 years. Rotational and planar slides and flows within the near surface materials occur within residual soils, completely
Block Slides on Extremely Weak Clay Tectonic Seams in Openly Folded Tertiary Mud-Rocks at…
Fig. 7 Extract of shaft log 36 showing clay seam at 14.6 m depth and marker horizon at 13.55 m
weathered rock, and in slide debris (material that has been displaced in an earlier slide). At least two bedding-parallel clay seams ([1 mm thick) or clay coated fractures (\1 mm thick) were observed in a number of shafts, trenches and stream-bed exposures. In combination with tectonic uplift, stream down-cutting and weathering, these defects (joints, faults, bedding planes) control the extent and depth of landslides observed at the site. In general, ridges are formed over anticlines and streams along synclines in the gently dipping, openly folded weak bedrock. Bedding parallel clay seams and steeply disjunctive fractures are thought to form the basal ruptures and side shears respectively, of block slides which then break down into debris slides. The enclosing boundary ridges of the site separate it from streams with similar landslides. The eastern ridge in particular (Fig. 3) is a narrow steep-sided spine, which separates the proposed landfill basin from the headwaters of a large stream flowing out to the sea at Whitford (Fig. 2). It was recognized that slope failure was possible not only within the proposed excavation for the landfill basin, but also of the eastern ridge under load from the completed landfill sliding eastwards into the stream, which
flows northwards to Whitford. This raised issues of possible damage to the stream valley, estuaries and coastline in the Whitford area. 4.1.1 Depth of Slope Movement Materials and Weathering Several of the shafts were drilled in areas of the site likely to be debris mounds or slide blocks as indicated from the geomorphology. The depth of dislocated material and of weathering identified in these shafts is shown in Table 1. Shafts indicate that dislocation and disturbance is mainly confined to the top 10–15 m of soil and rock. Weathering and a resultant general weakening of the rock mass is recorded to a maximum depth of 20 m in these shafts. 4.1.2 Clay Seams A thin clay seam or clay coated fracture was identified in most shafts and a similar feature was seen in a number of trenches. It is likely to have formed as a result of slight movements between adjacent beds (i.e. by flexural slip) at the time of folding. Locally, diverging and re-joining
123
W. M. Prebble, A. L. Williams Table 1 Depth of dislocated material and weathering at Auckland landfill site Shaft no./ depth (m)
Depth of dislocated material (m)
Depth of weathering (m)
37/28.8
4.2
4.2
39/36.6
0
6.0
41/21.0 42/20.0
0 8.2
5.8 8.5
52/34.1
9.3
10.0
55/40.0
13.6
20.0
splays of the main seam occur. These are together referred to as the clay seam. Figure 8 provides a summary of where the clay seam has been identified and its orientation. The clay seam varies in thickness over the site ranging from less than 1 mm up to 8 mm thick. This variation occurs over a single shaft diameter in S36 (Fig. 7). The clay seam was also observed to be bedding-parallel and wavy with bedding surfaces exhibiting an undulation of up to ±50 mm over a shaft diameter. Many small displacement faults have been observed in shafts and trenches (e.g. Fig. 6). These are shown to offset bedding planes by steps of up to 500 mm. The clay seam is also seen to be offset (for example offsets of 6 and 36 mm in S55). In combination with waviness, offset on sub-vertical faults results in local variations in the dip of bedding. Displacement of the clay seam by these minor faults suggests that the seam is an old tectonic feature (generated by shear along bedding planes), which may have facilitated recent block sliding by providing a low-angle basal slide plane, or is capable of providing a potential basal rupture surface. Along the in situ spine of the Eastern Ridge the clay seam is found at 15–35 m depth in the middle of a weak mudstone bed. Sandstone overlies and underlies the mudstone. In general, the rock is weaker and more openly fractured above and around the seam, and stronger and tighter below it. A second, upper clay seam (clayey crush zone) was located in two shafts (shafts 52 and 55) in the top 10–15 m and is thought to be the basal rupture surface for debris block slides at these locations. 4.2 Rangitikei North Bank The Rangitikei River and its tributaries have planed the Pliocene sandstones and mudstones into extensive flat to very gently dipping surfaces on which river gravels eroded from the greywacke ranges to the North and East have been deposited. Since then tectonic uplift, and down-cutting by the rivers, have left a pattern of gravel-capped terraces, probably in the order of 12,000 years old, into which the
123
river has cut a steep sided gorge 80 m deep. At the site the gorge walls are vertical. Bedding dips consistently throughout the site at 6–7 to the southwest (247). This simple structure extends throughout the surrounding region. The relationship of the dip to the geomorphology and geology of the site is shown in Figs. 9 and 10 which illustrate the site in plan and section. NNW to SSE striking active faults are mapped 7 km to the West and 5 km to the East of the site by Thompson (1981) and Journeaux et al. (1996). Parallel to these faults and 2.5 km east of the site is an active anticline. The faults and anticline extend for at least 30 km along the Rangitikei valley. Although block sliding is mapped by Thompson for several kilometres in every direction throughout the region surrounding the site, block sliding is concentrated around the faults and, in particular, around the active anticline. 4.2.1 Block Slides and Clay Seams Large block slides, from 300 m 9 300 m up to 4 km 9 2.5 km, exist for several kilometres upstream of the site and have developed along bedding-parallel clay seams. Some of these seams are close to the sandstone/mudstone transitional boundary and others are within these units. In the sandstones around the site a number of 2–20 mm thick clay seams were found. Large slope failures and clay seams are very infrequent downstream of the site in the thick, massive mudstone. Incipient block slides develop into short displacement slides. Where there was sufficient slope down-dip, block slides with large displacements have developed. These are often part of complex areas of coherent block slides, disintegrated block slides, debris flows and earth flows (Thompson 1981; Prebble 1979, 1995). The incipient block slide in the North bank of the bridge site is within sandstone, but at a stratigraphic level where two thin (5–8 mm thick) bedding-parallel clay seams have been found approximately 2 m apart, in the cliff face exposures and in shafts and boreholes. The seams consist of grey and brown, greasy, sheared and slickensided clay, containing chlorite as the abundant clay mineral. The clay seams control the depth and extent of block sliding. A short displacement block slide 200 m downstream of the site (Fig. 11) is indicated by a small detached hillock, approximately half the height of the terrace and separated from it by a 30 m wide gap. Further back there is a shallow, irregular, 20 m wide trough across the terrace. This trough is considered to be an incipient pull-away zone. Short displacement block slides have developed elsewhere in the region, very close to river banks, where there is room only for a short sliding displacement compared to the depth to the basal rupture surface. This is around 50 m deep
Block Slides on Extremely Weak Clay Tectonic Seams in Openly Folded Tertiary Mud-Rocks at… Fig. 8 Geological plan showing clay seam orientation
compared to 30 m sliding displacement straight down dip at this downstream locality (Fig. 11). The basal rupture dips beneath the river bed, limiting the amount of possible sliding. 4.2.2 Clay Seams and Open Fractures in the North Bank Promontory A lower seam dips 1 to a few degrees directly towards the river. It has slickensides in that direction and is assumed to have been a basal rupture surface to sliding of fracture-
bound blocks observed immediately downstream of the North bank promontory. Greasy shear surfaces exist within the clay seams and slickensides were seen on the seams and bounding surfaces. Open fractured zones, vertical to very steeply dipping towards the river bank were found throughout the sandstone and gravel on top of the North Bank promontory and probably extend down to the clay seam or close to it. Four main sets of fractured zones and several minor sets were uncovered in trenches cut into the top of the sandstone and in the overlying gravels. All of the zones strike parallel to
123
W. M. Prebble, A. L. Williams
Fig. 9 Elevation and section in position of bridge showing borehole positions and clay seam; north bank on the right side
North bank promontory. Substantial drilling water circulation losses in the fractured rock above the clay seams indicated high fracture permeability.
5 Mineralogy and Lab Scale Shear Strength of the Clay Seams 5.1 Mineralogy
Fig. 10 North Rangitikei bridge site; north bank on right side of photograph in sandstone
the river bank and one of them is continuous with a zone in the pile shaft extending down to just above the upper clay seam. Other fractured zones correlate with those found in the trenches. The fractured zones post-date the deposition of the 12,000 year old terrace gravels, thus providing a maximum age for the onset of incipient block sliding in the
123
The clay seams in the Southern Landslide Zone of Auckland are dominated by smectite. This is similar to the clay content of the enclosing host rock which also contains subordinate illite and lesser chlorite (Wylie 1989). At Rangitikei, Stout (1977) refers to ultra-thin montmorillonite (smectite) clay seams containing some illite, from samples collected 3 km north of the bridge and others in the area around the Bridge site. Clay seams closer to the Bridge consist of dark greenish grey waxy fissile clays with 45–60 % smectite and Ca/Mg exchangeable cations (Thompson 1981). A clay sample from the sheared surface of a shear box test specimen, cut from a borehole core at the site, consisted of mixed layer clays containing chlorite,
Block Slides on Extremely Weak Clay Tectonic Seams in Openly Folded Tertiary Mud-Rocks at… Fig. 11 Elevation and section showing incipient block slide
montmorillonite (smectite) and muscovite. Hence the seams contain a mixture of clays, of which smectite is the most prevalent. 5.2 Shear Strength Seventeen clay seam samples from the Auckland site were tested in a standard shear box apparatus to assess the resistance to sliding. The samples were tested under a range of normal stresses (25–600 kPa). One sample (S55/8) was sheared back and forth 20 times under a normal stress of 5 kPa at the end of the standard test and tested again to define the residual shear stress by conventional methods.
Results of shear box testing of the lower continuous clay seam are presented in Fig. 12. In general, intact rock tested in the shear box apparatus shows a well-defined peak, followed by a rapid and then gradual fall off. Because the clay seam and clay coated fracture are planes formed by tectonic shearing, the curve rises more gradually and asymptotes to a ‘‘residual’’ value. This is indicative of a material that has sheared in the past. In some tests of the clay coated fracture, shearing has occurred in part through the fracture and in part through rock. When this happens, the curve continues to rise and gives a high value of shear strength. These values have not been taken into account in further evaluation of the test data (box symbol in Fig. 12).
123
W. M. Prebble, A. L. Williams Fig. 12 Shear box test results for clay seams and clay-coated fractures. In some tests shearing has occurred in part through the fracture and in part through rock. These values are indicated by the box symbol
Regression analysis suggests an average shear strength of c0 = 13 kPa and ø0 = 13 for the clay seam and clay coated fracture (comparable to parameters obtained for clay seams encountered in Waitemata Group by Williams and Prebble 2004). The lowest parameters measured were c0 = 0 kPa and ø0 = 8. Variation in thickness, waviness and displacement of the clay seam combine to increase the resistance to sliding. Wylie (1989) performed ring shear tests on a beddingparallel clay seam from a tunnel excavation 5 km south of the landfill site. The results showed that with a normal load of 200 kPa the residual angle of friction was 10 and cohesion was 0 kPa. This is very similar to the lower bound values shown in Fig. 12. 5.2.1 Shear Strength of the Clay Seams at Rangitikei In the field, the clay seams are extremely weak, ‘‘soft’’ and certainly much weaker than the sandstone. A direct shear box test was performed with three stages on a 100 mm diameter core sample from a borehole on the North bank of North Rangitikei bridge site. The sample contained a natural shear plane parallel to bedding within a 100 mm layer of grey, greasy, wet clay within the sandstone rock mass. The sample also contained an irregular horizontal fracture. The shear plane dips approximately 10 to the horizontal, which is compatible with the local dip of 7 at the site. The sample was placed in the shear box with the shear plane oriented horizontal. Shearing took place on the natural shear plane, with maximum shearing stresses probably characterizing undrained residual shear strengths. Strength envelopes obtained were: Initial yield points, c = 40 kPa and / = 10. After 4 mm displacement, c = 240 kPa and / = 15. These shear strengths are greater than those recorded from the Auckland region (Fig. 12). A possible explanation is developed in the discussion Sect. 9.2.
123
6 Site Scale Models of Clay Seam Persistence and Geometry 6.1 Development of a Geological Model for the Auckland site The low dips and open folds of the simple structural domain have created slopes with beds dipping less than or parallel to the slope angle, exposing bedding planes and bedding–parallel defects such as clay-coated fractures and clay seams, in coastal cliffs and stream banks. Bedding dips in the order of 0–15 but is variable in direction. Steeply dipping tectonic joints and faults with vertical displacements ranging from millimetres to several metres are common at the site and throughout the Howick– Alfriston area. The nature, continuity and strength of the clay seam are critical to stability. It was therefore important to determine whether the clay seam identified at each shaft, trench or exposure was part of a single continuous feature or whether several discontinuous seams underlie the site. Faulting and folding, lateral and vertical variation in bed thickness and grain size, and amalgamation and separation of beds make correlation of individual beds over tens of metres difficult, even when fully exposed. Correlation is particularly difficult where exposure is discrete and spread over hundreds of metres (i.e. from outcrop to outcrop or shaft to shaft). Shaft logs were examined in detail in an attempt to identify a specific sequence of beds, which could be used to correlate between shafts. Aside from the clay seam, the most notable correlative feature was a tuffaceous bed described as a 20–50 mm thick dark grey to black strong, brittle mudstone with up to 30 mm of pink medium sandstone at its base. Tuffaceous beds are rare within the Auckland Miocene strata. The tuffaceous sequence therefore provided a useful marker
Block Slides on Extremely Weak Clay Tectonic Seams in Openly Folded Tertiary Mud-Rocks at… Table 2 Clay seam and marker horizon beds in shafts Shaft no.
Seam description/depth (m)
Marker bed/depth (m)
36
Seam/14.6
Yes/13.55
37
Not Observed
Not observed
38
Seam/26
Yes/23.4
39
Seam/20.1
Yes/17.6
40
Clay coated fracture/34.7
Yes/32.25
41
Seam/20.6
Yes/18.5
42
Not observed
Not observed
49
Clay coated fracture/17.9
Not observed
50
Not observed
Not observed
52
Seam/25.7
Yes/22.95
54 55
Clay coated fracture/22.5 Clay coated fracture/18.3
Yes/20.0 Not observed
horizon occurring at an average of 2.3 m above the clay seam in the shafts as shown in Table 2. The tuffaceous marker bed was not observed in two of the nine shafts in which the clay seam was encountered. It is possible that the marker bed has been locally eroded or thinned immediately after sedimentation to the point where it was not recognised in these shafts. The consistent correlation of the clay seam and tuffaceous bed and the absence of confirmed evidence to the contrary indicate that the clay seam can be regarded as continuous over the site and parallel to bedding. If the clay seam is assumed to be a single continuous horizon, the geological model implied from the data on the map (Fig. 8)
and shown as sections in Fig. 13, can be developed from structural contours of the seam (Fig. 14). Figure 13 indicates that the current topography broadly mirrors the underlying geological structure. A gentle anticlinal fold underlies the Eastern Ridge and a gentle synclinal fold occurs about an axis which is approximately coincident with the central stream valley. An anticlinal fold axis is also seen outcropping west of the crest of the Northwestern Ridge. Hence, bedding and the clay seam dip gently into the landfill bowl from the Eastern, Southern and North-western Ridges, and gently away into the adjacent streams on the outside of the North-western and Eastern Ridges. Throughout the Southern Landslide Zone, anticlinal ridges and synclinal valleys are conducive to widespread slope instability on bedding-parallel clay seams. A comparison between the dip of the clay seam and the lab values for friction angle reveals a noticeable difference. The friction angles of the clay seam (regression output 13 and lower bound 8) are steeper than the dip of the clay seam (1–15, average 5.5). Most clay seam dip angles are between 3 and 8. 6.2 Development of a Geological Model for the Rangitikei site The importance of bedding-parallel clay seams as the basal rupture zones of large block slides and debris slides in the Rangitikei region was revealed by Stout (1977), Prebble (1979) and Thompson (1981). This work provided a conceptual geological model of large block slides moving
Fig. 13 Cross-section showing clay seam
123
W. M. Prebble, A. L. Williams
the gorge of the river. The tension fractures are considered to extend through to the surface of the river banks upstream and downstream of the site, isolating the potential slide blocks from any side shear resistance. Shear strength of the clay seam at the site scale is therefore critical.
7 Clay Seam Geometry and Shear Strength at the Site Scale 7.1 Clay Seam Geometry
Fig. 14 Structural contours of the clay seam (shown in m RL) mirror topography
down dip on clay seams, in particular at the boundary between sandstone and underlying mudstone. At North Rangitikei Bridge the orientation, continuity and strength of the clay seams are critical to the stability of the North bank abutment. 6.2.1 The Incipient Block Slide Model Given the consistent dip direction, dip inclination, and very consistent thickness of the sandstone at the site, in the immediate area and in the surrounding region, a similar continuity and consistent attitude of the clay seam found at 39.6 m depth in boreholes and a shaft are not surprising. Similarly the lower seam found in the North bank exposure could be expected to have the same continuity and attitude. The existence of a short displacement block slide 200 m downstream from the site, exactly in the direction of dip of the two clay seams increased confidence in the model. The trenches, boreholes and the shafts also demonstrated the presence of open continuous vertical tension fractures from the terrace gravels down to the clay seam. Stratigraphy showed they are younger than the c 12,000 year old terrace gravels. Trenches through the gravels demonstrated the continuity of the fractures (at least four) throughout the north bank promontory. The site engineering geological model is shown in map and cross-section in Fig. 11 and has been referred to as an incipient block slide or ‘‘proto block slide’’ (Prebble 1995). The slide sits on two clay seams and is bounded at the back by open tension fractures. The clay seams and tension fractures are considered to be continuous throughout the North bank promontory, which is bounded by a 90 bend in
123
At the Auckland landfill site, many faults with small displacements in trenches and shafts offset bedding vertically by steps of up to 500 mm. The clay seam is also offset, but to a lesser extent, with vertical steps of 6–36 mm in height. Locally, diverging and re-joining splays of the seam are seen in trenches and a shaft. The clay seam also varies in thickness from less than 1 to 8 mm over a distance of 1 m. This creates waviness of the seam walls. Over a distance of approximately 1 m in the walls of some shafts, there is waviness of up to ±50 mm. Similar features were also observed for 30 m along the clay seam in a tunnel excavation, 5 km south of the Auckland landfill site and within the southern landslide zone. Considerable relief as undulations, steps and variation in thickness of the seam were evident on a scale similar to that observed in shafts and trenches at the landfill site. In a few areas of the tunnel wall exposure, steps of 10–30 mm over a distance of 300 mm were recorded (Wylie 1989). Polished and crushed wall rock, foliated fabric in the seam and rounded inclusions of wall rock were common. Injection of clay up into fractures in the overlying rock was frequently observed. This may have taken place during flexural slip caused by folding and uplift, or it may be attributable to dilation of fractures above the seam during incipient block sliding (Prebble 1995). In an analysis of slope stability at the tunnel location, Wylie concluded that the slope was stable under dry conditions but elevated groundwater levels or a significant local earthquake would produce a factor of safety of \1 and potentially failure. A comparable degree of waviness and variation in thickness of the clay seams was observed at Rangitikei North Bank but stepping was not. The seams at Rangitikei are younger and not disrupted so that site scale strength may be closer to lab scale values than it is at Auckland. 7.2 Shear Strength at the Site Scale Shear strength of the clay seams at the site scale will be greater than the lab scale strengths, as a result of the waviness, thickness variation, vertical offsets and steps in
Block Slides on Extremely Weak Clay Tectonic Seams in Openly Folded Tertiary Mud-Rocks at…
the seams. These provide extra resistance to shearing, over and above the lab scale strength. In a regional study of landslide hazard assessment in Auckland during earthquake and heavy rainfall, Williams and Prebble (1998) indicated that 80 % of slopes in the Southern Landslide Zone would fail in the 2000 year return period earthquake and 20 % of slopes during the 100 year heavy rainfall event. It seems likely that even the lower bound values of lab scale shear strength for the Auckland site require elevated groundwater levels or a significant earthquake to cause failure. This reflects the effect of waviness, steps, splays and rock inclusions on site scale strength in the clay seams.
8 Impact on Landslide Formation, Type, Actvity, Age and Volume The continuity, extremely low residual strength, beddingparallel orientation and very low dips of the seams creates extensive areas of disrupted landslide topography resting on a gently dipping planar surface. First time failures are block slides. As displacements increase, disintegration of blocks produces debris slides and flows. Generations of these develop over time, resulting in a complex and somewhat chaotic landslide mass. Slope angles and dips of beds and seams are commonly that of the friction angle of the seams. In large complex block slides where considerable disintegration of blocks has produced clayey debris, secondary rotational failures and flows are widespread. Many younger slides are active, with reactivation of all or part of older slides also taking place. Thompson (1981) recognised imminent (incipient) and very young failures, modern (active) failures, old (dormant) failures and ancient (and remnants of ancient) failures. Andesitic and rhyolitic ashes from further north, in the Taupo Volcanic Zone (Fig. 1), have been distributed over the basin in the Late Quaternary. These provide marker horizons, which give some indication of the age of terraces, slope deposits and landslides (Thompson 1981). Nearly all failures in the Rangitikei region are younger than 11,000 years B.P. The majority are either less than 1800 or 1800–11,000 years. Block slides are younger than 1800 years. Older slides must be considered as sites for a number of possible reactivations as indicated by the ancient slide described by Stout (1977). Hence it is useful to consider age ranges and not discrete ages. Approximate volumes vary from 0.4 to 0.03 km3. Given the highly irregular profile and thickness of older slides it is useful to consider areal dimensions, which for the larger block slides range from 300 m 9 300 m up to 4 km 9 2.5
to 3.0 km. Smaller block slides are approximately 50 m 9 50 m, up to 100 m 9 60 m.
9 Discussion 9.1 Formation and Geological Origin of Clay Seams The Tertiary mud–rocks which host the seams have a detrital clay-rich matrix and a variable component of clayrich rock grains, derived from weathered rock in the sediment source areas. As a result, significant clay content was included in the original sediment and is present throughout the rock mass, including the un-weathered rock at depth. Concentrated shearing on a particular bedding surface would inevitably produce sheared and foliated clay from the clay fraction of the original host rock. Even where clay seams are encountered in deep shafts at depths where no slope failure is possible, the seams are polished and/or striated. Clay mineralogy of the seams is the same as the clay matrix of the enclosing rock. The seams are most common in the sandstone beds and are parallel to bedding, usually at or close to the boundary with thin mudstone beds. The physical contrast between the mudstone and sandstone probably assists shearing and clay seam development as a result of flexural slip during folding of the beds. A flexural slip origin for bedding plane shears was proposed and discussed by Fell et al. (1988) in gently dipping claystone and siltstone. The shears provided rupture surfaces for slope failures. Hutchinson (1988, 2001) considers that the capability and potential of flexural slip has been underestimated as an origin for pre-existing shears, which develop into basal ruptures of landslides. Possibly some seams were originally discrete mudstone beds or beds of weathered volcanic ash within sandstone. Others are not seams but clay coated fractures. All are parallel to bedding and considered to be of tectonic origin as a result of flexural slip. This could have taken place during regional compression, uplift and open folding, which are on-going in the Rangitikei Valley and appear to be related to active faults and a growing anticline within 5–10 km of the site. Therefore clay seams in openly folded Tertiary mudrocks are considered to have formed by flexural slip of the beds during open folding as a result of regional compression and uplift. The uplift in Auckland was mainly in Pliocene time but has continued since at a lower rate than elsewhere in the plate boundary. However at Rangitikei, significant uplift (5–8 mm year-1), thrusting, compression and folding are on-going today (Milne 1975; Thompson 1981).
123
W. M. Prebble, A. L. Williams
A similar tectonic style and rate prevails in other regions of the New Zealand convergent margin, where active thrusting and folding have formed bedding parallel clay seams in clay-rich Tertiary strata and uplift and erosion have allowed the initiation of block slides. Examples are seen at Hawkes Bay (Pettinga 1987), Dunedin (Coombs and Norris 1981; Hancox 2009), Marlborough (Prebble 1987) and south Nelson (Lensen and Suggate 1968). 9.2 Field Methods Required to Successfully Investigate Clay Seams Mapping, followed by large diameter shafts and trenches were the critical steps in proving location, orientation, continuity and characterisation of the clay seams. The preliminary geological model at Auckland of anticlinal ridges and synclinal valleys formed by block sliding along clay seams in the gentle open folded weak bedrock was confirmed by the subsurface investigation in combination with detailed geomorphic and geological mapping (Williams and Prebble 2010). Cored boreholes were inadequate, but inspection shafts highly successful in confirming depth, attitude and geotechnical properties of the clay seam. The increased value and reliability of shafts and drives over boreholes was emphasised by Gillon et al. (1992) in their investigation of the Cromwell Gorge schist landslides. Our experience with shafts at the Auckland landfill site and Rangitikei has shown similar advantages of direct observation of subsurface rock conditions and potential failure surfaces, improved correlation of defects and the ability to collect samples for laboratory testing. The investigation shafts at Rangitikei were decisive in identifying open, continuous fractures and the trenches established their continuity across the site in at least four zones parallel to the bank. In conjunction with the boreholes, the shafts also provided essential direct confirmation of the level, attitude, thickness, continuity and consistency of the clay seams. In common with the Auckland landfill site, the shafts were distinctly advantageous for in situ observation, measurement and sampling. The field data collected in this study would not have been obtained without the high quality detailed information gained by in situ inspection, logging and sampling of clay seams and fractures in the investigation shafts and trenches. Features in common from all of the examples discussed in this paper include the importance of using the geological and geomorphological context to develop a preliminary model. Detailed geomorphic mapping in the site area and for a considerable distance around it can be invaluable in providing the appropriate information. Investigation shafts and trenches are decisive in obtaining critical data of high quality at the site scale.
123
9.3 Mineralogy and Lab Scale Shear Strength of the Clay Seams The mineralogy of the clay seams at both Auckland and Rangitikei is mixed; smectite is dominant, with subordinate chlorite, illite and muscovite. The composition of this mixture closely parallels that of the host rock clay fraction, from which it has been derived during flexural slip shearing. The shear strengths of the seam at Rangitikei Bridge North Bank are greater than those recorded from the Auckland region. If the single specimen from Rangitikei Bridge is accepted as representative, a possible explanation may be the age difference and contrasting geological and geomorphic histories of the two regions. The rocks from Auckland are of Miocene age, around 20 Ma. They have been subjected to compression in a convergent tectonic margin, uplift, open folding and have been deformed to a far greater extent than the sandstones at Rangitikei which are much younger at Pliocene age, around 3–4 Ma. The Auckland clay seams have been tectonically sheared much more than the seams at Rangitikei and since uplift and exposure have experienced more block sliding shear. It is also possible that because the seams at North Rangitikei Bridge North abutment have apparently been subjected to less block-sliding shear (as the basal shear to an incipient block slide), they retain more strength than the seams tested from Auckland. In direct contrast, Stout (1977) calculated very low values of shear strength for the slip-surface materials from the reactivated slide 3 km north of the Rangitikei Bridge. At a safety factor of 0.85 (condition of movement), the values were a cohesion of 9.5 kPa and angle of internal friction of 3. At a safety factor of 1.0 (condition of temporary stability) and the same cohesion, the friction angle was 4. These are considered to be residual values given the on-going historic movement and evidence of considerable disruption and degradation of the landslide mass. The friction angles which Stout calculated are the same as the angle of dip of the basal slide plane in the direction of slide movement. This indicates that, as would be expected, considerable strength is lost during block sliding, especially during repeated sliding of a reactivated, ancient landslide. The Rangitikei values represent two extremes of the geomorphic spectrum, a basal clay seam of an incipient block slide and a basal slip surface to a reactivated ancient slide. 9.4 Site Scale Models of the Clay Seam Persistence and Geometry The pattern of folding described in the Auckland example, and subsequent shaping of the land by erosion have
Block Slides on Extremely Weak Clay Tectonic Seams in Openly Folded Tertiary Mud-Rocks at…
resulted in the bedding (and the clay seam which is parallel to it) dipping downslope in most places. As a result, the potential for deep seated sliding along the clay seam became the ubiquitous model. Structural contours of the seam allowed assessment of the stability of the site. Although the dominant bedding dip is near horizontal, the dip direction varies widely. These variations in bedding orientation are consistent with either folding or faulting (or both) of the beds. It is considered likely that in some places they are due to rotational displacements across near vertical faults. It is clear from both the regional and site evidence that the continuity of bedding and of the clay seam is disrupted by steps across small faults, indicating that the clay seam is unlikely to form very long continuous surfaces, but instead will be stepped into shorter sections. At Rangitikei, geomorphic assessment and comparisons within the local region enabled identification of the incipient slide model (proto block slides) in the North bank and led to the recognition of short displacement block slides. 9.5 Shear Strength at the Site Scale The steps and variations in dip cause the clay seam to be ‘‘wavy’’, increasing its resistance to sliding movement. As a result, strength at the site scale is expected to be greater than lab scale strength. However, the lab friction angles (13 and 8) are greater than the clay seam dip angles (average 5.5). It is considered that the waviness and steps in the seam would increase the strength. Large steps in the clay seam would result in subdivision of slide areas into a complex of smaller failures. This is probably the case at Auckland, where the slides are much smaller than at Rangitikei. Average strength parameters for the clay seam at Auckland were therefore considered appropriate for stability analyses. It also seems that earthquake or high intensity rainfall is necessary to induce slope failure. A lesser degree of deformation and simpler structure at Rangitikei has possibly resulted in less waviness and stepping at the site scale. Site scale strengths may be closer to lab scale values. 9.6 Impact on Landslide Formation, Type, Volume, Activity and Age The different tectonic history and current tectonic setting of the Auckland Landfill and Rangitikei Bridge sites have a different impact at each of those localities. When the regional setting of the southern landslide zone in Auckland is compared to that of the Rangitikei River gorge, it is apparent that the greater age and more intense geological history of the Auckland region have resulted in a more fractured and faulted rock mass. This produced smaller block slides on clay seams and more diverse
complex topography of anticlinal ridges and synclinal valleys. Most Auckland slides are dormant in the current less tectonically active regime, but are capable of reactivation in earthquake and intense rainfall. The age of existing landslides is uncertain but possibly ranges over the last few 1000 years. By contrast the area around the Rangitikei Bridge is notoriously unstable. In response to the currently very active tectonic regime, there are many active younger slides and reactivating ancient slides. Uplift, active folding and river down-cutting perpetuate block sliding and reactivation of debris. Landslides are much larger than at Auckland and more diverse. Block slides on clay seams are younger than 1800 years. Acknowledgments Investigation work in Auckland, commissioned by Waste Care (NZ) Ltd, a subsidiary of Browning Ferris Industries (BFI) Ltd, was undertaken by Beca Carter Hollings and Ferner Ltd (Beca) under the directorship of Dr Do Van Toan. Special thanks are extended to all those involved with the project and in particular to Graham Mansergh who undertook detailed logging of many of the shafts and to BFI for approving innovative investigation methods and for permission to publish this work. At Rangitikei the investigation work was commissioned by New Zealand Railways and carried out by Beca Carter Hollings and Ferner Ltd under the direction of John Hollings, Gavin Cormack, John Blakeley and Dr Do Van Toan.
References Brown IR (1974) The stability of slopes in Tertiary sedimentary rocks of New Zealand. N.Z. Geomechanics Society. In: Proceedings of the symposium on stability of slopes in natural ground, pp 7.23–7.33 Coombs DS, Norris RJ (1981) The East Abbotsford, Dunedin, New Zealand Landslide of August 8, 1979. An interim Report. Bulletin de liaison de Laboratoire des Ponts et Chaussees Paris. Special edition X. In: Proceedings of the 26th international geological congress, Paris; section 17, pp 27–34 Edbrooke SW (compiler) (2001) Geology of the Auckland area. Institute of Geological and Nuclear Sciences 1:250,000 geological map 3, 1 sheet ? 74 p. Institute of Geological and Nuclear Sciences Limited, Lower Hutt Fell R, Sullivan TD, MacGregor JP (1988) The influence of bedding plane shears on slope instability in sedimentary rocks. In: Proceedings of the 5th international symposium on landslides, Lausanne vol 1, pp 129–134 Gillon MD, Denton BN, McFarlane DF (1992) Field investigation of the Cromwell Gorge landslides. In: Bell DH (ed) Landslides vol 1. Balkema. Proceedings of the 6th international symposium on landslides, Lausanne, vol 1, pp 3–35 Hancox GT (2009) The 1979 Abbotsford landslide in Dunedin, New Zealand: a retrospective case history thirty years after the disaster. New Zealand Geomechanics News, December 2009, issue 78, pp 24–44 Hawley JG, Riddolls BW (1975) Soft rock engineering in the central North Island of New Zealand: Australia—New Zealand Conference on Geomechanics, 2nd, pp 201–206 Hutchinson JN (1988) Morphological and geotechnical parameters of landslides in relation to geology and hydrogeology. In: Proceedings of the 5th international symposium on landslides, Lausanne, vol 1, pp 3–35
123
W. M. Prebble, A. L. Williams Hutchinson JN (2001) Landslide hazard assessment. In: Bell DH (ed) Landslides, vol 3. Balkema. Proceedings of the 6th international symposium on landslides, Christchurch, New Zealand Journeaux TD, Kamp PJJ, Naish T (1996) Middle Pliocene cyclothems, Mangaweka region, Wanganui Basin, New Zealand: a lithostratigraphic framework. NZ J Geol Geophys 39(1):135–149 Ker DS (1970) Renewed movement on a slump at Utiku. NZ J Geol Geophys 13:996–1017 Kermode LO (1986) Sheet N42/9 Whitford, Geological map of New Zealand, 1:25,000. Industrial Series. Department of Scientific and Industrial Research, New Zealand Kermode LO (1992) Geology of the Auckland urban area. Scale 1: 50,000. Institute of Geological and Nuclear Sciences geological map 2. 1 sheet ? 63 p. Institute of Geological and Nuclear Sciences Ltd, Lower Hutt Lensen GJ, Suggate RP (1968) Inangahua Earthquake—preliminary account of the Geology. In: Preliminary Reports on the Inangahua Earthquake, New Zealand, May 1968. Department of Scientific and Industrial Research Bulletin, vol 193, pp 17–36 Milne JDG (1975) Upper Quaternary Geology of the Rangitikei Drainage Basin, North Island New Zealand. PhD thesis, Victoria University of Wellington Pettinga JR (1987) Waipoapoa Landslide: a deep seated complex block slide in Tertiary weak rock flysch, Southern Hawkes Bay, New Zealand. NZ J Geol Geophys 30:401–414 Prebble WM (1979) Report on North Bank of North Rangitikei Railway Bridge. Geological Report on Fractured Zones. August 1979, Amended 6th December, 1979. Prepared for Beca, Carter, Hollings & Ferner. Applied Research Office. The University of Auckland. ARO/628 Prebble WM (1987) Slope movements in limestones and shales, North-East Marlborough, New Zealand. PhD thesis in Geology, The University of Auckland Prebble WM (1995) Landslides in New Zealand. Keynote Paper. In: Bell DH (ed) Proceedings of the Sixth International Symposium
123
on Landslides. Christchurch, New Zealand. February 1992, vol 3, pp 2101–2123 Prebble WM (2001) Hazardous Terrain—an engineering geological perspective. In: New Zealand geotechnical society symposium on engineering and development in hazardous terrain. Christchurch, August 2001. Institution of Professional Engineers New Zealand. Proceedings of Technical Groups vol 29 Issue 2 (GM) 2001, pp 1–31 Stout ML (1977) Utiku landslide, North Island, New Zealand. Geol Soc Am Rev Eng Geol 111:171–184 Thompson RC (1981) Landsliding in Cenozoic soft rocks of the Taihape-Mangaweka area, North Island, New Zealand. Bulletin de liaison de laboratoire des ponts et chausses, Paris. Special edition. In: Proceedings of the 26th international geological congress, Paris; section 17, pp 93–100 Tilsley S (1993) Engineering geology and mass movement of the Howick–Alfriston–Whitford district, South East Auckland, New Zealand. MSc thesis in Geology, The University of Auckland Williams AL, Prebble WM (1998) Auckland engineering lifelines project: seismic hazard assessment. In: Proceedings of the 8th congress of the international association for engineering geology and the environment. Vancouver, September 1998, pp 809–815 Williams AL, Prebble WM (2004) Engineering properties of clay seams within the Waitemata Group rocks of Auckland. In: Proceedings 9th Australia New Zealand conference on geomechanics, Auckland, vol 2, pp 840–844 Williams AL, Prebble WM (2010) Subsurface investigation of defects; proposed sanitary landfill, Whitford, New Zealand. In: Williams AL, Pinches GM, Chin CY, McMorran TJ, Massey CI (eds) Proceedings of the 11th IAEG Congress. CD. CRC Press Taylor & Francis Group, Auckland Wylie CA (1989) Engineering Geology of a Soft, Weak Rock Tunnel: ARA Redoubt Road No. 2 Inlet Tunnel. Appendix II. Slope stability analysis of the Redoubt Road reservoir site. MSc thesis in Geology, The University of Auckland pp 139–142