ADDENDUM STUDY REPORT

No. 120A (2008)

SOIL EXPANSIVITY IN THE

AUCKLAND REGION
Fraser Thomas Ltd
(B.J. Brown, J.P.M. Shorten,
D.N.R. Dravitzki, P.R. Goldsmith)

The work reported here was funded by the Building Research Levy
and Manukau City Council, Rodney District Council, North Shore City Council, Franklin District
Council and Auckland City Council

BRANZ 2008
ISSN: 1178-4938

SOIL EXPANSIVITY IN THE AUCKLAND REGION

BRANZ Study Report SR 120A (2008)

Fraser Thomas Ltd

B.J. Brown, J.P.M. Shorten,
D.N.R. Dravitzki, P.R. Goldsmith

Reference
Fraser Thomas Ltd, Brown BJ, Shorten JPM, Dravitzki DNR and Goldsmith PR. 2008. Soil
Expansivity in the Auckland Region. BRANZ Study Report 120A. BRANZ Ltd, Judgeford,
New Zealand.

Preface
This addendum report provides further assessment on the expansivity of soils at six sites
within the Auckland region and considers the applicability of the design methodology set out
in AS 2870:1996 Residential Slabs and Footings Construction for buildings constructed in
accordance with NZS 3604:1999 Timber Framed Buildings.

Acknowledgments
This work was jointly funded by the Building Research Levy and and Manukau City Council,
Rodney District Council, North Shore City Council, Franklin District Council and Auckland
City Council.
On behalf of Fraser Thomas Ltd, this report was written by Mr Barry J Brown, Dr Peter R
Goldsmith, Mr J Patrick M Shorten and Mr David NR Dravitzki and was peer reviewed by
Dr Peter Mitchell, Adelaide and Professor Michael Pender, University of Auckland.
This document draws on and incorporates work undertaken within the preceding Stage I
project, whose authors were Mr Barry J Brown, Dr Peter R Goldsmith, Mr J Patrick M
Shorten and Mrs Leanne Henderson.

Note
This report is intended for researchers, geotechnical and structural engineers, property
developers and other workers in the field of building construction to NZS 3604:1999.

Contents:

Page

1.0

INTRODUCTION .........................................................................................................1

2.0

RESEARCH APPROACH ............................................................................................2

2.1
Introduction ........................................................................................................2
2.2
Staged investigation ...........................................................................................2
2.3
Overall investigation programme .......................................................................3
2.4
Extensometers ....................................................................................................4
2.5
Building damage survey .....................................................................................4

3.0

LITERATURE REVIEW ..............................................................................................5

4.0

NEW ZEALAND AND RELATED STANDARDS .....................................................5

4.1
Introduction ........................................................................................................5
4.2
NZS 3604:1990 Code of Practice for Light Timber Framed Buildings Not
Requiring Specific Design .................................................................................6
4.3
NZS 3604:1999 Timber Framed Buildings ........................................................6
4.4
AS 2870:1996 Residential Slabs and Footings Construction..........................7
4.5
AASHTO Designation: T 258-81 Standard Method of Test for
Determining Expansive Soils.............................................................................7

5.0

SITE CLASSIFICATION UNDER AS 2870 ................................................................8

5.1
Introduction ........................................................................................................8
5.2
Calculation of characteristic surface movement ................................................8
5.3
Soil suction change profile .................................................................................9
5.4
Instability index ................................................................................................11
5.5
Other relevant considerations relating to AS 2870 ..........................................13

6.0

CLIMATE ....................................................................................................................19
6.1
Introduction ......................................................................................................19
6.2
Auckland climate information ..........................................................................19
6.3
Climate factors for Australia and Auckland .....................................................21
6.4
Comparison of Australian and Auckland climates ...........................................22

7.0

MINERALOGY OF SOILS ........................................................................................23

7.1
General .............................................................................................................23
7.2
Australian soils .................................................................................................23
7.3
Auckland soils ..................................................................................................25
7.4
Comparison of soils from Auckland and selected Australian centres ..............27

8.0

FIELD AND LABORATORY TESTING FOR STAGE II ........................................28

8.1
Site selection ....................................................................................................28
8.2
Extensometer field............................................................................................28
8.3
Sampling and monitoring .................................................................................31
8.4
Laboratory testing ............................................................................................36

9.0

RESULTS OF FIELD EXTENSOMETERS AND LABORATORY TESTING .......38

9.1
General .............................................................................................................38
9.2
Extensometer results ........................................................................................38
9.3
Laboratory test results ......................................................................................39
9.4
Correlations between shrinkage index and other soil classification
parameters ........................................................................................................42

10.0

CLASSIFICATION OF SUMMER CONDITIONS ...................................................44

10.1 Extrapolation for design drought conditions ....................................................44
10.2 Soil suction results ...........................................................................................47
ii

10.3

Measured ground surface movements and correlation with inferred soil

suction change ..................................................................................................50

11.0

DRY SUMMER PREDICTIONS FOR AUCKLAND REGION ...............................53

11.1 General .............................................................................................................53
11.2 Statistics extreme value analysis ...................................................................54
11.3 Results and extrapolations ................................................................................56
11.4 Scaling factors ..................................................................................................59
11.5 Foundation performance assessment ................................................................60
11.6 Effects of climate change .................................................................................61
11.7 Soil suction change profile for regions other than Auckland ...........................61

12.0

RESIDUAL SWELLING POTENTIAL IN AUCKLAND SOILS.............................62

12.1 Objective ..........................................................................................................62
12.2 Option for limiting soil expansivity .................................................................62
12.3 Conclusion ........................................................................................................62
12.4 Limited swelling potential ................................................................................63
12.5 Application of AS 2870 standard methods ......................................................64

13.0

FOUNDATION ANALYSIS: PERFORMANCE FRAMEWORK ............................65

14.0

NEW ZEALAND BUILDING CODE REQUIREMENT ...........................................68

15.0

FOUNDATION DESIGN DECISION TREE AND FLOW DIAGRAM ................72

16.0

CONCLUSIONS AND RECOMMENDATIONS ......................................................74

16.1 Conclusions ......................................................................................................74
16.2 Recommendations ............................................................................................79

APPENDICES
A
B
C
D
E
F

iii

Bibliography
Borehole logs
Site summary sheets
Cumulative deficit data and Gumbel Analyses
Appendices B & C from AS 2870 Standard
Foundation analysis

1.0

INTRODUCTION
Expansive soils are those that experience appreciable volume change when the soil
moisture is altered. Soil moisture may be altered by a number of factors which may
act in combination including seasonal influence, the effects of trees, drains, roads etc.
The swelling and shrinking of soils can adversely affect buildings.
A significant proportion of new residential construction in New Zealand has concrete
slab-on-ground floors. Before the introduction of New Zealand Standard
NZS 3604:1999 Timber Framed Buildings a minimum founding depth of 450 mm
below cleared ground levels was specified in NZS 3604:1990 Code of Practice for
Light Timber Framed Buildings Not Requiring Specific Design, its predecessor
Standard, as a means of mitigating the effects of expansive soils on light timber
framed buildings in New Zealand. However, NZS 3604:1999 specifically excludes
foundations on expansive soils from its scope and refers the designer to Section 17 of
the Standard for additional information on expansive soils. In Section 17 it is
suggested that the designer refer to the Australian Standard AS 2870:1996 Residential
Slabs and Footings Construction as a means of classification of expansive soil sites
and providing a standard footing design, or that a specific engineering design be
provided.
Traditionally, the founding depth of 450 mm below cleared ground level has been the
benchmark for residential building construction in New Zealand for buildings
supported on conventional shallow foundations. There is, however, a move in recent
years towards waffle or rib-raft slab construction for residential buildings, which are
founded at the ground surface and are stiffened to AS 2870 standards according to
the site expansive soil classification.
There is uncertainty as to the relevance or applicability of AS 2870 to the design and
construction of foundations for residential buildings in New Zealand. In particular,
AS 2870 does not provide any New Zealand-specific design parameters to support its
application to New Zealand climatic and soil conditions. AS 2870 specifically relies
on knowledge of the characteristic change in soil suction profile for any particular
region or soil profile, as well as the shrink-swell properties of the soil and the depth of
seasonal shrinkage cracking.
Recognising the issues surrounding the conditions imposed by the 1999 edition of
NZS 3604, the Building Research Association of New Zealand (BRANZ) and the
Manukau City Council jointly funded an investigation of the expansive characteristics
of soils in the Auckland region. A report was produced by Fraser Thomas Ltd entitled
BRANZ Study Report 120 (2003) Soil Expansivity in the Auckland Region.
This current report has been prepared as an addendum report to the 2003 Study Report
and follows the same methodology presented in this, and should therefore be read in
conjunction with it. Some sections of the 2003 Study Report have been incorporated
in the current report for ease of reference.
The research reported herein has involved field investigations and laboratory testing
at six locations within the Auckland region, referred to as Sites 2A to 2F, between
July 2004 and April 2006, and included two summer and two winter seasons. A
pattern of extensometers at six different depths and two surface monuments have been
installed at each of the six sites to measure the range of ground movements occurring
between summer and winter periods within the Auckland region.

The laboratory testing reported herein was undertaken by the Geomechanics

Laboratory of the University of Auckland School of Engineering and through a
private provider (Geotechnics Laboratory Ltd). Climatic data was provided by the
Climate Research and Information Services, National Institute of Water &
Atmospheric Research (NIWA).
As an additional part of the investigation into the expansive characteristics of soils in
the Auckland region, a review was undertaken of the geotechnical reports held by the
local territorial authorities (TAs) with the largest areas of recent and immediate future
development within the Auckland region, these being the Manukau, Waitakere and
North Shore City Councils.
The information within the various geotechnical investigation and completion reports
held by these councils has been reviewed and copies have been obtained of all the
shrink-swell laboratory test information contained within those files. An analysis of
the individual swell-strain and shrink-strain components making up the shrink-swell
index has been undertaken to determine the relative influence that each of the strain
components has on the overall index, and to determine if the relative influence
changes significantly with soil type, time of season or sample depth.
2.0

RESEARCH APPROACH

2.1

Introduction
For the results of a research project of this type to be meaningful the following criteria
should be met:

2.2

(a)

That there be sufficient test results on which to determine geotechnical

properties.

(b)

That the conclusions arising from the research are able to be supported
through the correlation of theoretical analyses with physical measurements
and observations of building performance.

Staged investigation
The research project was conceived in two parts:
(a)

Stage I involving the measurement and determination of geotechnical field

and laboratory parameters.

(b)

Stage II involving the correlation of the predicted soil shrink-swell

movements with the measured ground surface movements and building
performance observations.

Stage I, which is the subject of the 2003 Study Report, was jointly funded by the
Building Research Levy and Manukau City Council.
Funding was obtained for Stage II from the Building Research Levy, Manukau City
Council, Auckland City Council, Rodney District Council, North Shore City Council
and Franklin District Council. The level of funding that was put in place was
sufficient to provide for the installation of extensometers and further laboratory
2

testing over a two-year period, but was not sufficient to extend the study to the
observation and evaluation of building foundation performance.
Stage II has therefore been separated into two components, Stage IIA and Stage IIB,
corresponding to the extensometer and laboratory investigation reported herein and to
the observation of building performance respectively. The interaction of the
investigations and their related stages are shown in Figure 1.

Figure 1

2.3

Staged investigation showing overlap between Stage I, IIA and IIB areas of the
expansive soils research.

Overall investigation programme

2.3.1

Introduction

The Stage I investigation comprised eight sites selected from representative soil types
within the Auckland region. It had been envisaged that the Stage IIA investigation
would follow on from the completion of Stage I in October 2003 to provide
continuous record of soil suction data. However, funding was not in place sufficiently
early to enable the extensometers to be set up in time to monitor the 2003/04 summer
period.
Stage IIA was commenced in April 2004 and ran until April 2006. The Stage IIA sites
were selected to enable Stage IIB to be carried out should funding for Stage IIB be
made available at a future time. The extensometer installation for Stage IIA therefore
involved a review of the performance of the Stage I sites to determine if they would
be appropriate to be used in Stages IIA and IIB. The level of funding received from
the various local TAs was also a factor in selecting sites for Stage IIA.
2.3.2

Monthly testing

It was proposed that monthly soil suction testing should be carried out at all six sites
selected for the Stage IIA investigation, to provide an indication of the trend of soil
suction change over an 18 month period from winter 2004 to the end of summer 2006.
The monthly testing aimed to provide a guide for extrapolation of the measured
results to provide data for the climatic extremes and correlation with the extensometer
readings. Based on the level of funding for the project it was considered that eight
discrete months could be tested, comprising two winter months (one in 2004 and
one in 2005) and three months within each of the 2004/05 and 2005/06 summers.
The sampling times were selected on the basis of monitoring the soil moisture deficit
3

(SMD) data for the Auckland International Airport on the National Climate Centre
Climate Now website operated by NIWA in order to specifically target the driest
periods during the summer months.
2.3.3

Seasonal testing

Soil classification tests, comprising Atterberg Limits and linear shrinkage tests, were
obtained during the winter 2005 sampling round and submitted for testing. Core
shrinkage (Ics) tests were carried out on samples obtained during the 2005/06 summer.
The foregoing tests enable comparisons to be made for samples collected at the same
time between the core shrinkage (Ics) test (used in the Stage I report) and the shrinkswell (Iss) test (the test most regularly carried out by consulting engineers in the
Auckland region), in order to establish the validity of the assumptions adopted in the
projections made within the Stage I report.
2.4

Extensometers
Extensometers were installed up to depths of approximately 4 m at the six Stage IIA
sites and used to determine the depth at which no seasonal soil movement occurred,
with the maximum depth confidently assumed to be a zero for measurements. The
soil movements were generally measured at the same time as the laboratory test
samples were obtained.
The extensometer data enabled a correlation to be obtained between the calculated
soil movement from the soil suction readings and the actual movements measured on
site for the soil moisture conditions existing at the time of measurement, and provides
greater accuracy in the projection of the depth and amount of soil movement that may
occur in drought conditions within the Auckland region.
A set of six extensometers were installed at each site, comprising steel rods cemented
at depths of approximately 0.5 m, 1 m, 1.5 m, 2 m, 2.5 m and 4 m below the existing
ground surface at the sites. The extensometers were buried below the ground surface
to avoid accidental damage or vandalism to the equipment. Two surface monuments
were also installed at each site at the top of the soil profile, one set below the topsoil
layer and one set through the topsoil layer to the ground surface.

2.5

Building damage survey

A building damage survey was proposed to assess the performance, over time, of
existing buildings built in accordance with NZS 3604. This survey was aimed at
assessing the condition of five buildings in the vicinity of each test site based on:
(a)

A review of building records and drawings held by the property owner or local
TA.

(b)

An assessment of damage to the building structure and fabric.

(c)

A survey to assess uniformity of level of concrete floors.

Although funding has not been made available for the building damage survey to be
carried out, part of the Stage IIA investigation reported herein was to establish test
sites in proximity to buildings that could be used within the building survey should
4

funding lines be established for that work at a future time.

The site selection process required that buildings be identified that were publicly
owned, by either the local TA or Housing New Zealand Corporation, and that the
buildings had been constructed to approximate NZS 3604:1999 standards and be of a
construction type that would show structural damage should any distortion to the
building have occurred. Concrete slab-on-ground floor with conventional shallow pad
or strip footings with full masonry or brick veneer cladding was a typical construction
considered suitable for the purposes of the study. A minimum age of approximately
10 years was adopted in order to ensure that the buildings had experienced a
reasonable history of summer and winter conditions.
The Stage II sites that comprise this part of the study have been identified as having a
minimum of five buildings that meet the foregoing criteria, with the buildings
identified being located at distances ranging from approximately 10 m to 500 m from
the test site. Details of the locations of the identified buildings are held on Fraser
Thomas Ltd file records.
3.0

LITERATURE REVIEW
A literature review was undertaken as part of the Stage I investigation and report. The
published papers and other references that were reviewed for Stage I and which may
be referred to in this report are presented in the bibliography in Appendix A of this
report. Readers are referred to the 2003 Study Report for the literature review and
associated discussion.

4.0

NEW ZEALAND AND RELATED STANDARDS

4.1

Introduction
NZS 3604:1999 Timber Framed Buildings was introduced in June 1999. From
June 1999 to May 2000 both NZS 3604:1999 and its predecessor NZS 3604:1990
Code of Practice for Light Timber Framed Buildings Not Requiring Specific Design,
were both accepted as design and construction standards i.e. there was a one-year
overlap period.
The 1990 Standard specified minimum foundation embedment depths to mitigate soil
expansivity effects. The 1999 Standard removed the specified minimum foundation
embedment depths and introduced dependency on the processes and requirements of
the Australian Standard AS 2870:1996 Residential Slabs and Footings
Construction.
The requirements and processes of the Australian Standard have evolved over time
and have included the development of a considerable information base of soil
properties and performance and methodologies that reflect this data. In contrast, a
similar database has yet to be developed for New Zealand soils.
While not referred to in the New Zealand or the Australian Standards, the American
Association of State Highway and Transportation Officials (AASHTO) has also
developed a soil expansivity test method, which is discussed further in later sections.

4.2

NZS 3604:1990 Code of Practice for Light Timber Framed Buildings Not
Requiring Specific Design
NZS 3604:1990 was a prescriptive Standard for light timber framed buildings and was
used by designers and builders for the types of construction defined within the
Standard where generic solutions could be applied. The Standard provided for specific
design beyond the limitations of the generic solutions.
NZS 3604:1990 referred to buildings on expansive soils sites. Section 3.2.2 Expansive
Clay provided for the following criteria for assessment:
3.2.2.1
For the purpose of 3.3.2(b) expansive clay shall be assumed to be present in the soil supporting the
foundations unless:
(a) Reasonable enquiry does not reveal any incidence of major cracks in dry weather on the
building site itself or in the surrounding locality;
(b) The locality has not been identified as an area where expansive clay is likely to be found;
(c) Excavation for foundations does not reveal plastic clay.

Section 3.3.2 of the Standard then required that foundations in expansive clays be
founded at a minimum depth of 450 mm below the cleared ground level and all other
foundations (not into rock) be founded at a minimum depth of 300 mm. It included a
comment: The cleared ground level is used as the depth datum because this level is
not usually altered by future landscaping, thus retaining the lateral support of the
building.
In July 1992, Amendment 1 was issued which, among other things, removed
Sections 3.2.2 and 3.2.3 including the definition of expansive and plastic clays
from NZS 3604:1990. It also revised Section 3.1.1 so that the foundation provisions
of the Standard only applied to foundations supported on good ground which, with
respect to expansive soils, excluded:
(b)

Expansive soils being those that have a liquid limit of more than 50% when tested in
accordance with NZS 4402 Test 2.2, and a linear shrinkage of more than 15% when tested
in accordance with NZS 4402 Test 2.6.

However, Section 3.3.2 of the Standard was retained, which provided for a minimum
founding depth of 450 mm in expansive clay.
Notwithstanding the 1992 amendment, it appears that geotechnical practitioners
generally continued to rely on the original provisions of Sections 3.2.2 and 3.2.3 of
NZS 3604:1990 until the introduction of NZS 3604:1999 some eight years later.
4.3

NZS 3604:1999 Timber Framed Buildings

The 1999 revision of NZS 3604 introduced a provision in Section 1.1.2 that buildings
designed to the Standard were required to be founded on good ground, which is
defined in Section 1.3 as:
Any soil or rock capable of permanently withstanding an ultimate bearing capacity of 300 kPa (i.e.
an allowable bearing of 100 kPa using a safety factor of 3.0), but excludes:

(a)

Potentially compressible ground such as top soil, soft soils such as clay which can be
moulded easily in the fingers, and uncompacted loose gravel which contains obvious
voids;

(b)

Expansive soils being those that have a liquid limit of more than 50% when tested in
accordance with NZS 4402 Test 2.2, and a linear shrinkage of more than 15% when tested
in accordance with NZS 4402 Test 2.6; and

(c)

Any ground which could foreseeably experience movement of 25 mm or greater for any
reason including one or a combination of: land instability, ground creep, subsidence,
seasonal swelling and shrinking, frost heave, changing ground water level, erosion,
dissolution of soil in water, and effects of tree roots.

In circumstances where expansive soils are encountered the designer is referred to

Section 17 Expansive Soils, which in turn refers to AS 2870 for classification of the
soil into expansivity Classes S, M, H or E and for the methods to be used in the design
of the footings.
4.4

AS 2870:1996 Residential Slabs and Footings Construction

The preface to AS 2870 states:
... the purpose of this Standard is to establish performance requirements and specific designs for
footing systems for foundation conditions commonly found in Australia and to provide guidance
on the design of footing systems by engineering principles.

AS 2870 leads the designer through a process of site classification, standard designs,
design by engineering principles, detailing and construction requirements. These are
discussed in subsequent sections of this report.
It is relevant to note that AS 2870 does not contain prescriptive references to soil
parameters, such as Atterberg Limits or linear shrinkage, as a means of determining
whether a soil is expansive or has a particular degree of expansivity.
4.5

AASHTO Designation: T 258-81 Standard Method of Test for Determining

Expansive Soils
AASHTO prescribe a method to detect whether a soil is expansive and to predict the
amount of swell. This is done by relating the Atterberg Limits of the soil to the natural
soil suction at the time of construction, as shown in Table 1.
Table 1

AASHTO guidelines on assessing expansive soils (based on AASHTO T 258-81

Table 1)

Degree of
expansivity

Liquid limit
%

Plasticity index
%

Low
Marginal
High

<50
50-60
>60

<25
25-35
>35

Soil suction nat (tsf)

kPa
pF (1)
<144
<3.17
144-383
3.17-3.59
>383
>3.59

Note 1. The authors have included the translation of soil suction units from kPa to pF to provide data
that is comparable with the findings of this report. The conversion is based on Equation 1,
taken from Clause C2.2.3(a) of the Commentary to AS 2870 viz:

u(pF) = 1.01 + log10 [u (kPa)]

Equation 1

5.0

SITE CLASSIFICATION UNDER AS 2870

5.1

Introduction
AS 2870:1996 provides for the classification of sites in terms of soil expansivity
based on:
(a)

Visual inspection of the soil profile and the use of existing knowledge of the
performance of existing residential footing systems within the surrounding
region which are not less than 10 years old on similar soil profiles; or

(b)

Estimation of the characteristic surface movement (ys). The dimension ys

relates to the ground surface movement that occurs as the moisture condition
of the soil profile changes from wet to dry design conditions. The estimation
of ys requires knowledge of the design soil moisture conditions and the soil
shrinkage index.

Very little historic data is held for Auckland soils in terms of their expansivity. The
application of the procedures of AS 2870 to Auckland conditions therefore requires
that the designer estimate the characteristic surface movement in order to classify the
expansivity of a site for foundation design purposes, as shown in Table 2.
Table 2

Classification by characteristic surface movement (from Table 2.3 of AS

2870:1996)
Characteristic surface movement
0 mm <ys 20 mm
20 mm <ys 40 mm
40 mm <ys 70 mm
70 mm <ys

Classification of site
S Slightly reactive
M Moderately reactive
H Highly reactive
E Extremely reactive

The design parameters that are required for determination of the characteristic surface
movement (ys) are addressed in Sections 5.2 to 5.4.
Note that a classification above the lower end of the moderately reactive range will
fall outside the definition of good ground from Section 1.3 of NZS 3604:1999 as set
out in Section 4.3, which has a limiting maximum movement of 25 mm.
5.2

Calculation of characteristic surface movement

The commentary to AS 2870 defines the characteristic surface movement (ys) as the
characteristic value that has a five percent chance of being exceeded in the life of the
house which may be taken as 50 years. The parameter ys is calculated using Equation
2, with u and Hs being determined from the soil suction change profiles shown on
Table 2.4 and Figure 2.1 of AS 2870. Table 2.4 and Figure 2.1 are presented as Table
3 and Figure 2 of this report respectively.
1
ys =
100

Hs

Ipt udh
0

where
ys
Ipt
8

= design characteristic surface movement (mm)

= instability index (%)

Equation 2

u
h
Hs

= suction change at depth (h) from the surface (pF)

= thickness of soil layer under consideration (mm)
= depth below which no moisture or soil suction change occurs (mm)

A worked example of Equation 2 is provided in the Australian Handbook HB 28:1997

The Design of Residential Slabs and Footings.
In AS 2870, the design soil moisture conditions are expressed in terms of soil suction,
u, which has units of pF. When a soil is saturated, it has a relatively low suction value
of 3.2 pF or less, which increases to approximately 4.2 pF when the soil dries to the
wilting point of vegetation. Sunflowers in pots are commonly used as the test plant, as
discussed in Section 2.16.4 of HB 28:1997, with the permanent wilting point being
defined as the moisture in a soil when plants in pots start to wilt and not recover.
Recommended values of soil suction change at the soil surface, us, and depth of
design soil suction change, Hs, for various locations in Australia are given in Table 3.
Table 3

Recommended soil suction change profiles for various locations (from Table 2.4
of AS 2870)
Change in suction at (1)
the soil surface (us) pF
1.2
1.2
1.2
1.5
1.5
1.2
1.2
1.5
1.2
1.5
1.2

Location
Adelaide
Albury/Wodonga
Brisbane/Ipswich
Hobart
Hunter Valley
Launceston
Melbourne
Newcastle/Gosford
Perth
Sydney
Toowomba

Depth of design
suction change (Hs) m
4.0
3.0
1.5 to 2.3
2.0
2.0
2.0
1.5 to 2.3
1.5
3.0
1.5
1.8 to 2.3

Note 1. The symbol us has been adopted to denote the u value at the soil surface.
u

Soil surface
level

Hs

Hs

Hs

(a) Consistent soil case

Figure 2

5.3

Watertable
Bedrock level

(b) Effect of bedrock

(c) Effect of ground water

Effect of bedrock or watertable on design soils suction change profiles

(from AS 2870:1996 Figure 2.1)

Soil suction change profile

Soil suction is made up of two components:
(a)

Matrix suction, due to the capillary action between soil particles.

9

(b)

Osmotic suction, due to the water-attracting action of salts.

The method for determining soil suction referenced in AS 2870 is detailed in AS 1289
Test Method 2.2.1:1998 Soil Moisture Content Test - Determination of the Total
Suction of a Soil Standard Method. This method involves the laboratory
determination of the relative humidity of a small air space in equilibrium with a sealed
soil sample, by measurement of the dewpoint temperature of a thermocouple. This
method was followed for the study reported herein except that the specified Wescor
HR33T microvoltmeter has been substituted with a Soil Mechanics Instrumentation
(SMI) transistor psychrometer which, in Australia, is considered a suitable equivalent.
In principle, the SMI transistor psychrometer is an electronic wet and dry bulb
thermometer, in which a wet and dry transistor probe is used instead of wet and
dry thermometer bulbs to measure the relative humidity of the air space in
equilibrium with a soil sample. The temperature depression of the wet transistor,
which holds a standard sized water drop, is measured and amplified within the probe.
The relationship between relative humidity and soil suction is used to determine the
soil suction. The transistor psychrometer improves on the thermistor or thermocouple
psychrometer and other forms of suction measuring equipment in that it has a larger
range (3 pF to 5 pF or 100 kPa to 10000 kPa), faster response and is compatible with
modem data logging facilities.
Soil suction readings from soil samples generally fall between 3.2 pF (wet) and 5 pF
(dry), with higher readings in the order of 6.5 to 6.9 pF applying to oven-dried soil
samples.
Table 2.4 of AS 2870, presented as Table 3 of this report, provides recommended soil
suction change profiles for various Australian locations in terms of values of change
in soil suction (u) at the ground surface and depth of design suction change (Hs).
Figure 2.1 of AS 2870, presented as Figure 2, shows triangular design suction change
profiles in terms of u and Hs. The triangular profiles are based on the assumption
that u decreases linearly with increasing depth below the ground surface, becoming
zero at a depth of Hs.
In the case of bedrock being encountered within the depth of Hs, the design profile is
truncated to become trapezoidal.
Note 3 of Clause 2.2.3 of AS 2870 states that: The designer may extrapolate to other
areas if due consideration is given to the climate and soil fabric. Alternatively,
published values of Hs, based on consideration of regional Thornthwaite moisture
indices using the general principles in Appendix D [of AS 2870] and based on at least
20 years of climate data, may be used.
Fityus et al (1998), Fox (2000), and Smith (1993) have proposed that the
Thornthwaite Moisture Index (TMI) be used to determine the depth of soil moisture
change, Hs, for the purpose of site classification in terms of AS 2870.
The TMI is an aridity climate parameter. Fityus et al (1998) analysed the
developments that the formulae for the TMI have undergone since first published by
Thornthwaite in 1948, resulting in Equation 3. In New Zealand, this equation has been
used by NIWA to calculate the TMI for Auckland Airport for a hypothetical soil
suction profile with a water storage of 100 mm, where water storage is the depth of
water available for plant use and ranges, for the purposes of the model, from field
10

capacity to permanent wilting point and is nominally in the top 1000 to 1200 mm of
soil depth.
TMI

100

P
1
PE

Equation 3

where
P
PE

= annual precipitation at a site (mm)

= net potential for evapo-transpiration at a site (mm).

The proposed correlation between TMI and Hs from Fityus et al (1998) is shown in
Table 4.
Table 4

Depth of moisture change based on TMI values

Climatic classification
Wet coastal/alpine
Wet temperate
Temperate
Dry temperate
Semi-arid

Thornthwaite Moisture Index

(TMI)
>40
10 to 40
-5 to 10
-25 to -5
<-25

Depth of moisture change

(Hs) m
1.5
1.8 to 1.5
2.3 to 1.8
3.0 to 2.3
3.0

The correlation shown in Table 4 above is the same as that proposed in Tables Dl and
D2 in Appendix D of AS 2870, except that a range of Hs values is proposed by Fityus
et al (1998) whereas specific values are proposed by AS 2870.
Based on the TMI value of +50 for Auckland Airport, i.e. the value provided by
NIWA, it is apparent that Auckland falls into the wet coastal/alpine climatic
category of Table 4 (TMI >40) for which an Hs value of 1.5 m is given.
5.4

Instability index
5.4.1

General

The instability index (Ipt) is defined in Appendix F of AS 2870:

as the percent vertical strain per unit change in suction [in terms of pF], taking into account the
expected design values of:
(a)

Applied stress;

(b)

Degree of lateral restraint; and

(c)

Suction range.

The instability index (Ipt) is derived from the shrinkage index (Ips) which may be
determined from shrink-swell, loaded shrinkage or core shrinkage tests, in accordance
with AS 1289 Test Methods 7.1.1, 7.1.2 or 7.1.3 respectively. In the case of this
study, Ips is a generic notation and has been determined using the core shrinkage
method detailed in AS 1289 Test Method 7.1.3:1998 Soil Reactivity Tests
Determination of the Shrinkage Index of a Soil Core Shrinkage Index, where the
shrinkage index is referred to as Ics.
11

The core shrinkage index method was chosen over the other two permissible options
(shrink-swell index and loaded shrinkage index), following personal
communications with Australian consultants, which indicated that the core shrinkage
index was the more commercially viable test and therefore the more likely one to be
adopted within the geotechnical testing industry of New Zealand.
However, as discussed in Cameron (1989), all methods of estimating the instability
index are known to have a degree of inaccuracy. For this reason, the commentary to
AS 2870 recommends that the calculated ys value is rounded up to the nearest 5 mm.
Grayson (2000) states that: normally, there is some clay in the topsoil on a site. The
reactivity of the topsoil is rarely tested, but is typically assumed as approximately
50% of the reactivity of the underlying clays. This assumption has been adopted in
this report.
5.4.2

Calculation of instability index

The relationship between Ipt and Ips given in AS 2870 is as follows:

Ipt = Ips

Equation 4

where
is a constraint effect coefficient and is taken as follows:
= 1.0 in the cracked zone (unrestrained), and
= 2.0 z/5 in the uncracked zone (restrained), where
z is the depth below the finished ground surface, m.
5.4.3

Determining cracked zone depth

The depth of the cracked zone refers to the depth in which predominantly vertical
shrinkage cracks exist seasonally (AS 2870: 1996 and HB 28:1997). AS 2870
provides values of cracked depths as shown in Table 5.
Table 5

Examples of cracked zone depths (from HB 28:1997 p20)

Region
Adelaide and Melbourne
Sydney and Newcastle/Gosford
Brisbane Ipswich

Depth of cracked zone

0.75 Hs
0.5 Hs
0.5 Hs

Cracked zones are incorporated into Equation 4, for Ipt, through the value, which
allows the designer to consider the cracked depth of a soil to be laterally unrestrained.
At the time of this report, insufficient data exists to allow crack depths to be
determined for Aucklands soils so as to allow the determination of parameters
corresponding to those shown in Table 5. The investigation of crack depths for soils
within the Auckland region is beyond the scope of this study, but comments are given
in the following sections on how appropriate allowances might be made.
Pender (2001) notes that: excavations in Auckland clays reveal that the upper part of
the soil profile, up to depths of a metre or so but usually less, is fissured ... One
12

possible explanation for the fissures is the cracking of the ground surface that occurs
in the summer.
For the purposes of this study, a cracked zone depth of 0.5 Hs has been adopted,
which is the same as recommended for Sydney, Newcastle and Brisbane in HB
28:1997, as shown in Table 5 of this report, and corresponds to a depth of 0.75 m if Hs
is taken as 1.5 m.
5.5

Other relevant considerations relating to AS 2870

5.5.1

General

The foregoing sections of this report address the composition and framework of the
process outlined in AS 2870 for the classification of a site.
The following sections of this report collate various other considerations that define
the applications of AS 2870 to the site classification process and comments on the
interpretation of some of the requirements, as determined from AS 2870, AS 2870
Supplement 1 and HB 28:1997.
5.5.2

Application

AS 2870 requires that all sites on which slabs and footings are to be constructed for
residential dwellings be classified in accordance with the process set out in the
Standard.
The sites are required to be classified as Class A, S, M, H, E or P.
Class A sites are defined as most sand and rock sites with little or no ground
movements from moisture change. Sites determined to be reactive are classified as
slightly (S), moderately (M), highly (H), or extremely (E) as discussed previously
(refer Table 2).
Those sites that incorporate ground conditions that cannot be classified within the
definitions for Classes A to E are classified as Class P sites. A classification of
Class P does not signify any particular severity of problem, but rather that the site is
disqualified from the criteria for the other classes and therefore requires special
considerations using engineering principles. Class P sites would include, for example,
soft soils, landslips, subsidence areas etc.
Filled sites may be classified as any of Classes A to P.
5.5.3

Foundation performance considerations

The underlying philosophy of AS 2870 is that footings designed and constructed on a

normal site in accordance with the requirements of the Standard are expected to
have a low risk of damage. A normal site is one which is:
(a)

Not subject to abnormal moisture conditions; and

(b)

Maintained such that the original site classification remains valid and abnormal moisture
conditions do not develop.

A normal site is further described as one where foundation moisture variations are
13

caused by seasonal and climatic changes, effect of the building and subdivision, and
normal garden conditions without abnormal moisture conditions..
On sites where abnormal moisture conditions apply, footings are expected to have a
higher probability of damage. Examples of abnormal moisture conditions are given
as:
(a)

Recent removal of an existing building or structure likely to have significantly modified

the soil moisture conditions under the proposed plan of the building;

(b)

Unusual moisture conditions caused by drains, channels, ponds, dams or tanks which are
to be maintained or removed from the site;

(c)

Recent removal of large trees prior to construction;

(d)

Growth of trees too close to a footing;

(e)

Excessive or irregular watering of gardens adjacent to the house; lack of maintenance of

site drainage;

(f)

Failure to repair plumbing leaks.

Guidance and advice is given in Appendix B of AS 2870 for the requirements for the
maintenance of a normal site and which relates to:
(a)

Drainage or wetting of the site;

(b)

Positioning and operation of gardens adjacent to a house;

(c)

Restrictions on the planting of trees near the foundations of a house or a

neighbouring house; and

(d)

Repair of leaks in plumbing, stormwater and sewerage systems.

The recommendations of the Standard were developed from research and experience
in the design and performance of house footings and slabs and are reproduced in
Appendix E.
In this regard the commentary to AS 2870 notes that:
(a)

The current costs of failure are modest compared with the cost of conservative design ...
Expectations of performance of footing systems on reactive sites depends upon the
adopted standard of post-construction maintenance.

(b)

To avoid extreme moisture conditions it is essential that owners become aware of their
responsibility to care for and adequately maintain a reactive clay site.

Given the foregoing, for New Zealand use we propose a modification for the
Acceptable Solution B2/AS1 which defines the maintenance required for
foundations on expansive soils as follows:
Normal maintenance (Building Code B2/AS1 format)
Normal maintenance is that work generally recognised as necessary to achieve the
expected performance of the foundation located on expansive soils over time.
14

Unless otherwise specified by the designer, and noted on the drawings, basic normal
maintenance tasks shall ensure that:
(a)

The drainage and wetting of the site is controlled so that extremes of wetting
or drying of the soils is prevented;

(b)

The positions and operation of gardens adjacent to the dwelling are controlled,
and the planting of trees near to foundations of houses is suitably restricted;

(c)

Any leaks which develop in plumbing, stormwater and sanitary sewage

systems are reported promptly;

with the level of implementation matched to the expansivity of the underlying soils
and the distortion tolerance of the cladding.
5.5.4

Site classification

AS 2870 requires that natural sites be classified as to the expected extent of soil
movement and the depth to which the movement extends. It defines a natural site as
a site which has not been subject to cut or fill.
For other than sites classified as Class P sites, AS 2870 requires that site
classification:
shall include one or more of the following methods:
(a)

(b)

Identification of the soil profile and either:

(i)

Established data on the performance of houses on the soil profile; or

(ii)

Interpretation of the current performance of existing buildings on the soil profile.

Estimation of the characteristic ground surface movement (ys).

Notwithstanding the foregoing, AS 2870 describes the properties of the foundation by

one parameter, ys, also described as the expected free surface movement. This is the
vertical movement range expected during the life of a house from a reasonable
estimate of dry conditions to a similar estimate of wet conditions and does not take
into account the moderating effect of the footing system.
The effects of trees, poor site drainage, leaking plumbing and exceptional moisture
induced movements are not taken into account in the calculation of ys.
As discussed in Section 5.5.2 abnormal site environment factors lead to a
classification of Class P. For a reactive clay site the classification is S, M, H or E,
based on comparison of numerical values calculated for ys. The commentary to
AS 2870 advises that the accuracy of these numerical values: should not be over
emphasised. Of equal importance, although less definite, is classification by existing
house performance or by soil profile identification.
The commentary also notes that to identify accurately the reactivity of a clay site by
means of tests on samples through the soil profile is too complex and expensive to be
used routinely on individual house sites.

15

The observation is made that:

Overall estimates of the range of potential for movements in a whole area based on many tests are
a more reliable guide to design at a site than limited testing on the individual site.

It is in this context, and acknowledging the basis of development of AS 2870

recommendations from research and experience in the design and performance of
house footings and slabs, that Appendix D of the Standard provides a ready guide to
the expected level of site classification for the principal areas of Melbourne and
environs, Victoria, Sydney and Adelaide.
The data to provide such a generic appreciation for the principal areas of New
Zealand, and more specifically Auckland, either does not exist or has not been
collected. The rectification of that situation is in part the purpose of the current study.
5.5.5

Classification of filled sites

A filled site is required by AS 2870 to be classified as Class P, except where the

provisions of the Standard allow another classification. Differentiation is made
between whether the fill is controlled i.e. engineered fill or uncontrolled i.e. nonengineered fill as follows:
(a)

Controlled fill (engineered fill)

(i)

(ii)

(b)

(a)

up to 0.8 m for sand

(b)

up to 0.4 m for other materials.

Deep fill
(a)

>0.8 m depth of sand may lead to a less severe reactive

classification

(b)

>0.4 m depth of other material requires the site to be considered

as Class P (i.e. the classification is subject to specific
engineering considerations).

Uncontrolled fill (non-engineered fill)

(i)

(ii)

16

Shallow fill the classification is required to be the same as the natural

site prior to filling where the depth of fill is:

Shallow fill unless building foundations are founded on natural soil

under the fill the site is required to be classified as Class P for:
(a)

up to 0.8 m depth of sand

(b)

up to 0.4 m depth of other material.

Deep fill for fill depths greater than 0.8 m for sand and 0.4 m for
other materials, the site is required to be classified as Class P.

5.5.6

Reclassification of filled sites

Subject to the following proviso, AS 2870 provides that Class P controlled fill sites
may be reclassified in accordance with engineering principles including consideration
of:
(a)

Expected long-term movement in the fill and the underlying soils; and

(b)

The depth of the cracked zone.

The proviso is:

The reclassification shall not be less severe than the natural site classification unless the controlled
fill consists of non-reactive material and is deeper than one metre or O.5 Hs, whichever is greater
[Hs being the depth of the design soil suction profile].

In addition, the Standard requires that:

The depth of the cracked zone should be taken as zero for reactive clay in controlled fill placed
less than five years prior to building construction.

5.5.7
(a)

Classification parameters discussion

General
The commentary to AS 2870 makes it clear that the reactivity of clay soils
cannot be clearly evaluated by tests. Reactive clays are clay soils that shrink as
they dry and swell as they wet up. If the movement is significant such clay
soils are termed reactive.
The amount of movement depends on:
(i)

The clay minerals present;

(ii)

The proportion of clay in the soil and in the profile;

(iii)

The moisture changes and their extent;

(iv)

Loading;

(v)

Lateral restraint.

Individual tests for clay reactivity are subject to wide scatter. Thus individual
high or low results may often represent testing variations rather than real
variations in the overall properties of the site.
It follows that there is no single test that can confidently assess a particular
site.
(b)

Soil suction
Soil suction is not simple to determine. It is useful in the analyses of reactive
clays because it is more strongly a function of the climate and vegetation than
it is of soil type. The distribution of soil suction is approximated in the
17

Standard to be triangular and to generally be conservative, but it is recognised

that near surface soil suctions may be underestimated slightly.
The design profile includes the expected influence of the building and the
garden and, to some extent, droughts. The design soil suction and hence
movement are not merely cyclic seasonal values. The effects of very large
trees, poor site drainage and long-term plumbing leaks are not included.
(c)

Cracked zone
This zone refers to the depth in which predominantly vertical shrinkage cracks
exist seasonally.

(d)

Characteristic design surface movement

The characteristic design surface movement, ys, is a hypothetical parameter.
This surface movement is described in HB 28:1997 as: a relative movement
within the site between a low point during a dry time and a high point during a
wet period including the effects of site development on the moisture regime.
Thus the two extremes occur at different times. It is not simply the extreme
range of seasonal movement experience in the field before development.
HB 28:1997 further advises that:
It is important that too much emphasis should not be placed on this method in comparison
with other techniques, as the results are rather imprecise.

And that:
... the use of suction profile and instability index values is the most accurate method of
calculation available (but not necessarily the most accurate method of classification).

As also noted in the foregoing, it is a requirement of AS 2870 that the

calculation of ys assumes that the maintenance of the site complies with
Appendix B of the Standard.
(e)

Practical and implementation considerations

Fundamental to the application of the requirements of AS 2870 is the need to
evaluate the classification of a site with regard to the effect of the reactivity of
the site soils on a structure.
The commentary to AS 2870 states that:
It needs to be emphasised that such data [local experimental data] must be relevant to the
definition of ys, e.g. data from an open field site subjected to seasonal moisture changes
will not be applicable [without consideration of the effects of site development].

AS 2870 requires the classification of a site to take into account the effect of
site works when these are known at the time of classification. When the effect
of site works is not taken into account the Standard requires that the
classification be reconsidered if:

18

(a)
(b)

the depth of cut on an S, M, H or E site exceeds 0.5 m; or

the depth of fill exceeds the limits [described in Section 5.5.5].

AS 2870 further requires that the soil type and site conditions at a building site
be inspected at footing excavation stage by the classifier to confirm the soil
profile.
Examples of the effect of cut or fill on the classification of a site, arising from
either sub-divisional or site development earthworks, include:
(i)

Increase in reactive movement by removal of part or all of a protective

non-reactive soil layer.

(ii)

Reactive movements worsened by the addition of clay fill.

(iii)

Reduction in reactive movement by the addition of an upper profile of

sand or non-reactive silt.

As noted in HB 28:1997:
It is difficult to see how a classifier can accurately assess the implication of future fill
except by warnings in the fine print attached to the classification that reconsideration of
the classification is needed if the fill is not shallow.

6.0

CLIMATE

6.1

Introduction
The expansivity of a soil is determined by the soil mineralogy and its response to the
change in soil moisture levels, which are a consequence of climatic changes. Soils
that experience little change in soil moisture are, in general, those that experience
little seasonal climate change and those with a low shrinkage index.

6.2

Auckland climate information

For the purposes of this study, the National Climate Centre for Monitoring and
Prediction (a division of NIWA) was contracted to provide daily rainfall, temperature
and TMIs for the Auckland Airport meteorological station for the 34-year period
between 1 January 1972 and 30 June 2006.
6.2.1

Thornthwaite Moisture Indices (TMIs)

Internationally TMIs are generally used for the following:

(a)

In the United States the TMI is correlated to design edge distance which
determines the width of a floor slab that is subject to surface movement when
using the Post Tension Institute (PTI) method of ground slab design.

(b)

In Australia, the TMI is used to determine the depth, Hs, below which soils do
not experience volume changes, which is then used to determine the site
expansivity classification in accordance with AS 2870.

As noted in Section 4.5, the American method also uses the Atterberg Limits to define
the clay mineralogy. However HB 28:1997 states that the American method has poor
19

correlation with movements measured in Australia.

In this investigation, the TMIs provided by NIWA have been used to compare the
Auckland climate with that at Australian locations for which the climate-related soil
expansivity factors are known. It has been found that the TM1 varies greatly between
the various regions of Australia.
The correlation between climatic zone and TMI, as proposed in Appendix D of
AS 2870, is reproduced in Table 4.
As noted in Section 5.3, NIWA calculated an annual TMI (over 34 years of record)
for Auckland Airport of approximately +50, corresponding to the wet coastal/alpine
classification in Table 4.
As shown in Table 6, Auckland Airport has an annual average rainfall of 1,240 mm.
The average annual rainfall for the eight Stage I test sites (Sites A to H) falls within
the range of 1200 mm to 1400 mm, except for Site E which falls within the range of
1400 mm to 1600 mm.
It is therefore apparent that Auckland Airport rainfall corresponds to the lower end of
the range of annual rainfall for the Auckland region. The TMI for Auckland Airport is
therefore likely to be conservatively low with respect to the eight test sites.
6.2.2

Water balance

NIWA has identified that daily TMI values for the Auckland region are likely to have
a high degree of variability and has recommended that a better indication of daily soil
moisture is the running water balance. The water balance is represented by
precipitation less evapo-transpiration less deep percolation which, along with other
meteorological data, yields a SMD for the particular location.
The SMD data (shown in Figure 3) has been used to:
(a)

Estimate the time of year that testing should take place to obtain the soil
suction profiles corresponding to the wettest and driest periods during the
2002/06 years; and

(b)

Relate the soil moisture conditions at the time and location of sampling to
extreme wet winter and drought conditions to allow the measured soil
suction values to be extrapolated to provide estimated values corresponding to
wet winter and drought conditions at the ground surface, as required by
AS 2870. The winter 2002 samples corresponded to saturated conditions, and
have therefore been assumed to be representative of wet winter conditions.

AS 2870 refers to design wet and dry conditions. For the purposes of this report,
design dry conditions have been assumed to be drought conditions.
The water balance is expressed as a soil moisture surplus or deficit. A zero soil
water balance indicates that the soil is saturated. Positive values indicate runoff (i.e. a
surplus) and negative values indicate a deficit.
NIWA advised that SMDs (i.e. negative values) greater than 90 mm indicate drought
conditions and deficits less than about 10 mm are likely to be saturated with some
20

runoff and drainage occurring.

100
80

Daily water balance (mm)

(Surplus)
(Deficit)

60
40
20
0
-20
-40
-60
-80
-100
Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Time period

Average

Maximum

Minimum

2002

2003

Figure 3 Daily water balance for Auckland Airport during 19712003

6.3

Climate factors for Australia and Auckland

As noted previously, the expansivity of soils is dependent on soil mineralogy and
climate. Relevant climate factors for Auckland and selected Australian cities are
summarised in Table 6. Average annual rainfall and temperature data for Auckland
and four selected Australian centres are shown on Figure 4.
Table 6

Climate factors (sourced from www.worldclimate.com, Fityus (1998), HB

28:1997 and Appendix D of AS 2870)

City

Auckland
Adelaide
Brisbane
Hobart
Launceston
Melbourne
Cape Otway, Victoria
Newcastle
Sydney

Rainfall
(mm)
1240
516
1150
598
694
656
892
1143
1222

Average Annual Readings

Minimum
Maximum
temperature
temperature
(C)
(C)
11.3
18.9
11.0
22.3
15.6
25.4
7.8
17.2
6.1
16.8
9.3
19.4
10.4
17.2
14.1
21.7
12.9
21.0

Thornthwaite
Moisture
Index
+50
-40
+20
+10
+80
+10
+40
+30 to +40
+20

21

140

80

120
Average monthly rainfall (mm)

70
100

60
80

60

50

40
40
20

30

0
J

-20
20

-40
10
-60

-80
Auckland (Rainfall)
Melbourne (Rainfall)
Sydney (Temp)

Figure 4

6.4

Average monthly temperature (deg C)

Month

0
Newcastle (Rainfall)
Adelaide (Rainfall)
Hobart (Temp)

Sydney (Rainfall)
Auckland (Temp)
Melbourne (Temp)

Hobart (Rainfall)
Newcastle (Temp)
Adelaide (Temp)

Average monthly rainfall and temperature measurements for Auckland Airport

and selected Australian cities (sourced from www.worldclimate.com)

Comparison of Australian and Auckland climates

From the climate data summarised in Table 6 and Figure 4 it is apparent that:

22

(a)

The TMI for Auckland (+50) is similar to that of Newcastle (+30 to +40) and
the Cape Otway area in eastern Victoria (+40), indicating that the respective
climates are similar.

(b)

Although the Newcastle average annual rainfall is similar to that of Auckland,

it is important to note that some of the higher rainfall months in Newcastle are
in summer when potential evapo-transpiration is at its highest. This would
tend to give lower annual percolation of water into the soil than in Auckland
where the higher rainfall months are in winter. Hence the TMI would be
expected to be higher in Auckland. Thus the TMI value of around +50 for
Auckland, when compared to the TMI value of +30 to +40 for Newcastle,
would seem to be reasonable.

(c)

The TMI for Auckland (+50) is significantly higher than that for Sydney and
Brisbane (+20), and Melbourne and Hobart (+10), indicating that the
Auckland climate is significantly wetter than these Australian centres.

(d)

Although the Auckland annual rainfall is similar to that of Sydney, Brisbane

and Newcastle, the minimum and maximum temperatures in Auckland are

lower. Auckland would therefore be expected to have a lower evapotranspiration rate and consequently a higher TMI value than these Australian
cities.
(e)

The Auckland annual rainfall is approximately double that for Adelaide,

Melbourne, Hobart and Launceston.

(f)

Whilst the average annual rainfall for Launceston is approximately half that of
Auckland, the average annual minimum and maximum temperatures are
significantly lower than Auckland. The TMI for Launceston (+80) is
significantly higher than that of Auckland (+50). This data suggests that the
Launceston climate is wetter than Auckland.

In summary, therefore, the TMI values indicate that within Australia the climates in
Newcastle and eastern Victoria are the closest comparisons to that of Auckland,
although the Auckland climate is slightly wetter than these two areas.
7.0

MINERALOGY OF SOILS

7.1

General
Soils swell on wetting and shrink on drying, resulting in ground movement. If the
ground movement is sufficiently large to affect any structures the soil is said to be
reactive.
One of the factors governing the reaction of a soil to moisture change is the
mineralogy of the individual soil types. Some clays, such as smectite, are extremely
reactive to moisture change while clays with a high kaolinite content are known to be
only slightly reactive. AS 2870 and HB 28:1997 both discuss the range of soils found
within the main geographical regions of Australia and then incorporate the effects of
regional climatic changes to provide the designer with foundation solutions.
This section provides a brief overview of the soils found within both Australia and
Auckland, with the aim of ascertaining whether there are direct comparisons between
the design factors specified within AS 2870 for the varying Australian soil types and
those soils found in Auckland.

7.2

Australian soils
7.2.1

Introduction

The following information has been taken from Australian Handbook HB 28:1997,
which provides a brief summary of the main soil types within each geographical area.
7.2.2

Sydney clays

Sydney clays have been found, in general, to derive from sandstones or shales,
although there are a few well-defined areas that are founded on alluvial clays which
form deeper deposits than the rest of Sydney and which are highly reactive. As shown
in Table 7 the reactivity of all clays, other than alluvial, tends to be related to depth as
opposed to mineralogy.

23

Table 7

Classification of Sydney clays (from HB 28:1997, p29)

Type of clay
Alluvial
Other
Other
Other

Depth of clay
All depths
<0.6 m
0.6 2.5 m
>2.5 m

Expansivity class
H Highly
S Slightly
M Moderately
H Highly

The data in Table 7 indicates that any sites in Sydney with clay depths greater than
2.5 m are classed as highly expansive (Class H).
7.2.3

Newcastle/Hunter Valley clays

A significant amount of research into soil expansivity in the Newcastle region has
recently been undertaken and this is ongoing. There is still insufficient information to
provide blanket recommendations of variables to be used in AS 2870 and testing is
still recommended in most parts of Newcastle.
The Newcastle area has a more variable geology than that of other Australian cities
(e.g. Sydney). The sedimentary rocks include mudstone, shale, sandstone and
conglomerate as well as coal seams. Some of the sedimentary rocks contain thin
layers of volcanic ash. The volcanic ash has been said to contain up to 10% smectite,
which can have a marked effect on the reactivity of the soil. Most clay sites derive
from sandstones and conglomerates, producing mainly Class M sites.
7.2.4

Melbourne clays

Melbourne is generally founded on residual soils weathered in place from the

underlying rock. Although most residual soils are classed M, it has been observed that
the clays derived from basic igneous lava flows (i.e. basalts) are more reactive and are
generally classed as H. Further investigation is required for possibly highly reactive
limestone and alluvial clays. The climate of Melbourne varies, with the west being
significantly drier and the east being more moderate (wet temperate). Testing has
shown that the clay sites are generally less reactive in the wetter areas of Melbourne.
7.2.5

Brisbane clays

The founding soils within the Brisbane region vary considerably between sites due to
topography and relatively complex geology. The following soils are found within the
Brisbane area:
(a)

Residual and alluvial soils weathered from basalt are considered highly or
extremely reactive.

(b)

Rhyolitic tuff can be highly reactive but generally not as much as (a).

(c)

Black or brown clays found around Ipswich are of high/extreme reactivity.

(d)

Clays derived from volcanic ash are extremely reactive.

The climate varies greatly in the east-west direction and classification can range from
Class A to H and sometimes E. Due to the variability of the clays, tests are still
generally carried out to confirm which expansivity classification applies to particular
sites.
24

7.2.6

Adelaide clays

Adelaide is generally founded on sediments which vary from red-brown earths of

moderate to high reactivity to highly reactive black earths and fissured Pleistocene
clays (50% illite, >20% kaolinite and <20% smectite), which also display evidence of
shrinkage cracking at the surface of the more reactive soils during summer and
autumn.
Ground water is generally deep. Leaching of lime layers occurs, which assists in
reducing the reactivity of the soil. The Pleistocene clays are considered to be less
reactive than the smectite-rich basaltic clays in Melbourne.
In general, the classification of soils within the Adelaide region is based on the
opinions of experienced soil classifiers who log bores for each site. The red-brown
earth group has been assigned nine typical profiles which correspond to either an M,
H or E classification, while the: sites underlain by Pleistocene clays or black earth
are generally E [classification] for clay layers greater than two metres [depth].
7.2.7

Generalisation of Australian soils

The soil types and their expansivity as detailed in Sections 7.2.2 to 7.2.6 are
summarised in Table 8.
Table 8

Generalisation of Australian soil classifications

Location
Sydney
Sydney
Melbourne
Newcastle
Adelaide
Melbourne
Sydney
Sydney
Brisbane
Newcastle
Adelaide
Brisbane
Brisbane/Ipswich
Brisbane
Adelaide
Melbourne

7.3

Clay type
Non-alluvial
Non-alluvial
Non basaltic residual
Sandstone/conglomerate
derived
Red/brown
Basaltic
Non-alluvial
Alluvial
Rhyolitic tuff
Volcanic ash derived
Red/brown
Residual/alluvial
Black/brown
Volcanic ash derived
Pleistocene/black earth
Limestone/alluvials

Depth of clay
<0.6 m
0.6 2.5 m
All depths
All depths

Expansivity Class
S Slight
M Moderate
M Moderate
M Moderate

All depths
All depths
>2.5 m
All depths
All depths
All depths
All depths
All depths
All depths
All depths
>2.0 m
All depths

M Moderate/H High
H High
H High
H High
H High
H High
H High/E Extreme
H High/E Extreme
H High/E Extreme
E Extreme
E Extreme
Further investigation necessary

Auckland soils
7.3.1

Introduction

The following information has been adopted from the handbook accompanying the
New Zealand Geological Map, Auckland Urban Area, Sheet R 11, scale 1:50000.
7.3.2

Waipapa Group

25

The oldest known rocks in the Auckland region are indurated marine sedimentary
strata constituting the greywacke basement of Late Triassic to Late Jurassic age.
The Waipapa Group forms the rolling to steep hills in the Whitford and Brookby
districts, in the Hunua Ranges and on Waiheke Island, and comprises indurated
sandstone and mudstone.
The Waipapa Group commonly comprises deep weathering profiles, with the surficial
soils comprising yellow-brown, sandy and silty clays.
7.3.3

Waitemata Group East Coast Bays formation

The Waitemata Group comprises alternating mudstone and lithic sandstone of

Miocene age and underlies most of urban Auckland. The East Coast Bays Formation
is the dominant member of the Waitemata Group within the Auckland region and
forms the conspicuous alternating beds exposed in cliffs and on intertidal platforms
around the Waitemata Harbour.
The greater part of the East Coast Bays Formation consists of graded turbidite
sandstones alternating with poorly sorted interturbidite mudstones. The residual soils
formed on this formation produce greyish white to orange-brown clays. The clay
mineralogy of the Waitemata Group residual soils, as indicated by X-ray diffraction,
comprises a mixture of kaolinite, illite and montmorillonite, with kaolinite being more
dominant at the ground surface and montmorillonite being more dominant at depth
(Harvey et al 1982).
7.3.4

Onerahi Chaos Breccia

The Onerahi Chaos Breccia forms part of the Northland Allochthon, where oceanic
crust was thrust above continental crust and tilted to allow the lower Miocene deposits
to slide and shear off, followed by sliding and shearing of the upper Cretaceous
deposits, resulting in inversion of the normal stratigraphy. The deposits occur both
above and below the Waitemata Group sandstones and siltstones of the lower
Miocene age (Beca Carter 1980).
The Onerahi Chaos Breccia comprises chaotic, irregularly-bedded rocks that are
present near the ground surface over wide areas of North Auckland. The deposit has
been associated with several large ground creep movements.
Residual soils formed on the Onerahi Formation mudstone or siltstone are very
smooth impervious clays. High montmorillonite contents are associated with areas
where ground movement has been encountered.
7.3.5

Tauranga Group

Tauranga Group sediments occur throughout the extensive lowlands mainly south and
west of Auckland City and were deposited in fluvial, lacustrine, estuarine and shallow
marine settings from the late Pliocene to late Pleistocene age.
(a)

26

Puketoka Formation (tp) this formation forms the lowlands to the west and
south of Auckland City and comprises undifferentiated, mainly pumiceous,
light-grey to orange-brown mud, sand and gravel formed in terrestrial to
estuarine environments.

The deposits typically comprise clay with occasional lenses of sand and peat.
The formation is characterised by a high variability in the nature and type of
the sediments resulting from the nature of the deposition of the formation.
(b)

Rhyolitic Pumice (tpp) this member of the Puketoka Formation comprises

rhyolitic pumice deposits derived from non-welded distal ignimbrites
originating in the Taupo volcanic zone and deposited into terrestrial, fluviatile,
or shallow marine environments.
Weathering of the rhyolitic pumice deposits results in white clay. Derived
from one or more non-welded distal ignimbrites, the deposits are often interbedded with carbonaceous deposits.

7.3.6

Auckland Volcanic Field basaltic ash

The Auckland Volcanic Field comprises basaltic deposits of the Pleistocene to

Holocene age erupted from numerous small volcanoes within a 360 km2 area centred
on One Tree Hill. The erupted material comprises basaltic lava, scoria, lithic tuff, ash
and lapilli.
The basaltic ash deposits can mantle the terrain up to several kilometres downwind
from some of the volcanoes. Owing to the distribution of the multiple volcanoes in the
Auckland region, ash deposits can be found over much of the area, particularly in the
overlapping volcanoes in Auckland City and less so in the more isolated volcanoes in
Manukau.
The basaltic ash deposits weather to form red-brown sandy clays.
7.4

Comparison of soils from Auckland and selected Australian centres

As discussed in Section 6.0, it is considered that only the Newcastle and eastern
Victoria regions of Australia have similar climatic conditions to Auckland.
Comparison of Auckland soils to Australian soils has therefore been limited to the
foregoing regions of Australia that have similar climatic conditions to those in
Auckland. It was noted in Section 7.2.4 that repeated testing has shown that the clay
sites are less reactive in the wetter areas than the drier areas of Melbourne. As
Auckland is comparatively wetter than the comparable regions in Australia, Auckland
soils may also be less reactive.
It is considered likely that the soils formed on the sedimentary coal measures
(Class M) of Newcastle would be similar to the Waitemata Group residual soils in
Auckland.
The clays formed on the sandstones and shales in Sydney (Class S to H) could be
comparable to the Waitemata Group residual soils and possibly even the Waipapa
Group residual soils in Auckland. It is, however, recognised that the clays formed on
the Sydney sandstones are of a lesser thickness and of a more uniform profile than
those formed on the Waitemata Group sandstone and mudstone. The alluvial soils in
Sydney (Class H) are likely to be similar to the Tauranga Group/Puketoka Formation
(tp) alluvial soils of Auckland.
It is possible that the clays formed on the basaltic deposits in eastern Victoria (Class S
to H) may be comparable to the basaltic ash deposits in Auckland.
27

Soils similar to the Onerahi Chaos Breccia and Tauranga Group/Rhyolitic Pumice
(tpp) are not found in the Newcastle/Sydney or eastern Victoria regions of Australia.
8.0

FIELD AND LABORATORY TESTING FOR STAGE II

8.1

Site selection
The test sites used for the Stage I investigation (refer Table 9) have in general been
carried through into Stage II with the following modifications:
(a)

The original Sites A (Newmarket), C (Brookby) and E (Swanson) have been

discontinued.

(b)

A new Site 2F, located at Princess Street Reserve, Pukekohe has been added to
gain geographic representation from the South Auckland Volcanic Group soils
located within the Papakura District Council and Franklin District Council
areas.

(c)

Site B (East Tamaki) and Site G (Hillcrest) have been replaced by sites at
Otara and Mairangi Bay, respectively, in order to provide greater site control
in the event that funding for building damage surveys becomes available.

Table 9
Stage I
site code
A
B
C
D
E
F
G
H

Stage II test locations in the Auckland region

Main soil type

Suburb

Basaltic ash
Tauranga Group (tpp)
Waipapa Group
Tauranga Group (tp)
Tauranga Group (tp)
Waitemata Group
Waitemata Group
Onerahi Chaos Breccia
South Auckland Volcanics

Newmarket
East Tamaki
Brookby
Manurewa
Swanson
Howick
Hillcrest
Red Beach
Pukekohe

Stage II
site code

2B

2A

2C
2D
2E
2F

Suburb

Otara

Manurewa

Howick
Mairangi Bay
Red Beach
Pukekohe

The actual site selections made generally reflect the need to obtain a range of soil
types and local climate conditions across the Auckland region on which to base the
research findings. The selections also reflect the relevant TAs willingness to provide
funding support for the current project work.
8.2

Extensometer field
From April to August 2004 the extensometers were installed at each of the Sites 2A to
2F inclusive. Six extensometers and two surface monuments were installed at each
site. The extensometers were installed so that they would sit below the ground surface
beneath a turf square in order to prevent damage occurring to the extensometers from
mowing or vandalism.
Boreholes to six different depths were put down at each site for the extensometers.
The extensometers comprised stainless steel rods with a welded base plate concreted
in a 0.2 m plug in the base of the borehole. A profile of the extensometer installation
is shown in Figure 5. The plug was set so that the centre of each plug would be at
0.5 m, 1 m, 1.5 m, 2 m, 2.5 m and 4 m depths. The plug comprised rapid-setting

28

cement that was tremmied to the base of the extensometer via plastic hose pipe.

Figure 5 Sketch detail of extensometer construction (4 m deep rod used for example)

A 56 mm outside diameter (OD) PVC tube was installed to prevent the borehole from
closing in on the extensometer. The PVC tube was installed approximately 100 mm
above the top of the plug so that the extensometer was free to move without having to
overcome any friction resistance between the PVC tube and the surrounding soil. A
smaller 43 mm OD inner PVC tube of approximately 200 mm length was put down to
29

obtain an overlap with the bottom of the main outer tube.

In order to mitigate against the risk of the PVC tube acting as a conduit for the ingress
of stormwater and/or seepage from the topsoil layer migrating down the tubing and
affecting the moisture conditions at the base of the extensometer, seals were installed
in the upper metre of the extensometer. In order to install the seals an 85 mm diameter
borehole was put down to approximately 1 m depth (less for the 0.5 m and 1 m
extensometers). The seal comprised compacted (tamped) clay between approximately
1 m and 0.6 m depth below the ground surface and tamped sand/bentonite mix (90:10)
between 0.6 m depth and the underside of the topsoil.
The two surface monuments constructed at each site have been designated as the
0.2 m and 0.0 m extensometers. The 0.2 m surface monument was constructed by
excavating a turf square to the base of the topsoil at site and a survey plate installed
into a shallow concrete plug. The 0.0 m extensometer comprised a hand auger to the
underside of the topsoil at the site with the borehole infilled with concrete and a
survey nail set into the concrete. Profiles of the surface monuments are shown in
Figure 6.

Figure 6 Sketch details of surface monuments

The extensometers and surface monuments were measured by level survey on each
monitoring occasion and were measured in relation to the 4 m deep extensometer,
which has been assumed to be located at a depth below which any ground movement
could occur. It is noted that as the soil materials became too hard to auger at
approximately 3.2 m below the ground surface at Site 2F, the 4 m deep baseline
extensometer at Site 2F is actually at 3.2 m depth.
In a parallel experiment the University of Auckland has installed two spider magnet
extensometers at each of Sites 2A to 2F inclusive. The design, construction,
monitoring, analysis and reporting of the data from the spider magnet extensometers
is the entitlement of the University of Auckland and does not form a part of the
Stage II research project. Any comparison between the two sets of extensometers is
also outside the scope of this report.

30

8.3

Sampling and monitoring

8.3.1

General

Sampling and monitoring at the selected sites was programmed to coincide with the
driest periods during the summers of 2004/05 and 2005/06 and typical wet periods
during the winters of 2004 and 2005, with the aim of capturing soil suction, soil
moisture and ground movement data that would be representative of wet and dry
conditions.
Typically 17 hand auger boreholes were put down at each site over the duration of the
project. The logs of the boreholes are presented in Appendix B. The borehole number
relates to each of the six sites, viz Borehole 2A relates to Site 2A, Borehole 2B to
Site 2B etc. The logs relate to the first borehole put down at each site during the
installation of the 4 m deep extensometer. The subsequent boreholes put down at each
site were located in an approximately 1 m grid from the original borehole and line of
extensometers and were generally spaced out according to the plan shown in Figure 7.
A cross-section through the extensometer field is shown in Figure 8.
While it is acknowledged that there is some variability in the shrink-swell
characteristics of the soil profile across each site, it is our opinion that the variability
is likely to be small, albeit that such variability could have affected the extensometer
data.
It was anticipated that there was only a slim chance that extreme dry or drought
period conditions would occur during the two summer seasons of the study, but that it
was likely that representative wet conditions would occur during the winter season.
In order to target the driest conditions within the 2004/05 and 2005/06 summers, the
Climate Now website operated by the National Climate Centre, NIWA, was
monitored on a regular basis. The website provides SMD values for the weather
stations within the region and is generally updated on a weekly basis. The SMD
values are a daily water balance for a theoretical 150 mm thick topsoil layer which
keeps track of the rainfall entering the pasture root zone and being lost from this zone
by evapo-transpiration or plant use. The plots are intended as a guide for agricultural
users to aid irrigation decisions.
Our monitoring of the website enabled more precise targeting of dry summer
conditions for the Stage II extensometer measurements and obtaining of soil samples
for the laboratory testing than was possible for the earlier Stage I investigation.
Following NIWA advice (refer Section 6.2), a SMD value of 90 mm for a theoretical
100 mm thick topsoil layer was taken to represent a drought condition for the Stage I
2003 report. The SMD over the duration of the project and the sampling times are
shown in Figure 9. The 2004/05 summer period provided three measurements at SMD
conditions approaching, but not reaching, the theoretical drought condition defined
in Section 6.2.

31

Figure 7 Sketch plan showing typical extensometer and borehole sampling layout

32

Figure 8 Cross-section of extensometer field

Note: Refer Figures 5 and 6 for detailed features of extensometer and surface monuments.

As discussed in Section 9.0 of this report, the extensometer readings and soil suction
values generally recorded minor variations during the 2004/05 summer period. In
order to prevent duplication of these results, the 2005/06 summer period sampling and
monitoring was scaled back in order to hold funding in reserve until more severe dry
conditions were reached.
20

May-06

Dec-05

Jul-05

Feb-05

Sep-04

Apr-04

Nov-03

Jun-03

Jan-03

Sep-02

Apr-02

20 Oct 05

10 Mar 03

5 Sep 02

-40

-60

13 Sep 04

-20

21 Mar 02

Soil Moisture Deficit (mm)

Nov-01

Figure 9

25 Mar 06

28 Feb 06

29 Apr 05

-100

30 Jan 05

Theoretical Drought Condition

17 Mar 05

-80

Soil moisture deficit variation over Stage I and Stage II monitoring periods.
Stage II sampling dates shown in green. Stage I theoretical drought condition
shown in red. Note the relatively low (dry) values obtained in the Stage II period
compared to Stage I.

33

20.0

May-06

Apr-06

Mar-06

Feb-06

Jan-06

Dec-05

Nov-05

Oct-05

Sep-05

Aug-05

Jul-05

Jun-05

May-05

Apr-05

Mar-05

Feb-05

Jan-05

Dec-04

Nov-04

Oct-04

Sep-04

Aug-04

0.0

Soil Moisture Deficit (mm)

-20.0
13.09.04

18.10.05

-40.0

-60.0

-80.0
Theoretical Drought Condition

28.04.05
28.01.05

22.03.05

25.03.06
24.02.06

-100.0

Figure 10

Soil moisture deficit variation over Stage II monitoring period. Note the
February 2006 sampling point approximates the theoretical drought condition.

An issue identified in the Stage I Study Report was that the Stage I monitoring period
was carried out over two relatively wet summer periods in 2002 and 2003, and that it
was difficult to target the driest conditions within those summer periods. Figure 9
illustrates the SMD values and sampling points over the Stage I and Stage II
monitoring periods.
As can be observed from Figure 9, the Stage II summer sampling points have
measured significantly drier SMD conditions than the Stage I summer sampling
points. Further, as can be seen in Figure 10, the 2005/06 summer sampling generated
SMD values approximating the 2004/05 summer period values i.e. similar summer
conditions were obtained over both the Stage II summers monitored.
In particular, the 24 February 2006 sampling day essentially reached the theoretical
drought condition, thereby providing an opportunity of evaluating the measured
conditions against the predictions of the Stage I report.
The -90 mm condition has been reached 12 times over the 44-year record available for
the airport weather recording station, simplistically approximating a one-in-four-year
return period drought condition.
The Stage II results, encompassing periods having significantly drier soil conditions
than the Stage I study, should be able to provide correspondingly more meaningful
data, giving greater accuracy to the analyses, interpretations, extrapolations and
conclusions.
For completeness, the Sampling and Monitoring record and Results of Field
Investigation and Testing from Sections 8 and 9 of the Stage I report are reproduced
in Appendix C.

34

8.3.2

Winter 2004

The winter 2004 sampling was carried out on 13 and 14 September 2004 after the
extensometers had been installed for a minimum period of one month prior to
sampling.
The sampling comprised:
(a)

Five undisturbed soil samples at 0.5 m depth intervals, between 0.5 m and
2.5 m depth, taken with a 45 mm diameter thin-walled stainless steel tube
driven into the base of each borehole at the required depth using a Scala
Penetrometer hammer and rods; and

(b)

Level survey measurement of extensometers and surface monuments.

The ends of the tube samples were sealed and carefully stored until the samples were
extruded and prepared for laboratory testing at the Geomechanics Laboratory at the
University of Auckland School of Engineering.
8.3.3

Summer 2004/05

Three sampling rounds were carried out over the 2004/05 summer period. The first
sampling round was carried out between 28 January and 1 February 2005 and
comprised:
(a)

One hand auger borehole at each site to provide five undisturbed soil
samples at 0.5 m depth intervals for laboratory testing as for 8.3.2(a) above;
and

(b)

Level survey measurement of extensometers and surface monuments.

After the evaluation of the extensometer and surface monument data from the above
sampling round it was apparent that no significant movement was being measured
within the extensometers, but that some movement was being measured in the surface
monuments. It was therefore inferred that taking soil suction samples to a depth of
2.5 m would measure the soil suction below the zone of seasonal changes and that
more information would be obtained by targeting the upper soil profile. Therefore, for
the balance of the Stage II study, the depth of the undisturbed sampling was reduced
to 2 m with an undisturbed soil sample obtained from immediately below the surficial
topsoil at each site i.e. at 0.2 m or 0.3 m depth below the ground surface.
The second and third summer sampling rounds were carried out between 17 and
23 March 2005 and 28 and 29 April 2005 respectively and comprised:
(c)

One hand auger borehole, to 2 m depth at each site to provide five

undisturbed soil samples at approximately 0.5 m depth intervals for
laboratory testing as for 8.3.2(a) above; and

(d)

Level survey measurement of extensometers and surface monuments.

In addition to the foregoing, samples were obtained during the second sampling round
for shrink-swell index testing, which comprised:

35

(e)

8.3.4

Taking two undisturbed soil samples at between 0.5 m and 1 m depth, with a
63 mm diameter thin-walled stainless steel tube driven into the base of each
borehole at the required depth using a Scala Penetrometer hammer and rods.
Winter 2005

The winter 2005 sampling was carried out on 1820 October 2005 and comprised:
(a)

One hand auger borehole to 2 m depth at each site to provide five

undisturbed soil samples at approximately 0.5 m depth intervals for
laboratory testing as for 8.3.2(a) above;

(b)

Taking three disturbed samples at approximately 0.5 m, 1 m and 1.5 m depths,

immediately above the level of the foregoing undisturbed samples;

(c)

Taking two undisturbed soil samples at 0.5 m and 1 m depth in an additional

hand augered borehole located adjacent to the first borehole for shrink-swell
index testing as for 8.3.3(e) above; and

(d)

Level survey measurement of extensometers and surface monuments.

The samples obtained in (a) and (b) above were submitted to the Auckland University
Geomechanics Laboratory, while the samples in (c) above were submitted to
Geotechnics Ltd, an IANZ accredited laboratory, for laboratory testing.
8.3.5

Summer 2005/06

Two sampling rounds were carried out over the 2005/06 summer period. The first
sampling round was carried out between 24 and 28 February 2006 and comprised:
(a)

One hand auger borehole to 2 m depth at Sites 2A, 2B and 2C to provide five
undisturbed soil samples at approximately 0.5 m depth intervals for
laboratory testing as for 8.3.2(a) above; and

(b)

Level survey measurement of extensometers and surface monuments.

The second sampling round was carried out on 24 and 25 March 2006 and comprised:

8.4

(c)

One hand auger borehole to 2 m depth at each site to provide five

undisturbed soil samples at approximately 0.5 m depth intervals for
laboratory testing as for 8.3.2(a) above;

(d)

Taking two undisturbed soil samples at 0.5 m and 1 m depth in an additional

hand augered borehole adjacent to the first borehole for shrink-swell index
testing as for 8.3.3(e) above; and

(e)

Level survey measurement of extensometers and surface monuments.

Laboratory testing
The soil laboratory testing, shown in Table 10 of this report, was generally undertaken
by the Geomechanics Laboratory of the University of Auckland, School of
Engineering during the periods discussed in Section 8.3, with the exception of the

36

shrink-swell testing which was carried out by Geotechnics Ltd.

The Atterberg Limits and linear shrinkage tests were undertaken to NZS 4402:1987
Methods of Testing Soils for Civil Engineering Purposes.
The core shrinkage tests were undertaken to AS 1289 Test Method 7.1.3:1998 Soil
Reactivity Tests Determination of the Shrinkage Index of a Soil Core Shrinkage
Index.
The soil suction tests were undertaken to AS 1289 Test Method 2.2.1:1998 Soil
Moisture Content Tests Determination of the Total Suction of a Soil Standard
Method, except that the thermocouple psychrometer referred to in the Standard was
replaced with a transistor psychrometer. The method is stated in the Standard as being
applicable for suctions ranging from 3.2pF to approximately 5pF. The transistor
psychrometer is discussed by Woodburn et al (1993) and Woodburn and Lucas
(1995).
The shrink-swell tests were undertaken to AS 1289 Test Method 7.1.1:1998 Soil
Reactivity Tests Determination of the Shrinkage Index of a Soil Shrink-Swell
Index.
Table 10

Laboratory test programme

Laboratory
test

Winter
2004

Summer 04/05
Month Month Month
1
2
3

Winter
2005

Summer 05/06
Month Month
1
2

Soil suction
(5 depths)

Atterberg Limits

Water content
(5 depths)

(3 depths)
Linear shrinkage
(3 depths)

Core shrinkage

(3 depths)
Shrink-swell Test

(2 depths)
Note 1. Core shrinkage test carried out on samples obtained instead of shrink-swell test.
2. No results available for tests as a component of the core shrinkage test was not obtained.

37

9.0

RESULTS OF FIELD EXTENSOMETERS AND LABORATORY TESTING

9.1

General
The soil profiles encountered in the boreholes at Sites 2A to 2F are shown on the
borehole logs presented in Appendix B of this report. The soil profiles at each site are
summarised in Table 11. The groundwater levels measured during the Stage II period
are shown on the individual site reports in Appendix C of this report.
The extensometer and laboratory test results will be discussed in general in Sections
9.2 and 9.3 before more detailed treatment of the as measured dry summer
conditions and analysis of the soil suction results in Sections 10.0 and 11.0.
Table 11

9.2

Summary of soil types at each test site

Site code

Suburb

Depth (m)

Soil unit

Soil description

2A

Manurewa

0.2-4.0

Tauranga Group (tp)

silty CLAY

2B

Otara

0.2-4.0

Tauranga Group (tp)

sandy and silty CLAY

2C

Howick

0.2-1.3
1.3-4.0

Waitemata Group
Waitemata Group

silty CLAY
sandy and clayey SILT

2D

Mairangi Bay
(North Shore)

0.3-4.0

Waitemata Group

silty CLAY

2E

Red Beach

2F

Pukekohe

0.3-3.0
3.0-4.0
0.2-1.5
1.5-2.4
2.4-3.2

Onerahi Chaos Breccia

Onerahi Chaos Breccia
Lithic Tuff / Basaltic Ash
Ash / Alluvials
Lithic Tuff

silty CLAY
clayey SILT
clayey SILT
silty CLAY
gravelly SILT

Extensometer results
9.2.1

General

The results of the measured survey levels of the extensometers installed at each of the
six sites are presented in item 4.0 of the site summary sheets presented in Appendix C
of this report. All values presented for the extensometers and surface monuments are
in millimetres (mm).
The extensometer values presented in Appendix C for each site have been compared
to the winter 2005 readings, which are considered to best represent zeroed
conditions. These are the most complete set of winter readings available as some of
the surface monuments were damaged during the 2004/05 summer period and had to
be reinstalled.
The negative values therefore indicate the amount of shrink that has occurred between
winter 2005 and the individual measured summer point relative to the 4 m deep
extensometer. As the level survey has an accuracy of 1 mm per measurement, the
extensometer readings presented in Appendix C, being the difference between two
readings, should all be considered to have a combined accuracy of 1.4 mm.
9.2.2

Observations

In the first instance it is noted that, in general, the order of magnitude of the measured
readings is relatively low in comparison to the soil expansivity classifications
38

presented in AS 2870. The range of maximum measured extensometer movements

over the Stage II monitoring presented in Appendix C is in the order of 15 mm to
25 mm which, if it took place over a full range of climate extremes, would correspond
to the slightly reactive and the lower end of moderately reactive soil classification
classes according to AS 2870 as indicated in Table 12.
Table 12

Classification by characteristic surface movement (ys)

(from Table 2.3 of AS 2870:1996)
Characteristic surface movement
0 mm <ys 20 mm
20 mm <ys 40 mm
40 mm <ys 70 mm
70 mm <ys

Classification of site
S Slightly reactive
M Moderately reactive
H Highly reactive
E Extremely reactive

Detectable movements were generally consistently obtained at the Manurewa, Otara,

Howick and Red Beach sites over the Stage II monitoring period. No significant
movement was measured in the extensometers and surface monuments at the North
Shore site. Detectable movement at the Pukekohe site was only identified in one of
the surface monuments installed.
Examining the depth of recorded movement it is noted that, as expected, the largest
movements recorded occur in the surface monuments and that the size of the
movements decreases with depth. What is of particular interest to the study is the
depth at which no measurable movement occurs. Examination of the extensometer
data indicates that, for Sites 2B to 2F inclusive, measurements greater than the survey
tolerance were obtained down to and including the 0.5 m deep extensometer, but not
in the 1 m deep extensometer. The exception to this observation is Site 2A
(Manurewa), which generally recorded measurable movements within the 1 m deep
extensometer and, in one instance, recorded measurable movement in the 1.5 m deep
extensometer.
9.3

Laboratory test results

9.3.1

General

The results of the soil classification tests (Atterburg Limits and linear shrinkage tests)
and the shrinkage index tests (shrink-swell index and core shrinkage tests) for
Sites 2A to 2F are shown on the individual site reports presented in Appendix C.
R-squared values are shown on the various following plots to indicate the correlation
between the parameters. The R-squared value provides a linear regression between the
actual test data points and the theoretical line and is an indication of the amount of
variation that is inherent in the linear model. An R-squared value of zero indicates
that there is no correlation between the parameters and a value of one shows a perfect
correlation. A rule of thumb is that R-squared values of below 0.3 are not considered
to be statistically significant in demonstrating the existence of a correlation at about
the 90% confidence interval.
9.3.2

Atterberg Limits and linear shrinkage

A Casagrande plot of the Atterberg Limits test data for Stage II is shown on Figure
11, which indicates that the soils plot slightly above or below the A line and are of
39

high to extremely high plasticity with liquid limits ranging from 50% to 115%. As
shown on Figure 12, the linear shrinkage values of the soils range from 11% to 25%.
Figure 12 indicates that approximately 70% of the test locations for Stage II fall
outside the definition of good ground as defined by NZS 3604:1999. The plot
shown in Figure 12 also illustrates that a good relationship exists between linear
shrinkage and liquid limit, which is not unexpected but was not, however, borne out
by the original Stage I data.

Plasticity Index (%)

80
70
60
50
40
30
20

A-Line

10
0
0

10

20

30

40

50

60

70

80

90

100

110

120

Liquid Limit (%)

Figure 11 Casagrande plot of plasticity index against liquid limit for Stage II test data
y = 0.1862x + 3.9665
R2 = 0.8451

30

"Expansive soils" in
terms of NZS 3604:1999

Linear Shrinkage (%)

25
20
15

"Good Ground" in
terms of NZS 3604:1999

10
5
0
0

10

20

30

40

50

60

70

80

90

100

110

120

Liquid Limit (%)

Figure 12 Plot of linear shrinkage against liquid limit for Stage II test data

9.3.3

Shrinkage index, Ips

The shrinkage index (Ips) is defined as the percent vertical strain per unit change in
soil suction (pF), determined from shrink-swell, loaded shrinkage or core shrinkage
tests, in accordance with AS 1289 Test Methods 7.1.1, 7.1.2 or 7.1.3 respectively. As
stated in Section 8.4, the shrinkage index testing for Stage II comprised two sets of
shrink-swell tests (Iss) and three sets of core shrinkage tests (Ics), with one of the core
shrinkage test sets being incomplete.
40

The purpose of carrying out two different test methods on the samples was to:
(a)

Compare and evaluate the test method carried out for the Stage I investigation
(core shrinkage) to the test method generally being carried out by geotechnical
practitioners in the Auckland region (shrink-swell); and

(b)

Compare and evaluate any differences in the test results from samples taken in
summer to those taken in winter.

As discussed in Sections 8.3.3 and 8.4 the samples submitted for shrink-swell testing
in the 2004/05 summer had the core shrinkage test carried out.
The results of the various shrinkage tests that were carried out during the Stage II
investigation are shown in Table 13.
Table 13

Stage II shrinkage index test results

A

Depth (m)

Oct-05

Mar-06

Mar-05

Mar-06

0.5
1.0

Iss
2.4
4.2

Iss
2.9
2.4

Ics
1.30
4.39

Ics
6.08
6.10

1.5
0.5

1.5

2.22

5.96
4.83

Otara

1.0

3.3

3.9

3.74

5.87

1.9
3.6

Howick

1.5
0.5
1.0

2.9
2.6

2.2
3.0

2.71
2.14

0.88
4.99
4.86

2.6
2.6

1.5
0.5
1.0

1.1

0.5
0.9

1.92
2.4

3.86
0.12
3.9

1.2
1.7

1.5
0.5
1.0

3.6
6.7

2.1
2.8

3.19
4.77

4.2
4.94
4.74

3.0
4.8

1.5
0.5

2.8

2.5

2.75

7.96
2.30

1.0

3.7

2.7

2.30

1.5

5.49

Average

3.3

2.3

2.8

4.5

Approx sample diameter (mm)

60

60

60

41

Approx L:D

2.0

2.0

1.7-1.8

1.7-1.8

Site

Manurewa

Mairangi Bay
(North Shore)

Red Beach

Pukekohe

Note: Iss = shrink-swell index

Average
shrinkage index
(A, B, C)
2.2
3.7

2.7
2.9
2.8

Ics = core shrinkage index

Although there is some variance between the October 2005 and March 2006 shrinkswell test results, they are generally of the same order, indicating that with respect to
(b) in Section 9.3.3, there is no significant difference between samples obtained and
tested in winter and summer on the basis of the test results reported herein.

41

9.4

Correlations between shrinkage index and other soil classification parameters

The relationships between the shrinkage index and the soil classification parameters
of linear shrinkage, liquid limit and plasticity index have been investigated. The
values of shrinkage index for each test site are plotted against the corresponding
values of plasticity index, linear shrinkage and liquid limit on Figures 13, 14 and 15
respectively.
It is apparent from Figures 13, 14 and 15 that there are weak correlations between the
shrinkage index and all three of the soil classification parameters. The R-squared
values increase to greater than 0.5 if the upper left value in the plots is considered to
be an outlier and removed from the data set.
Obtaining these correlations is in contrast to the analyses presented in the Stage I
Study Report which did not identify any readily discernable correlations for the
same classification parameters, but is in agreement with Cameron (1989) which found
a reasonably satisfactory linear correlation between the shrink-swell index (the
shrinkage index determined from a shrink-swell test) and linear shrinkage for several
soils from three Australian states.
NZS 3604:1999 defines expansive soils as being those that have a liquid limit of
more than 50% ... and a linear shrinkage of more than 15% .... If the linear
regression equations shown in Figures 14 and 15 are applied to these threshold values,
they yield corresponding shrinkage index values of 2% and 2.6% respectively.
It is therefore inferred that, on the basis of the Stage II data reported herein, that a
shrinkage index threshold value of around 2.5% could be adopted as a corresponding
value for classifying a soil as being within the definition of good ground as defined
by NZS 3604:1999.
y = 0.0291x + 1.3599
R2 = 0.231

5.0

Shrinkage Index (%)

4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0

10

20

30

40

50

Plasticity Index (%)

Figure 13 Shrinkage index vs plasticity index for Stage II test data

42

60

70

80

y = 0.0967x + 0.8946

5.0

R = 0.2148

4.5

Shrinkage Index (%)

4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0

10

15

20

25

30

Linear Shrinkage (%)

Figure 14 Shrinkage index vs linear shrinkage for Stage II test data

y = 0.0247x + 0.7693

5.0

R = 0.3289

Shrinkage Index (%)

4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0

20

40

60

80

100

120

140

Liquid Limit (%)

Figure 15 Shrinkage index vs liquid limit for Stage II test data

43

10.0

CLASSIFICATION OF SUMMER CONDITIONS

10.1

Extrapolation for design drought conditions

When the Stage II expansive soils investigation was set up it was anticipated that
some correlation between the extensometer readings and the SMD values on the day
of the reading would be obtained, and that this correlation could be extrapolated to the
designated drought condition. During the first order appraisal of the movements
recorded in the extensometers over the Stage II monitoring period and comparison
with the SMD values occurring at the time of measurement, it became apparent that
the expected correlation was not occurring, with the larger recorded movements
generally not matching the most negative SMD values. Examples from the Manurewa
and Otara test sites illustrating the lack of correlation (and even a negative
correlation) are presented in Figure 16.
Furthermore, it was observed that the largest recorded movements generally occurred
towards the end of the summer period, indicating that there is a lag between dry soil
moisture conditions and response within the soil zone. On reflection, this observation
is not unexpected in that the intensity and duration of preceding moisture conditions
would be expected to have a bearing on the soil response. For example, a spell of
prolonged dry weather without significant rainfall would be expected to have less
effect on subsurface conditions if it occurs at the start of a summer period than in the
middle of the summer after the subsoils have already experienced a period of
relatively dry weather.
A method of quantifying the intensity of summer conditions was therefore devised
that would be better able to reflect the climate conditions preceding the sampling
dates. Discussions were undertaken with Mr Alan Porteous of NIWA, which
confirmed that the proposed method was valid.
The method applies an area under the curve approach to the plot of SMD (mm)
against time (days). The deficit below a nominated threshold value leading up to the
sampling date is accumulated on a daily basis. The nominated threshold value is the
SMD (SMD) value corresponding to the onset of relatively dry conditions and has
been set at a value of -70 mm for the purposes of this study. No negative influence for
occasional significant rainfall events, which might otherwise reduce the cumulative
deficit value, has been allowed for. The area under the curve output so computed is
referred to as cumulative deficit days (CDD).
The CDD values are provided in the site summary reports presented in Appendix C.
The effectiveness of this approach is indicated in Figure 17 by the markedly increased
R2 values for the same Manurewa and Otara test sites from Figure 16.
In adopting this approach, it is acknowledged that it is a simplified method involving
judgement and that there are a number of areas where errors can arise as a
consequence. For example, a difficulty arises in evaluating and quantifying the
differences of dry periods in terms of the SMD values over shorter periods of more
intense drying, to values greater than say -90 mm, compared to a longer period with
less intense values varying between say -85 mm and -90 mm. In a specific example
from the Stage II data, it is noted that during the 2005 summer period the SMD
reading did not reach the theoretical drought condition of -90 mm. However, the total
rainfall that fell between January and April was the lowest on record for that four
month period.

44

(a) Surface Monument : Manurew a

Movement (mm)

y = -0.051x - 12.604
R2 = 0.0021
-18
-16
-14
-12
-10
-8
-6
-4
-2
0
-70

-75

-80

-85

-90

Soil Moisture Deficit (mm)

(b) Extensometer : Otara

y = 0.4812x + 30.183
2

R = 0.1759

Movement (mm)

-20
-15
-10
-5
0
5
-70

-75

-80

-85

-90

Soil Moisture Deficit (mm)

Figure 16

Extensometer results plotted against soil moisture deficit (SMD) values (mm)
for:
(a) Manurewa surface monument; and
(b) Otara 0.5 m extensometer.

45

(a) Surface Monument: Manurewa

y = -0.0144x - 2.1415
R2 = 0.5622

Movement (mm)

-12
-10
-8
-6
-4
-2
0
0

100

200

300

400

500

600

700

Cumulative Deficit Days

(b) Extensometer: Otara

Movement (mm)

y = -0.0085x - 1.7335
R2 = 0.6067
-8
-7
-6
-5
-4
-3
-2
-1
0
0

100

200

300

400

500

600

700

Cumulative Deficit Days

Figure 17

Extensometer results plotted against cumulative deficit days (CDD) for:

(a) Manurewa surface monument; and
(b) Otara 0.5 m extensometer

The CDD approach does, however, provide a means of estimating the dryness of the
ground conditions preceding the recording date and a best-guess estimation of the
movement that is likely to occur for extrapolating to future drought conditions.
Notwithstanding its acknowledged limitations, analyses have been carried out to
determine the robustness of the CDD methodology. SMD data has been examined for
the Auckland Airport and Auckland Owairaka (Mt Albert) weather stations extending
back to 1962 and 1954 respectively.
In order to test the robustness of the nominal -70 mm threshold that has been selected,
the CDD values for each summer were computed and ranked from highest to lowest
for the Auckland Airport data for threshold values of -67.5 mm, -70 mm and
-72.5 mm. All three threshold values yielded similar rankings of the summer periods,
indicating that the nominal value of -70 mm is appropriate.
The CDD values for each summer were computed and ranked from highest to lowest,
and then compared between the Auckland Airport and Auckland Owairaka stations. It
was observed that although some variation of the ranking of the years occurred, the
46

approach generally yielded similar results, again indicating that some degree of
robustness around the approach that has been used for this study.
It is the opinion of the authors that while acknowledging the simplifications and
assumptions within the CDD approach, the method provides a suitable means of
analysing and extrapolating the Stage II data so that:

10.2

(a)

Inferences can be drawn as to the expansive characteristics and behaviour of

soils in the Auckland region; and

(b)

The response of the soils under potential drought conditions can be predicted.

Soil suction results

As part of the Stage II investigation the measurement and testing of the soil suction
profile has been carried out at the six sites according to the methodology and practices
developed for the Stage I Study Report. The soil suction test results are shown in
Section 1.0 of the site summary sheets presented in Appendix C of this report.
The winter 2004 suction tests and the first set of the summer 2004/05 period suction
tests were taken at sample depths of 0.5 m, 1 m, 1.5 m, 2 m and 2.5 m. As discussed
in Section 9.2, the extensometer data obtained during the course of the Stage II
investigation indicated that the ground movements were occurring only within the
upper metre of the soil profile and did not extend to greater than approximately 1.5 m
depth. In order to better target the soils over the depth of soil suction change, the
remaining sets of suction samples were obtained at depths of 0.2 m or 0.3 m
(immediately below the topsoil layer), 0.5 m, 1 m, 1.5 m and 2 m.
In accordance with the Stage I methodology, soil suction test results of less than
3.2 pF for Sites 2A to 2F inclusive shown in Appendix C have been recorded as
3.2 pF. Graphic representations of the water content results and change in soil suction
values are also shown in Appendix C. In calculating the change in soil suction values
the summer test values have been generally compared to the preceding winter soil
suction values, with the exception of the 0.2 m/0.3 m suction test results, which are all
compared to the winter 2005 test results.
With respect to the soil suction test results presented on the site summary sheets in
Appendix C of this report, the following observations are made:
(a)

The maximum soil suction measurements recorded ranged from 3.70 pF at

Sites 2A and 2F (Manurewa and Pukekohe) to 4.05 at Site 2E (Red Beach).

(b)

In general, the maximum change in soil suction occurred in the samples

obtained from immediately below the ground surface for each of the summer
periods monitored, although there are some examples where the maximum
suction change was located in the deeper parts of the soil profile.

(c)

The maximum amount of soil suction change at the six sites was in the range
of 0.4 pF to 0.85 pF.

(d)

In some cases, particularly in the deeper parts of the profiles, negative changes
were identified i.e. the soil suction measured in summer was less than that
measured in winter.
47

(e)

The maximum change in soil suction values at the Manurewa, Howick and
Red Beach sites during Stage II was 0.5 pF, 0.4 pF and 0.6 pF. These are not
particularly greater than the maximums of 0.4 pF, 0.3 pF and 0.3 pF
respectively recorded for the significantly wetter summer periods during the
Stage I investigation.

(f)

The change in soil suction values for the 0.2/0.3 m deep test (and using the
0.5 m deep sample for S1) were plotted against the CDD values discussed in
Section 10.0. No discernable relationship was determined between these
parameters.

Table 14 shows the measured extensometer movements plotted against the change in
soil suction values over the various depths for all six sites.
As can be observed from the comparison of the data, examples can be shown where
high recorded suction changes match high extensometer movements, such as those
cases highlighted in yellow.
However, many more cases have been observed such as those highlighted in green,
where high extensometer movements were recorded with little or no soil suction
change and where high soil suction changes were determined with little or no
extensometer movement.
Cross-correlation of the recorded changes in soil suction with the measured
extensometer movements and calculated CDD values suggest that:

48

(a)

The recorded changes in soil suction are not readily able to be correlated with
the measured extensometer movements; and

(b)

There is only a weak correlation between soil suction change at the soil
surface and CDD.

Table 14

Comparison table between extensometer movement and soil suction change

49

10.3

Measured ground surface movements and correlation with inferred soil suction
change
The theoretical ground surface movement (ys) for Sites 2A to 2F have been calculated
for the measured soil suction change values obtained during the Stage II investigation
reported herein. Theoretical surface movements (ys) have been calculated according to
the relationship given by Equation 2 below (from Section 5.2) and using the shrinkage
index test results discussed in Section 9.3.3, a depth of design suction change (Hs) of
1.5 m, and the soil suction change values (u) obtained for each soil layer.
Hs

ys =

1
I
100 0

pt udh

Equation 5
(refer Section 5.2)

where
ys
Ipt
u
h
Hs

=
=
=
=
=

design characteristic surface movement (mm)

instability index (%)
suction change at depth (h) from the surface (pF)
thickness of soil layer under consideration (m)
depth below which no moisture or soil suction change
occurs (m)

The calculated theoretical ground surface movements for the five summer monitoring
periods are shown alongside the corresponding measured extensometer movements at
the top of the soil column (being the average of the two surface monuments) in
Table 15. The theoretical and measured movements for each site have been plotted
against each other on the graphics underlying Table 15.
As can be observed by the theoretical and measured ground surface movements and
the associated plots, there is generally poor or negative correlation between the
calculated theoretical and actual measured ground surface movements.

50

51

Table 15 Comparison between measured soil surface movements and calculated soil surface movements based on soil suction results

As the shrinkage index and the depth of moisture change are relatively simple
parameters to measure, it is considered that the more variable soil suction change
values are resulting in the lack of correlation between the measured and theoretical ys
values, as further discussed in Section 11.0.
Because of the higher confidence in the measured ys values obtained from the
extensometers, as compared with the theoretical ys values obtained from the soil
suction test data, it is considered more appropriate on the basis of the data obtained
during the Stage II investigation to adopt the measured ys values for back analysis to
obtain the change in soil suction value at the soil surface (us) for use in the
extrapolation to the drought conditions for the Auckland region discussed in Section
13.0.
As a check on this method an analysis was undertaken for the Stage II data to
determine if any correlation existed between the maximum recorded values rather
than the individual profiles. The maximum soil suction change value recorded at each
site over the Stage II period was entered into the theoretical movement calculations
for depths of moisture change of 1 m and 1.5 m and compared to the maximum
extensometer reading obtained at each site. These calculations and comparisons are
shown in Table 16.
Table 16

Comparison of theoretical movements using maximum soil suction change to

maximum measured extensometer movement for each site
Theoretical ys

SITE

Maximum
delta U (pF)

Hs = 1.0m

Hs = 1.5m

Maximum measured
ys (mm)

Manurewa

0.5

8.4

12.4

16

Otara

0.8

12.4

18.2

17

Howick

0.4

6.3

9.3

21

North Shore

0.85

7.2

10.6

Red Beach

0.6

13.5

19.8

22

Pukekohe

0.4

6.8

10.0

The calculated theoretical ys values shown in Table 16 give a reasonably good

approximation to the maximum ys values measured by the extensometers for a depth
of moisture change (Hs) of 1.5 m used with the maximum soil suction change over the
Stage II period. The main anomalies appear to be the Howick site where high
measured movement occurred with only small soil suction changes, and the North
Shore site where high soil suction changes were recorded but almost imperceptible
extensometer movement was measured over the Stage II period.
In general, however, it would appear that the predicted maximum assessment
process is sufficiently robust to give reasonable correlations between the extensometer
movements and the theoretical calculated movements. It is therefore considered that it
is appropriate to extrapolate the extensometer data for use in the back analyses to
determine the design soil suction profile for the Auckland region.

52

11.0

DRY SUMMER PREDICTIONS FOR AUCKLAND REGION

11.1

General
As discussed in Section 10.0, the CDD approach is considered to give the most
reliable method of evaluating dry summer soil conditions for the purposes of this
report.
Table 17 lists the top 10 ranked summer periods for the Auckland Airport data over
the period 1962 to 2006 on the basis of the CDD calculations.
Table 17

Cumulative deficit days (CDD) values for top 10 summer periods at Auckland
Airport
Rank

Year

CDD

Rank

Year

CDD

1
2

1974
1978

1376
1263

6
7

1983
1981

716
694

3
4

1973
1998

1181
1164

8
9

1994
2006

661
654

1970

827

10

1993

634

It is noted that the 2006 CDD value is ranked 9th out of the 44-year data set. The 2005
summer period was ranked 17th. These rankings are indicative of the general opinion
that the Stage II summer periods were reasonably dry, particularly when compared
to the Stage I summer periods of 2002 and 2003, which had rankings of 37th and 34th
respectively.
Of the years on record, the 1974 and 1998 summer periods encompassed the lowest
consecutive sets of daily SMD readings below the -90 mm threshold, being 34 days
and 23 days respectively. The CDD method bares this out, albeit imperfectly, ranking
these summers as 1st and 4th respectively. Our recollections of the 1998 summer as
being particularly dry, with some issues relating to expansive soils and subgrade
preparation occurring is, in our opinion, consistent with its ranking of 4th shown in
Table 17.
The commentary to AS 2870:1996 Supplement 1 states that:
The Standard describes the properties of the foundation by one parameter, the expected free
surface movement, ys. This is the vertical movement range expected during the life of the
house from a reasonable estimate of dry conditions to a reasonable estimate of wet conditions
and does not take into account the moderating effect of the footing system. The Standard
nominates 50 years as the life of the house and reasonable as the level that could be
expected for 19 houses out of every 20. This does not mean that the house is not expected to
last more than 50 years nor that 1 in 20 houses could fail. It is, however, more reliable than
using average conditions or an undefinable extreme concept.
In this assessment, ys should be interpreted as the characteristic value that has a 5 percent
chance of being exceeded in the life of the house which may be taken as 50 years.
The characteristic value is defined as the value that has a 95 percent chance of occurring in the
life of the structure. Thus it is not necessary to consider extremes of drying or wetting of the
profile.

A strict interpretation of the AS 2870 commentary would lead to the consideration

that in the 50-year design life of a residential building only one-in-20 buildings will
53

develop a problem relating to expansive soils i.e. that the probability of any one house
developing a problem due to expansive soils within any given year is:
= 1/20 x 1/50
= 0.001 or one-in-1000
= 0.1% annual exceedance probability (AEP).
This interpretation of the design probabilities is supported by HB 28:1997 which
states:
The definition also includes the concept that ys is a value that has a 5% chance of being
exceeded in the 50-year nominal life of a house. Thus it is not the most extreme value possible
at a particular site, although with only one chance in a thousand of occurring at a site in any
one year, it should not be a common event. The definition of the design surface movement is
similar to that used for wind and floor loads.

It is noted that a strict mathematical interpretation of a 5% chance in 50 years gives

a return period of 975 years i.e. 1000 years is an approximation of sufficient accuracy
for the purposes of this report.
To show:
A 5% probability in 50 years corresponds to a return period of 975 years (so that
1000 years is only an approximation).
Explanation:
Let P = Probability in one year
1-P = Probability not occurring in one year
(1-P)n = Probability not occurring in n years
1-(1-P)n = Probability occurring in n years
i.e. for 5% probability in 50 years
0.05= 1-(1-P)50
Rearranging to solve for P gives
P =1-0.951/50
=0.001025
=0.1025%
Let T be the period, then
T=1/P=975 years
Hence annual probability is 0.1%, corresponding to a return period of 975 years.

11.2

Statistics extreme value analysis

11.2.1

Theory

The analysis and statistical treatment of extreme events such as the return periods of
floods comes within Type 1 Extreme Value Theory, commonly known as Gumbel
Analysis. This provides a statistical treatment of data sets that follow a Gumbel
Distribution for many small events and few large events.
Although periods of dry weather and droughts follow a different spatial time span and
54

intensity than one-off events such as rainfall storms, Gumbel Analysis nevertheless
provides a means of estimating and modelling the true statistical distribution and
behaviour of dry periods represented by the CDD data adopted in this study.
For the Gumbel Analyses carried out for this report, the CDD values have been
calculated and ranked from lowest to highest from i = 1 for the lowest CDD values
and i = N for the highest ranked CDD value. Only positive CDD values were included
in the analyses so that CDD values of zero did not impose a skew to the tail of the
analysis.
For each CDD the estimated probability of a smaller CDD is given by:
Pi = i / (N+1)
and

Xi = -ln [-ln (i/N+1)], where

i=

rank of event (CDD value) for i = 1, 2, 3 .... N

where N = total number of events (i.e. CDD sample size)

Pi = estimated probability of the occurrence of a smaller CDD event

than i
Xi = dimensionless expression of rank of i to enable plotting with
CDD value of i
Plotting CDD against Xi gives a linear relationship that can be modelled by a simple
regression analysis from which the CDD value corresponding to a return period T is
given by:
XT = -ln [-ln (1-1/T)]
11.2.2

Data sets adopted in Stage II

The CDD data sets adopted for this report are provided in Appendix D. NIWA have
provided data extending from 1963 to 2006 for the Auckland Airport monitoring
station (used for the Manurewa, Otara, and Howick sites), from 1967 to 2006 for the
Albany station (used for Mairangi Bay and Red Beach sites), and from 1996 to 2006
for the Pukekohe station (Pukekohe site).
The return periods of interest to this study are the 20, 50, 100, 300, 500 and 1000-year
return period CDD values. The calculated CDD values associated with each of the
foregoing return periods for the Auckland Airport, Albany and Pukekohe weather
stations are presented in Table 18. The CDD values for the Stage II monitoring period
are also presented for comparison.
From Table 18 it can be observed that the 2005 summer event at Albany and the 2006
summer event at Pukekohe were greater than one-in-20-year return period events. In
particular, the 2005 Albany CDD value was the highest value within the 39-year
record available. In comparison, the 2006 Airport CDD value was not even a one-infive-year event (CDD value of 721) for that data set.
Table 18

Cumulative deficit days (CDD) calculated by Gumbel Analyses for various

return periods

55

Cumulative deficit days

Return period
(years)

Auckland
Airport

Albany

20
50
100
300
500

1178
1468
1686
2028
2186

980
1220
1402
1687
1820

631
800
926
1126
1218

1000

2404

2001

1345

2005 summer

394

1053

308

2006 summer

654

482

672

Pukekohe

It is considered that this variability of the CDD values highlights two important
factors:

11.3

(a)

The high variability of weather conditions in the Greater Auckland Region and
the difficulty in extrapolating weather phenomena. The corollary to this is that
there is uncertainty in the assumption that the data recorded at the weather
stations is representative of the weather conditions and soil responses at the
test sites. If further research is undertaken into the expansivity of soils in the
Auckland region, it is recommended that consideration be given to installing
and monitoring a set of extensometers adjacent to a weather recording station
to help remove some of this uncertainty.

(b)

A potential lack of robustness in the CDD approach. While considerable effort

has been applied to come up with an accurate method as possible to assess
drought/dry period conditions which is based on a reasonably scientific
approach, the method is unsubstantiated. It is a broad-brush approach that is
the best that the authors could devise within the confines and limitations of
this study.

Results and extrapolations

The correlations between the measured ground surface movements (measured by the
extensometers) and the CDD values for each of the five relevant sites (i.e. excluding
the North Shore site) are shown on Figure 18 of this report. The projection equations
from these correlations have been used to determine the predicted ground surface
movement (ys) for the CDD values that correspond to the 20, 50, 100, 300, 500 and
1000-year return periods shown in Table 18. The predicted ys values for the foregoing
return periods are shown in Table 20.

56

Figure 18

Measured soil surface movements plotted against CDD and return period

Having obtained the predicted ground surface movement (ys) for a particular return
period, knowing the tested shrinkage index values (Ips, Section 9.3.3), and adopting an
Hs value of 1.5 m, analyses have been undertaken using Equation 2 to determine the
theoretical soil suction at the soil surface (us) that is required to match the predicted
ground surface movement.
The equivalent Ips values for each site, for an Hs value of 1.5 m, are shown in
Table 19.
57

Table 19

Equivalent Ips values

Site
Manurewa
Otara
Howick

0 to 0.5 Hs
2.2
1.9
2.6

0.5 Hs to 1.0 Hs
3.7
3.6
2.6

Red Beach
Pukekohe

3.0
2.7

4.8
2.9

Note: Hs = 1.5 m

The predicted ys and the corresponding us values for the 20, 50, 100, 300, 500 and
1000-year return periods are presented in Table 20. As no reliable correlation or
measured extensometer movement was obtained from the North Shore site, that site
has been omitted from the extrapolations.
Table 20

Analysis results showing us values required to match ys movements predicted by

Gumbel Analyses for 20, 50, 100, 300, 500 and 1000-year return period events.

SITE

Prediction Equation

1 in 20 yr event

1 in 50 yr event

1 in 100 yr event

Ys Predicted

us required

Ys Predicted

us required

Ys Predicted

us required

Manurewa

Ys = -0.0144x - 2.1415

-19.1

0.77

-23.3

0.94

-26.4

1.06

Otara

Ys = -0.0225x - 1.0949

-27.6

1.21

-34.2

1.50

-39.0

1.71

Howick

Ys = -0.015x - 7.3598

-25.1

1.08

-29.4

1.26

-32.6

1.40

Red Beach

Ys = -0.0135x - 7.1449

-20.4

0.62

-23.6

0.72

-26.1

0.79

Pukekohe

Ys = -0.008x + 0.4126

-4.6

0.19

-6.0

0.24

-7.0

0.28

-19.4

0.77

-23.3

0.93

-26.2

1.05

Average

SITE

Prediction Equation

Manurewa
Otara

1 in 300 yr event

1 in 500 yr event

1 in 1000 yr event

Ys Predicted

us required

Ys Predicted

us required

Ys Predicted

us required

Ys = -0.0144x - 2.1415

-31.3

1.26

-33.6

1.35

-36.8

1.48

Ys = -0.0225x - 1.0949

-46.7

2.05

-50.3

2.21

-55.2

2.42

Howick

Ys = -0.015x - 7.3598

-37.8

1.62

-40.1

1.72

-43.4

1.86

Red Beach

Ys = -0.0135x - 7.1449

-29.9

0.91

-31.7

0.96

-34.2

1.04

Pukekohe

Ys = -0.008x + 0.4126

-8.6

0.35

-9.3

0.37

-10.3

0.41

-30.9

1.24

-33.0

1.32

-36.0

1.44

Average

Note: 1. The u values are based on an Hs value of 1.5 m.

2. The prediction equations correspond to the equations for the surface monuments presented in
Appendix C.

Using the 100 and 500-year return period drought events with a mid-range return
period of 300 years to define a benchmark range for foundation performance, the
following conclusions can be drawn from the data presented in Table 20:

58

(a)

For an assumed Hs of 1.5 m, the average us ranges between 1.05 pF and

1.32 pF, with a value of 1.24 pF for the 300-year return period drought event,
which approximates the lower u value of 1.2 pF in AS 2870 Table 2.4.

(b)

The predicted ground surface movements (ys) for four of the five sites are less
than or equal to 40 mm and therefore fall into the slight to moderate soil
classification category of AS 2870 (i.e. ys 40 mm).

(c)

Using the 1000-year return period event as the basis for the foundation
distortion limit for an extreme drought condition, for an assumed Hs of 1.5 m,
the average us is 1.44.

The Stage I report presented three possible soil suction profiles to be used in the
calculation of the design surface movement ys, viz Profiles Alpha, Beta and Gamma,
as shown in Table 21.
Table 21

Stage I soil suction profiles

Profile
Alpha
Beta
Gamma

Hs
2.0
1.5
1.5

us
1.5
1.5
1.2

In the Stage I study it was recommended that soil suction Profile Alpha be adopted as
a conservative measure until further evidence had been obtained. The projected us
values from the CDD values presented in Table 18 and (b) and (c) above, suggest that
soil suction Profile Gamma is appropriate for use in calculating the theoretical design
surface movement in assessing the expansivity of soils in the Auckland region.
Performance and correlation of extensometer movements and soil suction
measurements under severe drought conditions would, however, be required in order
to confirm the design conditions should such a severe drought occur.
11.4

Scaling factors
It is recommended in Section 14.0 that foundations be designed to perform under
drought conditions with a return period in the range 100 to 500 years, with a midrange value of 300 years, with the design verified for an extreme drought with a return
period of 1000 years.
If a scaling factor of 1.00 is adopted for the calculated ys corresponding to the 300year return period design drought, the scaling factors shown in Table 22 have been
determined for various other return period events, based on the predicted ys values
shown in Table 20.
For any selected return period drought event, the ys value calculated using soil suction
Profile Gamma (Hs = 1.5 m, us = 1.2 pF) may be scaled using the appropriate scaling
factor from Table 22.
Table 22

Scaling factors
Return period
20 years
50 years
100 years
300 years
500 years
1000 years

Scaling factor
0.65
0.78
0.88
1.00
1.11
1.21

It is suggested the scaling factors in Table 22 be used to assess foundation

performance as follows:
(a)

The serviceability performance of any particular foundation system be

determined for the design drought, with a 300-year return period, using the ys
value calculated for soil suction Profile Gamma and a scaling factor of 1.0.

59

(b)

The foundation performance for the extreme drought event, with a 1000-year
return period, be determined either:
(i)

By applying a scaling factor of 1.2 to the ys value calculated for soil

suction Profile Gamma for the purposes of the soil foundation
interaction analysis prescribed by AS 2870 Part 4 (e.g. Mitchell
Method); or

(ii)

By calculating the flexural demand (Mu) for foundation elements

calculated from Profile Gamma (Hs = 1.5, u = 1.2) to incorporate a
universal load factor of 1.2 rather than 1.0.

The result of this will be that foundation strength and stiffness will cover the
design drought (i.e. the mid-range one-in-300-year drought event) while the
extreme drought event is covered for strength limit state only. Refer also
Section 14.2.
The basis for this assessment of foundation performance, based on AS 2870/HB
28:1997 concepts, is discussed further in Appendix E of this report.
11.5

Foundation performance assessment

The shrink-swell index values for tests carried out in the Auckland region are
generally in the range of 1% to 7%. Using soil suction Profile Gamma (Hs = 1.5, u =
1.2), the incremental Iss values will correspond to the theoretical ys movements and
site classifications shown in Table 23.
Table 23

Correlation between shrink-swell index and ys and site classification

Iss (%)
1%
2%
3%
4%
5%
6%
7%

ys (mm)
10.8
21.6
32.4
43.1
53.9
64.7
75.5

Site classification
S
S to M
M
M to H
H
H
H to E

From Table 2, the ys values for the various site classifications are:
(a)

Class S = 020 mm;

(b)

Class M = 2040 mm;

(c)

Class H = 4070 mm; and

(d)

Class E = >70 mm.

It is therefore apparent that the greater part of the measured Iss range (<4%)
corresponds to Class S and M, which for normally clad frames will generally be
considered to be adequately mitigated by specifying a foundation embedment depth of
450 mm below cleared ground level.
For Iss values in the upper region of the natural range, i.e. >4% Iss < 6%, a minimum
60

foundation embedment depth of 600 mm below finished external ground levels would
generally be sufficient to mitigate against potential shrink-swell issues associated with
expansive soils for clad frame, masonry veneer and articulated veneer cladding.
These indicative foundation depth requirements will satisfy angular distortion
requirements for external wall cladding, but will not mitigate heave or sag which
might arise from insufficient site pre-treatment (refer Section 12.0).
Specific design of slab and foundation systems to limit angular distortion will
generally be required for foundations on Class H and E sites.
11.6

Effects of climate change

The methodology and field data correlation presented in this report is based on
investigations carried out in the period 2002 to 2006.
Because the primary determinant of the recurrence of the design drought condition is
the SMD record supplied by NIWA for the period up to 2006, any impact of future
climate change should be reflected through the ongoing monitoring of that record and
the CDD parameters derived from it.
Whilst domestic dwellings currently being constructed will have a design life
extending some 50 years into the future, the nature of the potential change to the SMD
record for Auckland projected over that period is not clear.
It is therefore recommended that the SMD profile for Auckland be re-evaluated at not
less than 20-year intervals using the methodology adopted in this report.
Any changes recorded would be reflected in the values of Hs and u recommended
for foundation design.

11.7

Soil suction change profile for regions other than Auckland

The soil suction change profile recommended in this report is specific to the Auckland
region.
The soil suction change profile for other regions in New Zealand depends on the SMD
profile for the particular region, which in turn depends on the local climate (plus any
change to that local climate in the future).

61

12.0

RESIDUAL SWELLING POTENTIAL IN AUCKLAND SOILS

12.1

Objective
Since the introduction of NZS 3604:1999, it has become progressively more common
for geotechnical practitioners to undertake laboratory testing of subsoils within a
subdivisional development in order to classify the expansive characteristics of the
subsoils.
The most common method for calculating the design characteristic surface movement
by geotechnical practitioners is by determining the shrink-swell index (Iss) in order to
estimate the instability index (Ipt) for use in calculating the design characteristic
surface movement (ys) value.
The investigation reported herein involved a review of the records on file at the
various local TAs in the Auckland region over the previous five years or so in order to
obtain copies of the shrink-swell index test data held in those files, and to investigate
the extent of swell contribution to the Iss index, and to determine if any significant
variation or trends could be established to show if relationships could be inferred for
Auckland soils by:
(a)

Depth; or

(b)

Soil type; or

(c)

Sampling condition (summer vs winter).

The investigation methodology and results obtained are discussed separately.

12.2

Option for limiting soil expansivity

AS 2870 Section F4 states that: on a site that is wet throughout the profile at the time
of construction, a reduction of ym for edge heave not exceeding 40% may be made.
This reduction factor is directly applicable to sites that have been pre-treated to
maintain a high water content for the full duration of the pre-construction period.
The adoption of a scaling factor of around 60% gives rise to the possibility of two
considerations that form the basis of the investigation and analyses reported herein:
(a)

To determine, on the basis of the available test data for the Auckland region, if
the amount of residual swelling can be limited or reduced further than the
factor of 0.7 recommended in AS 2870.

(b)

To investigate the possible effects of pre-treatment of building sites in the

Auckland region.

This report provides the results of the investigation of (a) and (b) above as an interim
report to Manukau City Council as part of the Stage II BRANZ investigation into the
expansivity characteristics of soils in the Auckland region.
12.3

Conclusion
Based on the analysis of the collected data, it can be inferred that the proportion of

62

swell component within the shrink-swell index test does not exhibit significant
variation by season or SMD value, soil type or sample depth.
It therefore follows that, for foundation design purposes, the swell component can be
treated equally across soils in the Auckland region. On the basis of the data collected
to date, a case can be argued for the soils in the Auckland region being generally at or
near their potential maximum swell, even for typical summer conditions.
This means that for a typical building platform, the foundations and grade slab will be
unlikely to be subject to significant heave conditions, provided that the subgrade
within the footprint of a dwelling is maintained at or close to its natural water content
during construction.
12.4

Limited swelling potential

On the basis of the foregoing, it is considered that soil expansivity at a particular site
can be limited by specific design (incorporating site pre-treatment), as suggested in
Section 12.4.1(b). As a consequence, the following reductions in ground deformation
due to shrink-swell can be assumed:
(a)

The free unloaded heave (ym) value can be reduced from a value of 0.7 ys to
0.4 ys.

(b)

For a site that is subject to pre-treatment or maintenance protection, by either

pre-saturation or by the placement of a barrier to subgrade moisture loss
respectively, then a design allowance of ym for edge heave not exceeding
0.2 ys may be made.

The impacts of these recommendations on sites with highly expansive soils are
described in Sections 12.4.1(a) and 12.4.1(b).
12.4.1
(a)

AS 2870Appendix F: Soil Parameters and Footing Design Methods

Design recommendation for stiffened raft/waffle slab foundations
incorporating perimeter piling on Class H and E sites
Appendix F4 (a) (on p67) states that:
On a site that is wet throughout the profile at the time of construction, a reduction of ym
for edge heave not exceeding 40% can be made.

For stiffened rafts/waffle slab foundations incorporating unconnected

perimeter piles, where the subgrade has been pre-treated to achieve nearsaturation conditions prior to casting slab this leads to the design cases
shown in Table 24.

63

Table 24

Design ym recommendations for slab foundations incorporating perimeter piles

cast on near saturated sites
Design case

Design ym
value(1)
+0.2 ys

1(a)

centre heave

1(b)

edge subsidence

-0.7 ys

2(a)

centre subsidence

-0.4 ys

2(b)

edge heave

+0.2 ys

Note: 1.
2.
3.

Remarks (2), (3)

Centre heave suppressed by subgrade pretreatment to near saturation conditions residual
design allowance say 0.2 ys max
Edge beam system to be supported by perimeter
piling to suppress this case
Refer AS 2870 Appendix F4(a), which allows a
40% reduction in ym for this case
Edge beam system either lifts off perimeter piles
or piles/edge beam system must resist swelling
pressures (say 20 kPa based on residual 10%
swelling) if structurally connected

Mound movements are (+) upwards, (-) downwards referenced against mound
surface level at time of construction.
Where perimeter edge beam supported on piles (as hard supports without
structural connection to foundation).
Piled foundations supporting perimeter beams to be designed for negative skin
friction effects, as well as factored design loads imposed onto the pile head.

Immediately prior to casting foundation slab, the residual swell strain must be
less than 15% of the confirmed shrink-swell index value when measured in
accordance with AS 1289:1998 Method 7.1.1 for the above design basis to
apply.
(b)

Site pre-treatment (saturation) recommendations in accordance with

AS 2870 Appendix F4(a) for Class H and E sites
(i)

Record shrink-swell tests (2No) for particular site to verify

recommendation of subdivision completion report.
NB: Output = Iss and ys

(ii)

Recover samples for swell tests (0.25 m and 0.75 m depth) at each of
2No locations on-site (1No under edge beam, 1No in central area of
the building envelope) and a single sample from 1.5 m depth within the
central area of the building envelope.
NB: Samples to be taken after saturation pre-treatment of subgrade and
immediately prior to casting of foundation slab, and to be tested in
accordance with AS 1289:1998 Method 7.1.1.

(iii)

12.5

Application of AS 2870 standard methods

(a)

64

If the swelling strain is less than 15% of the shrink-swell index Iss (%)
value, then the reduced centre subsidence (edge heave) provisions of
AS 2870 Appendix F4(a) can be used, following the (4No) load cases
specified by the design cases in Section 12.4.1(a).

The design values for ym derived for pre-treated sites described above are
directly applicable to standard designs formulated for use with AS 2870,
subject to:

(b)

(i)

Sites being re-classified consistent with the foundation system

adopted, by direct scaling of the derived ys values.

(ii)

Edge beams associated with standard designs being modified as

necessary to accommodate e.g. piled supports.

In general, specific design will be necessary to confirm performance adequacy

of foundation elements e.g. edge beams, piles.

13.0

FOUNDATION ANALYSIS: PERFORMANCE FRAMEWORK

13.1

In general, the performance requirement for buildings can be related to

13.2

13.3

(a)

The factors and events that buildings might be subject to;

(b)

The impacts on buildings, and building users, that society is prepared to

tolerate generally; and

(c)

How society might tolerate different impacts for different types of buildings.

The factors and events of Section 13.1(a) can be categorised into three groups:
(a)

Factors that affect buildings all the time, such as effects of gravity or human
activity;

(b)

Specific events, such as earthquake, storms that bring strong winds, heavy
rain or snow, and noise nuisance; and

(c)

Factors that affect the ability of a building to respond to demands over a long
period of use, such as corrosion, rot and decay, or exposure to UV radiation.

For the performance requirements of Section 13.1(b) and (c), the Building Code seeks
to provide reasonable protection from the effects of demands on buildings,
acknowledging that it would be uneconomic and too restrictive to aim to eliminate all
risk.
Society, therefore, tolerates some impacts on buildings in certain circumstances. The
impacts that are tolerated depend partly on the size of the event that caused them and
the likelihood that such an event will happen.

13.4

Using this approach, Code writers have developed descriptions for tolerable impact
levels (TIL) ranging from insignificant to extreme, setting out in each case what
society might tolerate in terms of:
(a)

Impacts for occupants;

(b)

Economic impacts;

(c)

Social impacts;

(d)

Environmental impacts;

65

and then classifying buildings by type and/or intended use against their tolerable
impact.
13.5

Society's tolerance of impacts also depends on how vulnerable the people in the
buildings are, and how important the building is to society. Regulators might therefore
classify different types of buildings into four main performance groups,
depending on e.g. life risk factors such as:
(a)

Function of the building;

(b)

Proportion of time the building is occupied by people;

(c)

Familiarity of occupants within the building; and

(d)

Whether vulnerable or special populations use the building.

13.6

Under this classification, Performance Group (PG) 4 at the high end of the scale
might cover buildings that are essential to post-disaster recovery, or are associated
with hazardous facilities (e.g. hospitals), whereas at the low end of the scale
Performance Group (PG) 1 might describe buildings posing low risk to human life, or
a low economic cost should the building fail (e.g. ancillary buildings).

13.7

In addressing the TIL for domestic/residential buildings, it is noted that there has
recently been an increased focus on prevention of economic loss (by adequate design,
construction, maintenance or otherwise) for homeowners. This is reflected in Building
Act 2004 (BA04) Section 4 Principles to be applied in performing functions or
duties, or exercising powers, under this Act.
BA(4)(2)(a)(i) states in part:
(a)

13.8

when dealing with any matter relating to 1 or more household units

(i)
the role that household units play in the lives of the people who
use them, and the importance of
(A)
the building code as it relates to household units; and
(B)
the need to ensure that household units comply with the
building code:
(ii)
the need to ensure that maintenance requirements of household
units are reasonable:
(iii)
the desirability of ensuring that owners of household units are
aware of the maintenance requirements of their household units.

It is appropriate that these principles also be acknowledged in the level of

conservatism brought to setting up of TILs for dwellings founded on expansive soils.
Tolerable impacts for buildings in
Performance Group 2

Range of specific events considered

TIL5

66

TIL4

TIL3

Flooding

1/200

1/100

1/50

Volcanic activity

1/2000

1/1000

1/500

Snow/ice

1/250

1/150

1/50

Wind

1/500

1/500

1/100

Earthquake

1/1000

1/500

1/100

Soil expansivity

1/1000

1/500

1/100

13.9

The Department of Building and Housing (DBH) has recently published a

consultation document Building for the 21st Century: Review of the Building Code
(July 2007) signalling the intended basis for the review of New Zealands Building
Code. From Table 10 of that document, the following comparable events are noted
that might impact on homeowners with the same TILs.
TABLE 10: PERFORMANCE FRAMEWORK
Tolerable impacts

Volcanic
Activity

Snow/Ice

Wind

Performance Group

Flooding

Earthquake

Annual Probability of Event

1/500

1/5000

1/500

1/2500

1/2500

Extremely
Low

1/200

1/2000

1/250

1/500

1/1000

Very Low

Tolerable Impact
Level 6
(Extreme)

Tolerable Impact
Level 5
(Very Severe)

Tolerable Impact
Level 4
(Severe)

Tolerable Impact
Level 4
(Severe)

1/100

1/1000

1/150

1/500

1/500

Low

Tolerable Impact
Level 5
(Very Severe)

Tolerable Impact
Level 4
(Severe)

Tolerable Impact
Level 3
(High)

Tolerable Impact
Level 2
(Moderate)

1/50

1/500

1/50

1/100

1/100

Medium

Tolerable Impact
Level 4
(Severe)

Tolerable Impact
Level 3
(High)

Tolerable Impact
Level 2
(Moderate)

Tolerable Impact
Level 1
(Mild)

1/20

1/25

1/25

1/25

1/25

High

Tolerable Impact
Level 2
(Moderate)

Tolerable Impact
Level 1
(Mild)

Tolerable Impact
Level 1
(Mild)

Tolerable Impact
Level 1
(Mild)

Tolerable Impact
Level 0
(Insignificant)

Tolerable Impact
Level 0
(Insignificant)

Tolerable Impact
Level 0
(Insignificant)

Tolerable Impact
Level 0
(Insignificant)

"Everyday"

Chances of
Event
1

Tolerable Impact
Level 6
(Extreme)

Tolerable Impact
Level 5
(Very Severe)

Tolerable Impact
Level 5
(Very Severe)

Notes:
1. This table shows the tolerable impacts for a range of events and performance groups. The annual
probabilities of events are intended to give an indication of the scale of event being considered in
setting the tolerable impacts. They have been chosen to align with values given in AS/NZS 1170.
2. The table has been derived for structural performance requirements, viz B1. The DBH intend to
develop the concept to cover all relevant disciplines.

13.10 For groups of domestic dwellings (Performance Group (PG) 2) within a community
which are, for example, located on expansive soils, and potentially subject to
economic damage through extreme drought conditions, the following is able to be
derived from the attached Table 10 Performance Framework which suits structural
design actions.
13.11 Given that domestic dwellings and/or residential buildings (as a general
ownership class) will be the main focus of the economic analysis, the following is
derived:
(a)

Domestic/residential buildings PG2

(b)

Design event for evaluation with corresponding TIL

(i)
(ii)
(iii)

Extreme drought event (TIL5/very severe) with event frequency rated

very low
Design drought event (TIL4/severe) with event frequency rated low
Serviceability drought event (TIL3/mild) with event frequency rated
mild.
67

(c)

Evaluation threshold 90% of buildings will perform better.

14.0

NEW ZEALAND BUILDING CODE REQUIREMENT

14.1

Performance statements (B1.3.1 and B1.3.2)

Structural performance requirements for foundations in terms of the New Zealand
Building Code Regulations are substantively:
B1.3.1 Buildings, building elements and sitework shall have a low probability of
rupturing, becoming unstable, losing equilibrium, or collapsing during
construction or alteration and throughout their lives.
B1.3.2 Buildings, building elements and sitework shall have a low probability of
causing loss of amenity through undue deformation, vibratory response,
degradation, or other physical characteristics throughout their lives, or during
construction or alteration when the building is in use.
B1.3.3 Account shall be taken of all physical conditions likely to affect the
stability of buildings, building elements and sitework, including:
(a) Self-weight,
(b) Imposed gravity loads arising from use

(m) Differential movement

(n) Vegetation

(q) Time dependent effects including creep and shrinkage, and

(r) Removal of support.
B1.3.4 Due allowance shall be made for:

(c) Effects of uncertainties resulting from construction activities, or the sequence

in which construction activities occur,
(d) Variation in the properties of materials and the characteristics of the site, and
(e) Accuracy limitations inherent in the methods used to predict the stability of
buildings.
B1.3.7 Any sitework and associated supports shall take account of the effects of:
(a) Changes in ground water level,
(b) Water, weather and vegetation, and
(c) Ground loss and slumping.

Under Clause A2 Interpretation the term siteworks (referred to in B1.3.7 etc)

means:
Sitework means work on a building site, including earthworks, preparatory to or
associated with the construction, alteration, demolition, or removal of a building.

Subsequent AS 2870 performance requirements should be read in the context of the

above-mentioned New Zealand Building Code requirements.
14.2

Basis of design
Given the stiffness-based design approach for the 5% probability of exceedance in
50 years i.e. 1000-year return period drought event advocated by the AS 2870

68

Standard, and the performance framework for building design set out in Section 11.0,
the impacts have been considered of using a purely strength-based analysis for the
representative floor slab analysed in Appendix F, Section F5.5.2 (Figure F3).
From the structural engineering perspective, one question that could be asked is:
Can we not proportion the foundation for a lesser baseline event, e.g. 300-year
drought, and cover the extreme event, viz 1000-year drought, by introduction of
appropriate load factors?

At face value, this ignores the fact that foundation stiffness is the primary design
driver for foundations on expansive soils, and merely providing a foundation system
with adequate flexural strength will generally not achieve a complying design i.e. a
design which limits superstructure damage to levels prescribed within AS 2870
Appendix C.
However, in order to check out the impacts of the normalised T=300 return period
decision, some parameter studies were undertaken to investigate the variation of
various design parameters, viz:
(a)

Foundation unsupported edge distance, viz [e(T)];

(b)

Foundation flexure under the centre heave, viz [M(T)];

(c)

Foundation deflection or stiffness, viz [(T)]; and

(d)

Angular distortion derived from (a) and (c), viz [(T)=(T)/e(T)];

on a stiffened foundation slab system subject to a drought-generated centre heave

ground profile, where the drought return period T varies over the range 20, 50, 100,
300, 500 and 1000 years.

Figure 19

Foundation plan and section

A representative rectangular stiffened slab system (8 x 16 m), as shown in Figure 19,

was analysed by the Mitchell Method using the soil structure interactive SLOG
software as described in Appendix F.
The raft comprised 400 deep x 300 wide foundation beams in a nominal 4 x 4 m grid
69

cast integrally into the slab, these being subject to a uniformly distributed load of 4
kPa and a line load of 8 kN/m to represent the weight of the external perimeter walls.
The results, normalised to the T=300-year design drought, are summarised for the
centre heave design condition for the short and long span directions in the following
table.
Long span: centre heave
Edge distance

Return
period T
(yrs)

e(m)

20
50
100
300
500
1000

0.65
0.70
0.76
0.85
0.89
0.95

Normalisation
factor
0.74
0.83
0.89
1.00
1.05
1.12

Foundation
deflection/stiffness

Normalisation
(mm)
factor
7.5
0.48
10.3
0.66
12.3
0.79
15.6
1.00
17.1
1.09
19.1
1.23

Foundation flexure
M
(kNm)
6.80
9.00
10.60
13.00
14.10
15.50

Normalisation
factor
0.52
0.69
0.82
1.00
1.08
1.19

Angular
distortion
Normalisation
factor = /e
0.66
0.79
0.88
1.00
1.04
1.10

Short span: centre heave

Edge distance

Return
period T
(yrs)

e(m)

20
50
100
300
500
1000

0.63
0.70
0.76
0.85
0.89
0.95

Normalisation
factor
0.74
0.83
0.89
1.00
1.05
1.12

Foundation
deflection/stiffness

Normalisation
(mm)
factor
9.4
0.53
12.6
0.71
14.4
0.81
17.8
1.00
19.2
1.08
21.3
1.20

Foundation flexure
M
(kNm)
9.90
12.50
14.40
17.20
18.50
20.30

Normalisation
factor
0.58
0.73
0.84
1.00
1.08
1.18

Angular
distortion
Normalisation
factor = /e
0.72
0.85
0.91
1.00
1.03
1.07

A plot of the normalisation

factors for Normalised
flexure, stiffness
and angular distortion for
SLOG Parameters
to T=300
the long span case under different
droughts
Longreturn
Span,period
Center
Heave is reproduced in Figure 20.
1.3

Parameter Scale Factor

1.2
1.1
1
Foundation Flexure (M)

0.9

Foundation Stiffness
0.8

Angular Distortion

0.7
0.6
0.5
0.4
0

200

400

600

800

1000

Return Period (Years)

Figure 20

SLOG parameters normalised to T=300 long span, centre heave

These indicate that:SLOG Series 1 Parameters Normalised to T=300

Short Span, Center Heave

70

1.3
1.2

(a)

Scaling the foundation ULS flexural moment calculated for the T=300-year
design drought condition by a factor of 1.2 is conservative in providing
reserve strength for the 1000-year extreme drought condition i.e. actually 1.19;
and

(b)

The corresponding angular distortion occurring in the extreme drought event

will exceed that specified for the normalised T=300-year design threshold by
around 10%. This is not considered critical given the limited potential for this
to occur.

Similar results are obtained when the normalisation is applied to the reverse flexure,
which occurs to foundations subject to the edge heave condition.
It is therefore recommended that conforming designs for Auckland be developed
using conditions for the T=300 design drought, scaled for strength only to
accommodate the extreme drought event, viz 1000-year return period case.
14.3

For foundation performance evaluation, the design basis shown in Figure 21 is

proposed for expansive soil sites.
DESIGN CASE

TIL3,4
AEP 1
300

LIMITING CASE

TIL5
AEP 1
1000

Load Factor
1.2
for strength

SERVICEABILITY
(DAMAGE LIMITATION)

ENHANCED STRENGTH
(DAMAGE LIMITATION
scaled x1.1)

"MILD TO SEVERE
DROUGHT"

"SEVERE DROUGHT"

Figure 21

Design basis for foundations on expansive soils

This reflects, in general terms, the performance framework for wind and/or
earthquake under the New Zealand Building Code referred to in Section 14.1.

71

15.0

FOUNDATION DESIGN DECISION TREE AND FLOW DIAGRAM

15.1

The background data presented in this report provides a basis for a design approach
that is matched to the characteristics of Auckland soils.

15.2

The design process essentially requires designers to address the following two
questions:
(a)

Is the site classified as having expansive soils? (Yes/No) as shown in

Figure 22?

(b)

If the answer to (a) is No, then the foundation design may proceed following
NZS 3604:1999 Sections 3 and 7 rules.
If the answer to (a) is Yes, then the foundation design needs to make
provision for the expansivity of the site for the specified design drought
conditions using AS 2870 rules, with two alternative solution paths possible,
as shown in Figure 23.
Option 1
A specific design is prepared to cover both centre heave and edge heave
ground profile conditions for the 300-year design drought event, with a ULS
load factor of 1.2 included to cover the extreme drought (1000-year return
period) case.
Option 2
A specific design is prepared which incorporates, for example, perimeter
piling to critical depths to cover the edge heave ground profile condition,
plus pre-treatment of the foundations soils to minimise the residual swelling
potential, thereby limiting the potential for the design to be dominated by the
centre heave ground profile condition covering the same drought events as
for Option 1.

15.3

The following flowcharts indicate the process involved:

Foundation Design
for Domestic
Dwelling

No

Is site
classified as
expansive

Foundation Design
to
NZS 3604
(R120 Appendix F)

End

Figure 22

72

General foundation design

Yes

Foundation
Design to
AS 2870
(modified to NZ
conditions)
See
Figure 23

Foundation Design to
AS 2870 (modified to
NZ conditions)

Can expansivity effects (eg

centre heave, edge
heave) be reduced by
specific design ?
(R120 Cl 12)

Yes

Option 1

Option 2
Soil Structure
Interaction analysis to
AS 2870 Appendix F using for
example Mitchell Method based
on :
(a)

(b)

No

Site pre-treatment to ensure

minimum swelling potential (to
control centre heave)
Piling to foundation perimeter
(to control edge heave)

Soil Structure
Interaction analysis to
AS 2870 using for example Mitchell
Method for design drought
(300 year) with ULS load
factor 1.2 (to cover extreme
drought case)
R120 Cl 11.4

R120 Cl 12.4

Figure 23

Option for expansive soil solutions

73

16.0

CONCLUSIONS AND RECOMMENDATIONS

The following conclusions and recommendations are based on the results of the
Stage I and Stage II studies and supersede those presented in the Stage I BRANZ
Study Report 120.

16.1

Conclusions
(a)

AS 2870 procedures:
(i)

In the absence of established data for houses on the soil profile,

AS 2870 requires that the designer estimate the characteristic surface
movement, ys, in order to determine the level of expansivity to be used
in foundation design.

(ii)

Based on the TMI value of +50 for Auckland Airport, it is apparent

that Auckland falls into the wet coastal/alpine climatic category (TMI
>40) defined by AS 2870, for which a depth of moisture change (Hs)
value of 1.5 m is proposed in Tables D1 and D2 of Appendix D of AS
2870 for Melbourne and Victoria and by Fityus et al (1998).

(iii)

Table 2.4 of AS 2870 provides a recommended Hs value of 1.5 m for

Newcastle and Sydney and 1.5 m or 1.8 m for eastern and coastal
Victoria. The value of Hs of 1.5 m for Auckland recommended in
Section 11.3 of this report is equal to the lower end of the range of
1.5 m to 1.8 m.

(iv)

The AS 2870 commentary states that ys should be interpreted as the

characteristic value that has a 5% chance of being exceeded in the life
of the dwelling, which may be taken as 50 years. This corresponds to a
return period of one-in-1000 years or an annual exceedance probability
(AEP) of 0.1%.
It is considered that one-in-300-year return period drought conditions
are applicable to the assessment of expansive soil movements in
residential buildings subject to:

(v)

(a)

Foundations being proportioned with sufficient strength and

stiffness to restrict differential movements within the dwelling
superstructure to limiting values prescribed in AS 2870
Clause 4.4 in the design drought (one-in-300-year return
period) condition; and

(b)

Foundations being provided with sufficient strength to resist

imposed ground deformations i.e. edge heave and/or centre
heave in the more severe extreme drought (one-in-1000-year
return period) condition.

AS 2870 relies on the loading Standard AS/NZS 1170 Part 1 in terms

of:
(a)
(b)

74

Limit state design philosophy being applied;

95% characteristic values of design actions and member

(c)

resistances being used;

Loads, combinations and load factors being consistently
applied;

and assumes that a structural analysis incorporating appropriate soil

structure interaction modelling, e.g. Mitchell method, will be used in
proportioning the foundation for the required stiffness.
The requirements of AS/NZS 1170 Part 1 are generally similar to those
prescribed in the New Zealand limit state Standard NZS 4203:1992
Code of Practice for General Structural Design and Design Loadings
for Buildings and the related materials Standards e.g. NZS 3101 Code
of Practice for the Design of Concrete Structures.
(vi)

(b)

(c)

For the purposes of satisfying the foundation structural performance

requirements of (a)(iv) above, the following load factors may be
applied to design strengths derived from the one-in-300-year return
period design drought condition:
(a)

Design drought (one-in-300-year) = 1.0;

(b)

Extreme drought (one-in-1000-year) = 1.2.

Classification of summer soil moisture conditions

(i)

The CDD value, defined as the SMD below a threshold value of -70
mm (representative of dry soil conditions) accumulated on a daily
basis over the period leading up to the sampling date, provides a
method of quantifying the intensity of summer conditions for any
particular sampling date.

(ii)

The CDD approach for determination of the intensity of summer

conditions provides a means for extrapolating the extensometer data to
predict the ground surface movement under the design drought
conditions.

(iii)

Gumbel Analysis (Type 1 Extreme Value Theory) has been adopted as

a means of estimating and modelling the statistical distribution of dry
periods, represented by the CDD data adopted for this study.

(iv)

Based on the Gumbel Analyses, the summer periods of 2005 and 2006
during the Stage II study period are found to correspond to less than a
one-in-five-year return period event for Auckland Airport. The
summer periods of 2005 and 2006 approximately correspond to the
one-in-20-year return period event for the Albany and Pukekohe
weather stations.

Results of field and laboratory investigations for this study are:

(i)

The Atterberg Limits for the soils from Sites 2A to 2F plot slightly
above or below the A line and are of high plasticity, with liquid limit
values ranging from 50 to 115. The linear shrinkage values of the soils
range from 11% to 25%.
75

(d)

(ii)

The shrinkage index for the soil samples from Sites 2A to 2F range
from 1.2% to 4.8% with a mean value of 2.74%. These shrinkage index
values are comparable to those reported by Fityus et al (1998) for soils
in the Newcastle/Hunter Valley region and by Coffey and Partners
(1985) for the Sydney region.

(iii)

There is a weak correlation between the shrinkage index and Atterberg

Limits or linear shrinkage for the soils from Sites 2A to 2F. In
particular, a weak correlation is apparent between the shrinkage index
and linear shrinkage or liquid limit, which indicates that the linear
shrinkage and liquid limit values are only indicative of soil
expansivity.

(iv)

The weak correlation between the shrinkage index and the linear
shrinkage or liquid limit is not unexpected, given that the shrinkage
index relates to undisturbed samples while the other two
classification parameters relate to fully remoulded samples.

(v)

Because of the higher confidence in the measured ys values obtained

from the extensometers as compared with the theoretical ys values
obtained from the soil suction test data, it is considered more
appropriate, on the basis of the data obtained during the Stage II
investigation, to adopt the measured ys values for back analysis to
obtain the change in soil suction value at the soil surface (us) for use
in the extrapolation to the design drought conditions for the Auckland
region.

(vi)

On the basis of the data collection and analyses carried out for the
Stage II Study Report discussed herein, it is considered that there is
sufficient evidence to support a recommendation that soil suction
Profile Alpha derived in the Stage I report be abandoned and soil
suction Profile Gamma be adopted for determining the expansive soil
characteristics of soils in the Auckland region, based on a notional
300-year return period drought event i.e. that a triangular suction
change profile, having an Hs of 1.5 m and a us (suction change at the
ground surface) of 1.2 pF, can be adopted.

(vii)

If the recommended design suction change Profile Gamma is adopted,

Sites 2A to 2F generally fall into or close to the M classification of
AS 2870.

In adopting the above-mentioned soil suction Profile Gamma (as

Section 17(c)(vi)), it is noted that the Stage II data extrapolation to an extreme
drought (1000-year return period) event justifies the use of an average u =
1.44 pF in lieu of the u = 1.2 pF proposed for the corresponding ys
calculation. To address this, the following approach to determining the
foundation performance for the extreme drought (1000-year return period)
event prescribed by AS 2870 is proposed:
(i)

76

By applying a scaling factor of 1.2 to the ys value calculated from soil

Profile Gamma for the purposes of soil foundation interaction analysis
e.g. Mitchell Method; or

(ii)

(e)

(f)

By calculating the flexural demand (Mu) for foundation elements

calculated from Profile Gamma (Hs = 1.5, u = 1.2) to incorporate a
universal load factor of 1.2 rather than 1.0.

The maximum soil suction change between the winter 2004 measurements and
either the summer 2005 or 2006 measurements, normalised to a base of 3.2 pF,
ranged from 0.4 pF to 0.85 pF. The following became apparent:
(i)

A weak relationship was apparent between the soil suction change at

the soil surface ( us) and the CDD value.

(ii)

No discernible relationship was apparent between the soil suction

change at the soil surface ( us) and the vertical ground movement
measured by the extensometers.

(iii)

The sampling and measurement of soil suction using the transistor

psychrometer requires careful execution by a skilled and experienced
technician, and has a high potential for the introduction of variability to
the calculation of ys values for any particular site.

Applicability of AS 2870 to Auckland

(i)

The method of site classification in terms of soil expansivity, involving

estimation of the characteristic surface movement (ys) in accordance
with AS 2870, appears to be applicable to Auckland sites.

(ii)

The soil shrinkage index (Ips) is able to be determined by means of the

core shrinkage test method given in AS 1289 Test Method 7.1.3 and
referred to in AS 2870.

(iii)

Weak correlation was apparent between Ips and other soil classification
tests, such as liquid limit, plasticity index and linear shrinkage.

(iv)

To calculate the characteristic surface movement (ys) for a site, the

instability index (Ipt) requires to be calculated, which in turn requires
the depth of the cracked zone (the zone in which predominantly
vertical shrinkage cracks exist seasonally) to be estimated.
Investigation of the depth of the cracked zone was beyond the scope of
the study reported herein. However, since the value of Ipt is sensitive to
the estimated depth of the cracked zone, as indicated by the equations
given in Section 5.4.2, it follows that ys is also sensitive to the
estimated depth of the cracked zone.

(v)

Investigation into the depth of the seasonal cracked zone in Auckland

soils is required to provide design values for determination of the
instability index (Ipt) and to enable reliable estimates of ys to be made.

(vi)

As set out in Appendix D to this report, the following can be concluded

with respect to the applicability of foundation details in NZS 3604, for
use on Auckland sites:
(1)

Type 1 foundations are unsuitable for construction on

77

expansive soil sites in the Auckland region without

modification.

(g)

78

(2)

Foundations where the floor slab is fully integrated with the

concrete masonry blockwork foundation wall to provide
structural performance similar to that of Type 2A foundations
also require modification from the details provided in
NZS 3604.

(3)

NZS 3604 appears to be suitable for construction on Class M

expansive soils sites in the Auckland region.

(4)

Foundation Type 2C requires no modification from the

detailing specified in NZS 3604 to be suitable for construction
on Class M expansive soils sites in the Auckland region where
the beam depth is at least 750 mm. For reduced beam depth
additional steel reinforcement will be required, as indicated in
Table 26 for Foundation Type 2B.

(5)

The analyses of the AS 2870 standard designs of stiffened raft

and waffle slab construction types using specialist (SLOG)
software indicates that these designs are generally suitable for
single-storey construction.

(6)

Section 2.2 of AS 2870 permits consideration of 10 years (or

more) of satisfactory field performance as providing guidance
for the design of foundations on expansive soils.

(7)

The angular distortions which occur in essentially brittle

cladding, e.g. masonry veneer, can be used to derive
permissible ys values at particular sites where particular
damage levels have been observed in that cladding.

(8)

Back analyses undertaken using the Type 2A foundations, from

Table 28 and Figure 25, have confirmed that Type 2A
foundations with an embedment depth of 450 mm supporting
clad frame, articulated masonry veneer and masonry veneer
buildings, perform satisfactorily in Auckland.

Design/interpretation procedure recommended for Auckland until such time as

the results of further studies are available:
(i)

That the soil expansivity of any particular site be determined by taking

core samples at, say, 0.5 m to 0.8 m depth and 1 m to 1.3m depth
below the ground surface and measuring the soil shrinkage index (Ips)
by means of either the shrink-swell index or core shrinkage index
methods given in AS 1289 Test Methods 7.1.1 or 7.1.3.

(ii)

That the instability index (Ipt) be calculated using the equations given
in Section 5.4.2, which are taken from Appendix F of AS 2870, and
assuming a cracked zone depth equivalent to 0.5 Hs.

(iii)

That the characteristic surface movement (ys) be calculated using the

equation given in the Section 5.2, which is taken from AS 2870, and on
the basis of a triangular suction profile, having an Hs value of 1.5 m
and us at the ground surface of 1.2 pF, and an Ipt value calculated as
discussed in (e)(ii) above.
(iv)

(h)

(i)

That the site classification be determined on the basis of ys, using

Table 2.3 of AS 2870.

Limitations on research
(i)

As the study period did not include a severe drought period, it was not
possible to measure the ground movement and soil suction profiles at
Sites 2A to 2F that relate to such dry weather conditions, in order to
determine the corresponding soil suction change profile that is required
for calculation of the characteristics surface movement (ys).

(ii)

It was therefore necessary to extrapolate the actual soil suction profile

for Sites 2A to 2F corresponding to the driest of the 2005 and 2006
summers in order to derive a suggested design soil suction change
profile for the Auckland region. The extrapolation was based on the
assumption that there is a correlation between the CDD values and
ground surface movement (ys), as discussed in Section 10.0.

Soil suction change profile for regions other than Auckland

The soil suction change profile recommended in this report is specific to the
Auckland region.
The soil suction change profile for any other region in New Zealand depends
on the SMD profile for the particular region, which in turn depends on the
local climate (plus any change to that local climate in the future).

16.2

Recommendations
(a)

To develop the understanding of expansive soils within the Auckland region

the following further work is recommended:
(i)

Building damage survey, as discussed in Section 2.5, to ascertain the

actual foundation performance of a prescribed number of dwellings
within Manukau City, which were constructed with concrete ground
bearing slabs and perimeter footings complying with Appendix E of
NZS 3604:1990 Code of Practice for Light Timber Framed Buildings
Not Requiring Specific Design.
The number of dwellings would be dependent on the availability of
suitable building developments within representative soil type zones.
The recommended building damage survey would allow site
classification on the basis of the method outlined in Section 2.2.2 of
AS 2870, for comparison with site classification based on the
alternative method, involving estimation of the characteristic surface
movement (ys) as outlined in Section 2.2.3 of AS 2870.

79

The output of the building damage survey would be a series of figures

similar to Figure 25 of this report for a range of building sizes to be
used by designers and local TAs to determine the suitability of
NZS 3604 specified footings or the need to use AS 2870 specified
standard or specific designs.
(ii)

(b)

Geotechnical investigation to determine the soil profiles, soil

properties and seasonal soil suction profiles at each of the sites selected
for the foregoing building damage survey.

Additional work would be required at Sites 2A to 2F, over a number of years,

to resolve limitations in the work carried out for this study:
(i)

Continue to monitor the SMD on the NIWA Climate Now website

during summer periods in order to determine whether a severe drought
event occurs i.e. conditions similar to, or approaching, the 1998 event.
If such a drought event occurs, measure the ground movement at Sites
2A to 2F by means of the extensometers and take soil suction samples
to determine the soil suction profile for each site.

(ii)

If such drought conditions occur, review the correlation between the

ground movement and CDD values, including the values relating to the
severe drought, and verify the design us value.

(iii)

If Recommendations 16.2(b)(i) and (ii) are adopted, undertake

monitoring of the extensometers at Sites 2A to 2F on a regular basis
(say twice per year) to confirm that the extensometers are continuing to
function satisfactorily.

(iv)

Excavation of test pits at selected sites at the end of a representative

dry summer period, in order to determine the depth of seasonal
shrinkage cracks, is required to determine the instability index (Ipt).

(v)

It is recommended, during the course of an identified future dry period,

that Fraser Thomas Ltd be engaged to undertake monitoring of the
extensometers to identify the lag between dry conditions and the
response in the soil column.

(c)

If further research is undertaken into the expansivity of soils in the Auckland

region, it is recommended that consideration be given to installing and
monitoring a set of extensometers adjacent to a weather recording station to
help remove some of the foregoing uncertainties.

(d)

Length of time to develop database

Given that AS 2870 refers to the validity of a 10-year experience record to
evaluate the performance of existing structures, it is recommended that the
length of time to develop the database should record results for at least five
years before the preliminary recommendations of this report are modified, but
otherwise should not be less than 10 years.

(e)

80

Climate change

It is recommended that the SMD profile for Auckland be re-evaluated at not

less than 20-year intervals, using the methodology adopted in this report, in
order to take account of any future climate change.
(f)

Proposed form of reporting by geotechnical practitioners

The authors recommend that every residential land subdivision site on soils
which are potentially expansive should be subject to a geotechnical
investigation report in support of a sub-division consent, which provides for
the reactivity classification of the development site as follows:
(i)

Site classification in accordance with the processes set out in AS 2870,

and reporting of the parameters on which this classification is made.

(ii)

The design characteristic surface movement, ys, value calculated using

soil suction Profile Gamma, as discussed in Section 16.1(e).

(iii)

Recommendations for reassessments of the site classification in the

geotechnical completion report on the completion of land development
and at the building consent stage to reflect any variation in
classification arising from land development earthworks or building
site works.

(iv)

Recommendations on foundation types and embedment depth for

particular building construction types.

These requirements will provide the building designer with the information
required to design the buildings foundations and provide the local TAs with
the information upon which to build a database for their region.
(g)

Applicants for building consents on such sites be advised of the importance of

ongoing foundation maintenance as set out in Appendix B of this report in
order to ensure owner obligations become more widely known.

81

Appendix A
Bibliography

BIBLIOGRAPHY OF LITERATURE STUDY

Aitchison GD (Ed). 1965. Moisture Equilibria and Moisture Changes in Soils Beneath
Covered Areas. Butterworths, Sydney.
American Association of State Highway and Transportation Officials (AASHTO). Standard
Method of Test for Determining Expansive Soils. AASHTO Designation: T 258-81.
American Association of State Highway and Transportation Officials (AASHTO). 1996.
Standard Method of Test for Soil Suction. AASHTO Designation: T 273-86.
American Society for Testing and Materials. 1995. Standard Method for Shrinkage Factors
of Soils by the Wax Method. ASTM Designation: D 4943-95.
Barthur R, Jaksa MB and Mitchell PW. 1996. Design of Residential Buildings Built on
Expansive Soil Using Probabilistic Methods. Proceedings 7th Australian/New Zealand
Conference on Geomechanics 369374.
Beca Carter. 1980. Report on Onerahi Chaos Breccia for Rodney County Council. Report
No. 130149/3.
Bell FG and de Bruyn IA. 1997. Sensitive, Expansive, Dispersive and Collapsive Soils.
Bull. Int. Assoc. Engineering Geology 56: 1938.
Cameron DA. 1989. Tests for Reactivity and Prediction of Ground Movement. Civil
Engineering Trans. Institute of Engineers Aust 121132.
Coffey & Partners Pty Ltd. 1985. Sydney Swelling Soils Study: Analysis of Data. Report to
the Building Licensing Board, Report No. S7032/2-AD.
Crawford S. 1999. Shrinkage of Auckland North Shore Clays (Comment). SESOC Journal
12(2): 78.
CSIRO. 1996. Guide to Homeowners on Foundation Maintenance and Footing
Performance. Information Sheet No. 10-91. CSIRO Publishing, Collingwood.
Delaney MG, Allman MA and Sloan SW. 1996. A Network of Field Sites for Reactive Soil
Monitoring in the Newcastle-Hunter Region. In Jaksa MB, Kaggwa WS and Cameron DA
(Eds). Proceedings 7th Australian/New Zealand Conference on Geomechanics 381387.
Department of Building and Housing. 2007. Building for the 21st Century: Review of the
Building Code. DBH, Wellington.
El-Garhy BM, Youssef AA and Wray WK. 2000. Using Soil Diffusion to Design Slab-OnGround Foundations on Expansive Soils. In Toll DG and Leong EC (Eds). Unsaturated Soils
for Asia Proceedings of the Asian Conference on Unsaturated Soils 223229.
Farquhar G. 1999. Shrinkage of Auckland North Shore Clays (Comment). SESOC Journal
12(2): 11.
Fityus SG. 1999. A Soil Moisture Based Method of Estimating ys. Proceedings 8th
Australian/New Zealand Conference on Geomechanics 2: 809815.

A1

Fityus SG. 1996. The Effect of Initial Moisture Content and Remoulding on the ShrinkSwell Index, Iss. In Jaksa MB, Kaggwa WS and Cameron DA (Eds). Proceedings 7th
Australian/New Zealand Conference on Geomechanics 388393. Valley. Conference
Publications, Springwood, Australia. 251265.
Fityus SG, Walsh PF and Kleeman PW. 1998. The Influence of Climate as Expressed by the
Thornthwaite Index on the Design Depth of Moisture Change of Clay Soils in the Hunter
Valley. Conference on Geotechnical Engineering and Engineering Geology in the Hunter
Valley. Conference Publications, Springwood, Australia 251265.
Fityus SG and Welbourne JC. 1996. Trends in Shrink-Swell Test Results in the Newcastle
Region. In Jaksa MB, Kaggwa WS and Cameron DA (Eds). Proceedings 7th Australian/New
Zealand Conference on Geomechanics 394399.
Fox E. 2000. A Climate-Based Design Depth of Moisture Change Map of Queensland and
the Use of Such Maps to Classify Sites Under AS 2870:1996. Australian Geomechanics
35(4): 5360
Francis CA. 1999. Shrinkage of Auckland North Shore Clays. SESOC Journal 12(1): 46
48.
Fredlund DG and Rahardjo H. 1993. Soil Mechanics for Unsaturated Soils. J Wiley & Sons
Ltd, New York. 517 pp.
Freeman TJ, Littlejohn GS and Driscoll RMC. 1994. Has Your House Got Cracks?: A Guide
to Subsidence and Heave of Buildings on Clay. The Cromwell Press, London. 114 pp.
Grant-Murray J. 1999. Shrinkage of Auckland North Shore Clays (Comment). SESOC
Journal 12(2): 79.
Grayson BJ. 2000. Expansive Clays: Translating AS 2870 Into NZS 3604. NZ
Geomechanics News 59: 5256.
Harvey CC, Riley PB and Pickens GA. 1982. Problems Associated with the Shrinkage of
Auckland Clays. NZ Geomechanics News 25: 3846.
Holden JC.1996. Some Developments in Urban Root Barriers. In Jaksa MB, Kaggwa WS
and Cameron DA (Eds). Proceedings 7th Australian/New Zealand Conference on
Geomechanics 406411.
Holland JE and Richards J. 1984. The Practical Design of Foundations for Light Structures
on Expansive Clays. Proceedings 5th International Conference on Expansive Clays,
Adelaide 154158.
Holland JE. 1981. The Design, Performance and Repair of Housing Foundations. Swinburne
Ltd, Melbourne. 72 pp.
Hung VQ and Fredlund DG. 2000. Volume Change Predictions in Expansive Soils Using a
Two-Dimensional Finite-Element Method. In Toll DG and Leong EC (Eds). Unsaturated
Soils for Asia Proceedings of the Asian Conference on Unsaturated Soils 231237.

A2

Li J, Cameron DA and Mills KG. 1996. Numerical Modelling of Covers and Slabs Subject
to Seasonal Suction Variations. In Jaksa MB, Kaggwa WS and Cameron DA (Eds).
Proceedings 7th Australian/New Zealand Conference on Geomechanics 424429.
Marsh ET and Thoeny SA. 1999. Damage and Distortion Criteria for Residential Slab-OnGrade Structures. Journal Performance of Constructed Facilities 13(2): 122127.
Mitchell PW. 2003. Peer Review Comments.
Mitchell PW. 1984. A Simple Method of Design of Shallow Footings on Expansive Soil.
Proceedings 5th International Conference on Expansive Clays, Adelaide 159164.
Mitchell PW and Avalle DL. 1984. A Technique to Predict Expansive Soil Movements.
Proceedings 5th International Conference on Expansive Clays, Adelaide 124130.
Pender MJ. 2001. The Effect of Fissuring in Auckland Residual Clays on the Capacity of
Shallow Foundations. SESOC Journal 14(2): 3541 (September).
Poulos HG. 1984. Parametric Solutions for Strip Footings on Swelling and Shrinking Soils.
Proceedings 5th International Conference on Expansive Clays, Adelaide 149153. Mining
Engineering Res. Rep. 456. 49 pp.
Poulos HG. 1983. Analysis of Strip Footings on Expansive Soils. Univ. of Sydney Sch. Civil
and
Smith RL. 1993. Estimating Soil Movements in New Areas. From a seminar Extending the
Code Beyond Residential Slabs and Footings. The Institution of Engineers, Australia.
Standards Association of New Zealand. 1999. NZS 3604 Timber Framed Buildings. SNZ,
Wellington.
Standards Association of New Zealand. 1995. NZS 3101 Code of Practice for the Design of
Concrete Structures. SNZ, Wellington.
Standards Association of New Zealand. 1992. NZS 4203 Code of Practice for General
Structural Design and Design Loadings for Buildings. SNZ, Wellington.
Standards Association of New Zealand. 1990. NZS 3604 Code of Practice for Light Timber
Frame Buildings Not Requiring Specific Design. SNZ, Wellington.
Standards Association of New Zealand. 1987. NZS 4402 Methods of Testing Soils for Civil
Engineering Purposes. SNZ, Wellington.
Standards Australia. 1998. AS 1289.7.1.1. Methods for Testing Soils for Engineering
Purposes: Method 7.1.1: Soil Reactivity Tests. Determination of the Shrinkage Limit of a Soil
Shrink-Swell Index.
Standards Australia. 1998. AS 1289.7.1.2 Methods for Testing Soils for Engineering
Purposes: Method 7.1.2: Soil Reactivity Tests. Determination of the Shrinkage Limit of a Soil
Loaded Shrinkage Index.

A3

Standards Australia. 1998. AS 1289.7.1.3 Methods for Testing Soils for Engineering
Purposes: Method 7.1.3: Soil Reactivity Tests. Determination of the Shrinkage Limit of a Soil
Core Shrinkage Index.
Standards Australia. 1998. AS 1289.2.2.1. Methods for Testing Soils for Engineering
Purposes: Method 2.2.1: Soil Moisture Content Tests. Determination of the Total Suction of a
Soil Standard Method.
Standards Australia. 1997. Handbook HB 28:1997 The Design of Residential Slabs and
Footings. By Walsh P and Cameron D.
Standards Australia. 1996. AS 2870 Residential Slabs and Footings Construction.
Standards Australia. 1996. AS 2870 Supplement 1 Residential Slabs and Footings
Construction Commentary.
Standards Australia. 1993. AS 1726 Geotechnical Site Investigations.
Stapleton M. 1999. Shrinkage of Auckland North Shore Clays (Comment). SESOC Journal
12(2): 76.
Van der Woude F. 2000. User Manual Design of Stiffened Slabs on Reactive Soils.
Available online through RAFT Software.
Van der Woude F. 1999. Radical Re-examination of Stiffened Raft Footing Technology.
Proceedings 8th Australian/New Zealand Conference on Geomechanics. Hobart, 1719
February.
Walsh KD, Brashford HH and Mason BCA. 2001. State of Practice of Residential Floor Slab
Flatness. ASCE Journal of Performance of Constructed Facilities 15(4).
Walsh PF and Walsh SF. 1987. Structure/Reactive-Clay Model for a Microcomputer.
CSIRO Division of Building Research Report 86/9. 32 pp.
Woodburn JA. 2003. Personal Communications.
Woodburn JA. 2002. Manual for the 8-Probe Transistor Psychrometer. Soil Mechanics
Instrumentation, Adelaide. 18 pp.
Woodburn JA, Holden JC and Peter P. 1993. The Transistor Psychrometer: A New
Instrument for Measuring Soil Suction. Am. Soc. Civil Engineers Ann. Convention and
Exposition, Dallas, Texas.
Woodburn JA and Lucas B. 1995. New Approaches to the Laboratory and Field
Measurement of Soil Suction. In Alonso EE and Delange P (Eds). Unsaturated Soils
Proceedings 1st International Conference on Unsaturated Soils 667671.
Wray WK (Ed). 1995. So Your Home is Built on Expansive Soils A Discussion of How
Expansive Soils Affect Buildings. American Society of Civil Engineers, Virginia. 59 pp.

A4

Appendix B
Borehole Logs

B1

B2

B3

B4

B5

B6

B7

Appendix C
Site Summary Sheets

C1

C2

4.0 EXTENSOMETER READINGS - SITE 2A: MANUREWA

Date of Sampling
Deficit Value at Reading (mm)
Cumulative Deficit Days
Surface Monument at 0.0 mbgl
Surface Monument at 0.2 mbgl
Rod at 0.5 mbgl
Rod at 1.0 mbgl
Rod at 1.5 mbgl
Rod at 2.0 mbgl
Rod at 2.5 mbgl

Summer 04/05
28.01.05 17.03.05 28.04.05
-76.3
-81.3
-72.2
15.6
216.3
384.2
-2
-7
-3
0
-12
-7
-1
-11
-9
-1
-6
-6
0
-1
-2
0
0
-1
0
0
-1

Summer 05/06
24.02.06 25.03.06
-85.4
-77.4
252.1
576.7
-3
-6
-7
-16
-6
-16
-3
-11
0
-5
0
-1
1
0

NOTES: 1. Extensometer readings in mm compared to Winter 05 readings

2. NR = no reading due to damage of surface monument
3. Cumulative Deficit Days = sum of values below -70 mm threshold

Movement (mm)

Surface Monument Average

y = -0.014x - 2.141
R = 0.562

-20
-15
-10
-5
0
0

200

400

600

800

Cumulative Deficit Days

y = -0.024x - 1.649
R = 0.800

Movement (mm)

0.5m Extensometer
-20
-15
-10
-5
0
0

100

200

300

400

500

600

700

Cumulative Deficit Days

C3

C4

C5

4.0 EXTENSOMETER READINGS - SITE 2B: OTARA

Date of Sampling
Deficit Value at Reading (mm)
Cumulative Deficit Days
Surface Monument at 0.0 mbgl
Surface Monument at 0.2 mbgl
Rod at 0.5 mbgl
Rod at 1.0 mbgl
Rod at 1.5 mbgl
Rod at 2.0 mbgl
Rod at 2.5 mbgl

Summer 04/05
28.01.05 17.03.05 28.04.05
-76.3
-81.3
-72.2
15.6
216.3
384.2
4
-6
-5
-3
-8
-6
-1
-5
-3
-2
-2
-1
-1
-1
0
-1
-1
0
-1
-2
0

Summer 05/06
24.02.06 25.03.06
-85.4
-77.4
252.1
576.7
-13
-17
-10
-12
-5
-7
-2
-2
-1
-1
-1
-1
-1
-1

NOTES: 1. Extensometer readings in mm compared to Winter 05 readings

2. NR = no reading due to damage of surface monument
3. Cumulative Deficit Days = sum of values below -70 mm threshold

Movement (mm)

Surface Monument Average

y = -0.022x - 1.094
R = 0.659

-20
-15
-10
-5
0
0

200

400

600

800

Cumulative Deficit Days

0.5m Extensometer

y = -0.008x - 1.733
R = 0.606

Movement (mm)

-20
-15
-10
-5
0
0

200

400

Cumulative Deficit Days

C6

600

800

C7

C8

4.0 EXTENSOMETER READINGS - SITE 2C: HOWICK

Date of Sampling
Deficit Value at Reading (mm)
Cumulative Deficit Days
Surface Monument at 0.0 mbgl
Surface Monument at 0.2 mbgl
Rod at 0.5 mbgl
Rod at 1.0 mbgl
Rod at 1.5 mbgl
Rod at 2.0 mbgl
Rod at 2.5 mbgl

Summer 04/05
01.02.05 17.03.05 28.04.05
-76.3
-81.3
-72.2
15.6
216.3
384.2
-8
-12
-2
NR
NR
-2
-2
-3
-1
-1
0
-1
0
1
-1
0
1
0
0
1
1

Summer 05/06
24.02.06 25.03.06
-85.4
-77.4
252.1
576.7
-21
-28
-10
-14
-2
-4
-2
-1
0
0
0
0
0
0

NOTES: 1. Extensometer readings in mm compared to Winter 05 readings

2. NR = no reading due to damage of surface monument
3. Cumulative Deficit Days = sum of values below -70 mm threshold

Surface Monument Average

y = -0.015x - 7.359
R = 0.187

Movement (mm)

-30
-20
-10
0
0

200

400

600

800

Cumulative Deficit Days

Movement (mm)

0.5m Extensometer

y = -0.002x - 1.720
R = 0.184

-20
-15
-10
-5
0
0

200

400

600

800

Cumulative Deficit Days

C9

C10

C11

4.0 EXTENSOMETER READINGS - SITE 2D: MAIRANGI BAY

Date of Sampling
Deficit Value at Reading (mm)
Cumulative Deficit Days
Surface Monument at 0.0 mbgl
Surface Monument at 0.2 mbgl
Rod at 0.5 mbgl
Rod at 1.0 mbgl
Rod at 1.5 mbgl
Rod at 2.0 mbgl
Rod at 2.5 mbgl

Summer 04/05
01.02.05 17.03.05 29.04.05
-76.3
-81.3
-73.0
87.3
334.6
1013.4
0
0
-1
-1
0
0
-1
0
0
0
0
0
-3
-1
-3
-1
1
-1
0
1
-1

Summer 05/06
27.02.06 24.03.06
-87.7
-78.0
323.6
560.5
-2
1
-2
0
-1
-1
0
1
-3
-2
-1
0
-2
1

NOTES: 1. Extensometer readings in mm compared to Winter 05 readings

2. NR = no reading due to damage of surface monument
3. Cumulative Deficit Days = sum of values below -70 mm threshold

Surface Monument Average

y = 0.000x - 0.729
R = 0.034

Movement (mm)

-15.0
-10.0
-5.0
0.0
5.0
0

200

400

600

800

1000

1200

Cumulative Deficit Days

Movement (mm)

0.5m Extensometer
y = 0.000x - 0.998
R = 0.300

-15.0
-10.0
-5.0
0.0
0

200

400

600

Cumulative Deficit Days

C12

800

1000

1200

C13

C14

4.0 EXTENSOMETER READINGS - SITE 2E: RED BEACH

Date of Sampling
Deficit Value at Reading (mm)
Cumulative Deficit Days
Surface Monument at 0.0 mbgl
Surface Monument at 0.2 mbgl
Rod at 0.5 mbgl
Rod at 1.0 mbgl
Rod at 1.5 mbgl
Rod at 2.0 mbgl
Rod at 2.5 mbgl

Summer 04/05
01.02.05 17.03.05 29.04.05
-76.3
-85.6
-73.0
87.3
334.6
1013.4
-1
-9
-14
NR
-14
-20
1
-1
-5
0
-1
0
0
-1
0
0
-1
0
0
-2
-1

Summer 05/06
27.02.06 24.03.06
-87.7
-78.0
323.6
560.5
-16
-15
-22
-22
-7
-8
-1
-1
0
1
-1
0
-1
-1

NOTES: 1. Extensometer readings in mm compared to Winter 05 readings

2. NR = no reading due to damage of surface monument
3. Cumulative Deficit Days = sum of values below -70 mm threshold
y = -0.013x - 7.144
R = 0.390

Surface Monument Average

Movement (mm)

-25
-20
-15
-10
-5
0
0

200

400

600

800

1000

1200

Cumulative Deficit Days

0.5m Extensometer

y = -0.005x - 1.36
R = 0.264

Movement (mm)

-15
-10
-5
0
0

200

400

600

800

1000

1200

Cumulative Deficit Days

C15

C16

C17

4.0 EXTENSOMETER READINGS - SITE 2F: PUKEKOHE

Date of Sampling
Deficit Value at Reading (mm)
Cumulative Deficit Days
Surface Monument at 0.0 mbgl
Surface Monument at 0.2 mbgl
Rod at 0.5 mbgl
Rod at 1.0 mbgl
Rod at 1.5 mbgl
Rod at 2.0 mbgl
Rod at 2.5 mbgl

Summer 04/05
02.02.05 17.03.05 28.04.05
-76.3
-81.3
-72.2
63.3
215.9
308.3
NR
-4
-2
1
0
0
1
0
0
2
0
1
1
0
0
1
0
0
2
1
1

Summer 05/06
28.02.06 25.03.06
-88.2
-77.4
265.3
661.5
-6
-7
-1
-2
0
-1
0
-1
-1
-1
0
0
1
0

NOTES: 1. Extensometer readings in mm compared to Winter 05 readings

2. NR = no reading due to damage of surface monument
3. Cumulative Deficit Days = sum of values below -70 mm threshold

Movement (mm)

Surface Monument Average

-15
-10
-5
0

y = -0.008x + 0.412
R = 0.668

200

400

600

Cumulative Deficit Days

800

0.5m Extensometer
Movement (mm)

-15
-10
-5

y = -0.001x + 0.682
R = 0.233

0
5
0

200

400

600

Cumulative Deficit Days

C18

800

Appendix D
Cumulative Deficit Data
and Gumbel Analyses

AIRPORT CUMULATIVE DEFICITY DAYS DATA AND GUMBEL ANALYSES

Year

CDD

Ordered by CDD

Rank

Pi

Xi

1963

117.7

1974

1376.3

38

0.97

3.65

1964

428.6

1978

1263.0

37

0.95

2.94

1965

0.0

1973

1180.7

36

0.92

2.53

1966

0.0

1998

1164.0

35

0.90

2.22

1967

0.0

1970

826.5

34

0.87

1.99

1968

158.1

1983

716.2

33

0.85

1.79

1969

26.6

1981

693.5

32

0.82

1.62

1970

826.5

1994

661.0

31

0.79

1.47

1971

447.8

2006

654.0

30

0.77

1.34

1972

0.0

1993

636.4

29

0.74

1.22

1973

1180.7

1982

608.3

28

0.72

1.10

1974

1376.3

1991

559.3

27

0.69

1.00

1975

137.3

1979

498.6

26

0.67

0.90

1976

306.2

1971

447.8

25

0.64

0.81

1977

297.1

1987

441.8

24

0.62

0.72

1978

1263.0

1964

428.6

23

0.59

0.64

1979

498.6

2005

394.0

22

0.56

0.56

1980

0.0

1990

377.3

21

0.54

0.48

1981

693.5

1995

334.5

20

0.51

0.40

1982

608.3

1976

306.2

19

0.49

0.33

1983

716.2

1997

302.7

18

0.46

0.26

1984

0.0

1977

297.1

17

0.44

0.19

1985

22.8

1999

271.6

16

0.41

0.12

1986

0.6

1989

244.3

15

0.38

0.05

1987

441.8

2001

195.5

14

0.36

-0.02

1988

180.6

2000

186.5

13

0.33

-0.09

1989

244.3

1988

180.6

12

0.31

-0.16

1990

377.3

1968

158.1

11

0.28

-0.24

1991

559.3

1996

153.2

10

0.26

-0.31

1992

99.1

1975

137.3

0.23

-0.38

1993

636.4

1963

117.7

0.21

-0.46

1994

661.0

1992

99.1

0.18

-0.54

1995

334.5

2004

80.1

0.15

-0.63

1996

153.2

2003

37.4

0.13

-0.72

1997

302.7

1969

26.6

0.10

-0.82

1998

1164.0

1985

22.8

0.08

-0.94

1999

271.6

2002

15.7

0.05

-1.09

2000

186.5

1986

0.6

0.03

-1.30

2001

195.5

1965

0.0

2002

15.7

1966

0.0

2003

37.4

1967

0.0

2004

80.1

1972

0.0

2005

394.0

1980

0.0

2006

654.0

1984

0.0

D1

Cumulative Deficit Days

Airport Gumbel Analysis

1600.0
1400.0
1200.0
1000.0
800.0
600.0
400.0
200.0
0.0

y = 311.04x + 254.85
R2 = 0.9688

-2.00

0.00

2.00
Xi

D2

4.00

ALBANY CUMULATIVE DEFICIT DAYS DATA AND GUMBEL ANAYLSES

Year

CDD

Rank

Pi

Xi

1967

0.0

Ordered by CDD
2005

1053.2

37

0.97

3.62

1968

6.6

1974

917.8

36

0.95

2.92

1969

80.0

1973

896.0

35

0.92

2.50

1970

728.7

1978

882.7

34

0.89

2.20

1971

214.1

1994

760.7

33

0.87

1.96

1972

153.8

1970

728.7

32

0.84

1.76

1973

896.0

1998

701.6

31

0.82

1.59

1974

917.8

2000

522.4

30

0.79

1.44

1975

15.3

1977

518.8

29

0.76

1.31

1976

489.8

1983

494.1

28

0.74

1.19

1977

518.8

1976

489.8

27

0.71

1.07

1978

882.7

1990

482.8

26

0.68

0.97

1979

330.0

2006

481.7

25

0.66

0.87

1980

0.0

1987

473.3

24

0.63

0.78

1981

306.2

1995

436.1

23

0.61

0.69

1982

156.5

1993

430.9

22

0.58

0.60

1983

494.1

1979

330.0

21

0.55

0.52

1984

0.0

1991

328.9

20

0.53

0.44

1985

2.6

1981

306.2

19

0.50

0.37

1986

7.6

1989

239.4

18

0.47

0.29

1987

473.3

1971

214.1

17

0.45

0.22

1988

99.8

1999

198.1

16

0.42

0.15

1989

239.4

1982

156.5

15

0.39

0.07

1990

482.8

1972

153.8

14

0.37

0.00

1991

328.9

2003

129.2

13

0.34

-0.07

1992

124.2

1992

124.2

12

0.32

-0.14

1993

430.9

2004

102.3

11

0.29

-0.21

1994

760.7

1988

99.8

10

0.26

-0.29

1995

436.1

1969

80.0

0.24

-0.36

1996

2.3

1997

72.9

0.21

-0.44

1997

72.9

2001

28.5

0.18

-0.53

1998

701.6

1975

15.3

0.16

-0.61

1999

198.1

1986

7.6

0.13

-0.71

2000

522.4

1968

6.6

0.11

-0.81

2001

28.5

1985

2.6

0.08

-0.93

2002

1.2

1996

2.3

0.05

-1.08

2003

129.2

2002

1.2

0.03

-1.29

2004

102.3

1984

0.0

2005

1053.2

1980

0.0

2006

481.7

1967

0.0

D3

Albany Gumbel Analysis

Cumulative Deficit Days

1400.0
1200.0

y = 259.72x + 207.14
R2 = 0.9676

1000.0
800.0
600.0
400.0
200.0
0.0
-2.00

0.00

2.00
Xi

D4

4.00

PUKEKOHE CUMULATIVE DEFICIT DAYS DATA AND GUMBEL ANAYLSES

Year

CDD

Ordered by CDD

Rank

Pi

Xi

1996

70.7

2006

671.7

11

0.92

2.44

1997

211.3

1998

339.3

10

0.83

1.70

1998

339.3

2005

308.3

0.75

1.25

1999

157.4

1997

211.3

0.67

0.90

2000

75.5

1999

157.4

0.58

0.62

2001

57.4

2004

77.7

0.50

0.37

2002

6.6

2000

75.5

0.42

0.13

2003

40.2

1996

70.7

0.33

-0.09

2004

77.7

2001

57.4

0.25

-0.33

2005

308.3

2003

40.2

0.17

-0.58

2006

671.7

2002

6.6

0.08

-0.91

Pukekohe Gumbel Analysis

Cumulative Deficit Days

800
700

y = 181.21x + 92.749
R2 = 0.883

600
500
400
300
200
100
0
-2.00

-1.00

0.00

1.00

2.00

3.00

D5

Appendix E
Appendices B and C
from AS 2870 Standard

APPENDIX B
PERFORMANCE CRITERIA AND FOUNDATION MAINTENANCE
(Informative)
BI GENERAL
The designs and design methods given in the Standard are based on the performance
requirement that significant damage can be avoided provided that foundation site conditions
are properly maintained. This is expressed in Section 1 by the statement that the probability
of failure for reasonable site conditions is low, but is higher if extreme conditions are
encountered. It is neither possible nor economical to design for the extreme conditions that
could occur in the foundation if a site is not properly maintained. The expected standard of
foundation maintenance is described in Paragraph B2.
Some minor cracking and movement will occur in a significant proportion of houses,
particularly those on reactive clays, and the various levels of damage are discussed in
Paragraph B3.
The performance requirements of a concrete floor in respect to shrinkage cracking and
moisture reaction with adhesives are discussed in Paragraph B4.
A more extensive discussion of the material in Paragraphs B2 to B4 is contained in the
CSIRO Pamphlet 10-91: Guide to Home Owners on Foundation Maintenance and Footing
Performance and its recommendations should be followed.
B2 FOUNDATION MAINTENANCE
B2.1 Foundation soils
All soils are affected by water. Silts are weakened by water and some sands can settle if
heavily watered, but most problems arise on clay foundations. Clays swell and shrink due to
changes in moisture content and the potential amount of the movement is implied in the site
classification in this Standard, which is designated as follows:
(a)
(b)
(c)
(d)
(e)

A means stable (non-reactive)

S means slightly reactive
M means moderately reactive
H means highly reactive
E means extremely reactive.

Sites classified Class A and S may be treated as non-reactive sites in accordance with
Paragraph B2.2. Sites classified as M, H and E should comply with the recommendations
given in Paragraph B2.3.
B2.2 Class A and S sites
Sands, silts and clays should be protected from becoming extremely wet by adequate
attention to site drainage and prompt repair of plumbing leaks.
B2.3 Class M, II and E Sites
Sites classified as M, H or E should be maintained at essentially stable moisture conditions
and extremes of wetting and drying prevented. This will require attention to the following:
E1

(a)

Drainage of the site: the site should be graded or drained so that water cannot pond
against or near the house. The ground immediately adjacent to the house should be
graded to a uniform fall of 50 mm minimum away from the house over the first metre.
The subfloor space for houses with suspended floors should be graded or drained to
prevent ponding where this may affect the performance of the footing system.
The site drainage recommendations should be maintained for the economic life of the
building.

(b)

Limitations on gardens: the development of the gardens should not interfere with the
drainage requirements or the subfloor ventilation and weephole drainage systems.
Garden beds adjacent to the house should be avoided. Care should be taken to avoid
over-watering of gardens close to the house footings.

(c)

Restrictions on trees and shrubs: planting of trees should be avoided near the
foundation of a house or neighbouring house on reactive sites as they can cause
damage due to drying of the clay at substantial distances. To reduce, but not
eliminate, the possibility of damage, tree planting should be restricted to a distance
from the house of:
(i)
(ii)
(iii)

1 x mature height for Class E sites

1 x mature height for Class H sites
x mature height for Class M sites.

Where rows or groups of trees are involved, the distance from the building should be
increased. Removal of trees from the site can also cause similar problems.
(d)

Repair of leaks: leaks in plumbing, including stormwater and sewerage drainage

should be repaired promptly.

The level to which these measures are implemented depends on the reactivity of the site. The
measures apply mainly to masonry houses and masonry veneer houses. For frame houses clad
with timber or sheeting, lesser precautions may be appropriate.
B3 PERFORMANCE REQUIREMENT FOR WALLS
It is acknowledged that minor foundation movements occur on nearly all sites and that it is
impossible to design a footing system that will protect the house from movement under all
circumstances. The expected performance of footing systems designed in accordance with the
Standard is defined in terms of the damage classifications in Table Cl, Appendix C.
Crack width is used as the major criterion for damage assessment, although tilting and
twisting distortions can also influence the assessment. Local deviations of slope of walls
exceeding 1/150 are undesirable. The assessment of damage may also be affected by where it
occurs and the function of the building, although these effects are not likely to be significant
in conventional housing. In the classification of damage, account should also be taken of the
history of cracking. For most situations Category 0 or 1 should be the limit. However, under
adverse conditions, Category 2 should be expected although such damage should be rare.
Significant damage is defined as Category 3 or worse.
For Category 1 or 2 damage, remedial action should consist of stabilising the moisture
conditions of the clay and paying attention to repairing or disguising the visual damage. This
E2

should be regarded as part of the normal maintenance of houses on reactive clays.

Even significant masonry cracking with crack widths over 5 mm often has no influence on
the function of the wall and only presents an aesthetic problem. Generally, the remedial
action for such damage should start with an investigation to establish the cause of the
damage. In many cases the treatment should consist of stabilising moisture conditions by
physical barriers or paths or replenishing moisture in dry foundations. This can be followed
by repair of the masonry and wherever possible added articulation should be included while
repairs are being carried out. Structural repairs to the footing system such as deep
underpinning should only be considered as the last resort.
Underpinning should generally be avoided where the problem is related to reactive clays,
although it is recognised there may be occasional situations where underpinning or other
structural augmentation work is appropriate. None of this structural augmentation work
should be undertaken without proper engineering appraisal.
In some cases, walls may be designed to span sagging footings and cantilever beyond
hogging footings. In such cases, satisfactory performance involves the wall remaining free of
cracks and articulation joint movements, remaining within the limits for the particular
jointing system.
B4 PERFORMANCE REQUIREMENTS FOR A CONCRETE FLOOR
Shrinkage cracking can be expected in concrete floors. Concrete floors can also be damaged
by swelling of reactive clays or settlement of fill. The categories of damage are given in
Table C2, Appendix C. In the classification, account should be taken of whether the damage
is stable or likely to increase, and an allowance should be made for any deviations in level
which resulted from or during construction.
The time of attachment of floor coverings and the selection of the adhesive for them should
take into account the moisture in the concrete floor and its possible effect on adhesion.
Concrete floors can take a considerable time to dry (three to nine months).
Floor coverings and their adhesives can be damaged by moisture in the concrete and by the
shrinkage that occurs as the concrete dries. Drying could take three months or more. The time
of fixing of floor coverings and the selection of the adhesive should take these factors into
account (see AS 3958.1).

E3

APPENDIX C
CLASSIFICATION OF DAMAGE DUE TO FOUNDATION MOVEMENTS
(Normative)
TABLE Cl
CLASSIFICATION OF DAMAGE WITH REFERENCE TO WALLS
Description of typical damage and required r e pai r

Approximate c r a c k width limit

(see Note 3)

Damage
category

<0.1 mm

Hairline cracks.
Fine cracks which do not need repair.

<1 mm

Cracks noticeable but easily filled.

Doors and windows stick slightly.

<5 mm

Cracks can be repaired and possibly a small amount of wall will need to be
replaced. Doors and windows stick. Service pipes can fracture.
Weathertightness often impaired.

5 mm to 15 mm (or a number of
cracks 3 mm or more in one group)

Extensive repair work involving breaking-out and replacing sections of

walls, especially over doors and windows. Window and door frames
distort. Walls lean or bulge noticeably, some loss of bearing in beams.
Service pipes disrupted.

15 mm to 25 mm but also depends

on number of cracks

TABLE C2
CLASSIFICATION OF DAMAGE WITH REFERENCE TO CONCRETE FLOORS

Description of typical damage

Approximate
crack width
limit in floor

Change in offset from

a 3 m straight edge
centred over defect
(see Note 6)

Damage
category

Hairline cracks, insignificant movement of slab from level.

<0.3 mm

<8 mm

Fine but noticeable cracks. Slab reasonably level.

<1.0 mm

<10 mm

Distinct cracks. Slab noticeably curved or changed in level.

<2.0 mm

<15 mm

Wide cracks. Obvious curvature or change in level.

2 mm to 4 mm

15 mm to 25 mm

Gaps in slab. Disturbing curvature or change in level.

4 mm to 10 mm

>25 mm

Notes:
1. Crack width is the main factor by which damage to walls is categorised. The width may be supplemented
by other factors, including serviceability, in assessing category of damage.
2. In assessing the degree of damage, account shall be taken of the location in the building or structure where
it occurs, and also of the function of the building or structure.
3. Where the cracking occurs in easily repaired plasterboard or similar clad-framed partitions, the crack width
limits may be increased by 50% for each damage category.
4. Local deviation of slope, from the horizontal or vertical, of more than 1/100 will normally be clearly
visible. Overall deviations in excess of 1/150 are undesirable.
5. Account should be taken of the past history of damage in order to assess whether it is stable or likely to
increase.
6. The straight edge is centred over the defect, usually, and supported at its ends by equal height spacers. The
change in offset is then measured relative to this straight edge.

E4

Appendix F
Foundation Analysis

FOUNDATION ANALYSIS
Review of AS 2870/HB 28:197 Design Approach
F1

INTRODUCTION
The analyses provided in the following summarises the requirements of the Australian
Standard AS 2870 and comments on the applicability of these requirements for
buildings within the Auckland region.

F2

BUILDINGS INCLUDED IN NZS 3604

Clause 1.1.2 of NZS 3604 describes the buildings (and sites) which are covered by the
Standard, and to which the requirements of NZS 3604 apply.
The following types of construction covered by the Standard are summarised from
NZS 3604 Figure 1.2:
(a)

One and two-storey buildings slab-on-ground with clad framing or masonry

veneer.

(b)

One and two-storey buildings with a foundation wall no higher than 2 m and
with clad framing or masonry veneer.

(c)

Three-storey buildings with a foundation wall no higher than 2 m and with the
lower storey in concrete masonry.

F3

FOUNDATION DESIGN TO AS 2870

F3.1

General
Clause 1.4.2 of AS 2870 requires that foundations are to be designed for both
serviceability and strength for foundation movement and the effects of gravity loads.
AS 2870 further provides a building foundation designer with two options with
respect to designing the foundations of a structure. They are:
(a)

Standard designs these are prescriptive designs for a range of site

classifications and construction types. The Standard provides solutions for
stiffened raft, waffle raft and strip footings.

(b)

Specified designs these are to be based on engineering design principles and

detail in Section 4 of AS 2870, which allow the designer to alter the standard
designs for footing types listed in Clause 4.3 of AS 2870, viz:
(i)
(ii)
(iii)

Raft footing systems supporting a superstructure that relies entirely on the raft to
resist cracking,
Footing systems for walls which are (themselves) able to cantilever without
cracking,
Other footing systems.

This section also allows a designer, who is designated in AS 2870 as a qualified

engineer, to utilise the engineering design principles of AS 2870 for buildings that
are generally excluded from the standard designs of AS 2870 i.e. buildings beyond the
limits prescribed in Section F3.1(a). However for standard draft designs, AS 2870
specifies acceptable ranges for design parameters in Clause 4.5.1 of the Standard.
F1

Although Clause 1.1 of AS 2870 indicates that the Standard will generally be applied
to Class 1 and 10A Buildings, i.e. residential dwellings and non-habitable auxiliary
buildings, it is understood that AS 2870 is often applied also to commercial, industrial
and educational buildings where the construction types fall within the clad frame and
masonry construction types.
F3.2

Buildings excluded from standard design under AS 2870

Clause 3.1.1 of AS 2870 identifies the situations where the standard designs provided
by the Standard cannot be applied:
(a)

Class E or P sites [Expansivity classes (E)xtreme and (P)roblem sites];

(b)

Buildings longer than 30 m;

(c)

Slabs containing permanent joints e.g. contraction or control joints;

(d)

Two-storey construction with a suspended concrete floor at the first floor level except in
accordance with Clause 3.5 [which specifies geometric limitations of concrete floors for
buildings on Class A and S sites];

(e)

Two-storey construction in excess of the height limitations [of 8 m];

(f)

Support of columns or fireplaces not complying with Clause 3.6 [which specifies footing
construction for columns and fireplaces];

(g)

Buildings including wing-walls or masonry arches unless they are detailed for movement
in accordance with TN 61 [which is an industry guideline published by the Cement and
Concrete Association of Australia for Articulated Walling];

(h)

Construction of three or more storeys; or

(i)

Single-leaf earth or stone masonry walls greater than 3 m height.

Where the standard designs are precluded from use, AS 2870 provides for a qualified
engineer to design the footings in accordance with the aforementioned Section 4 of
AS 2870.
F3.3

Construction types included in AS 2870

AS 2870 provides for the design of foundations for buildings of brick (or earth)
masonry and clad framing construction. The three main construction types commonly
used in Auckland come within the AS 2870 definitions as follows:

F2

(a)

Clad frame is defined in Clause 1.7.9 of AS 2870 as timber or metal frame

construction with the exterior wall clad with timber or sheet material not
sensitive to minor movements. Includes substructure masonry walls up to
1.5 m high.

(b)

Articulated masonry veneer is defined in Clause 1.7.3 of AS 2870 as

masonry veneer construction in which the provisions for articulated masonry
have been applied to the masonry veneer i.e. construction joints etc.

(c)

Masonry veneer is defined in Clause 1.7.32 of AS 2870 as house

construction consisting of a load-bearing frame clad with an outer leaf of

masonry.
Where mixed construction types are used within a building, the equivalent
construction types of buildings including masonry for some or all the walls are
detailed in the Standard. Those requirements are summarised in the following
Table F1.
Table F1: Equivalent constructions for application in AS 2870 (taken from AS 2870 Table 3.1)
External walls
Single leaf masonry
Reinforced single-leaf masonry
Reinforced single-leaf masonry
Reinforced single-leaf masonry
Articulated single-leaf masonry
Articulated single-leaf masonry
Other single-leaf masonry
Other single-leaf masonry
Mixed construction
Full masonry
Articulated full masonry
Earth masonry
Infill panels of earth masonry
Load-bearing earth masonry

F3.4

Internal walls

Equivalent construction

Articulated masonry on Class A

and S sites or framed
Articulated masonry or
reinforced single leaf masonry
Masonry
Articulated masonry
Masonry
Framed
Masonry

Articulated masonry veneer

Framed
Framed

Articulated full masonry

Masonry veneer

Framed earth masonry

Load-bearing earth masonry

Articulated masonry veneer

Articulated full masonry

Masonry veneer
Articulated full masonry
Articulated full masonry
Articulated full masonry
Articulated full masonry
Full masonry

AS 2870 design philosophy for foundation movement

Although building foundations are often unsymmetrical and irregular in layout it is a
common design approach that the footing layout, for design calculations, is
approximated to overlapping rectangular areas which can be defined to have a
mound shape i.e. a profile that the soil changes to when it experiences a moisture
change.
For design purposes it is assumed that the mound is symmetrical and experiences
centre heave and/or edge heave as shown in Figure F1. The mound shape is defined
by two parameters the edge distance (e) and the free unloaded mound heave
within the confines of the plan area of the structure (ym).

F3.5

Design parameters in the consideration of foundation movement

The following soil parameters are taken from AS 2870:
(a)

Mound stiffness is defined in Appendix F4(c) of AS 2870 and is in the range

of 4001,500 kPa/m for beams in contact with swelling soil, with the further
limitation of 100 q (but not less than 1000 kPa/m) where q is the total building
load divided by the slab area. It is also noted that computed forces and
displacements are generally not particularly sensitive to the value of [mound
stiffness] used except for certain heave situations.

(b)

Soil heave the free unloaded heave (ym) is calculated using Equation F1 as:
ym = 0.7 ys

Equation F1
F3

where
ys = design characteristic surface movement (mm)
ym is always less than ys due to the slab stiffness and weight of the structure
above. Due to the inaccuracy of field assessment of these values (as discussed
in Section 5.4.2) the maximum in the range is always used for analysis to
ensure conservatism.
Theoretical mound due
to moisture effect but
before load effects

ym

A - Centre heave condition

Theoretical mound due
to moisture effect but
before load effects

ym

Where:
= Distortion in structure
ym = Free unloaded mound heave within
confines of structure plan

B - Edge heave condition

Figure F1: Soil structure interaction (from HB 28:1997 Figures 1.6 and 5.4)

AS 2870 Section F4 states that on a site that is wet throughout the profile at
the time of construction, a reduction of ym for edge heave not exceeding 40%
may be made. This reduction factor is directly applicable to sites that have
been pre-treated to maintain a high water content. This is further discussed in
Section 15.0 Residual swelling potential.
(c)

Edge distance
e = (Hs/8 + ym/36)
e = 0.2 slab length, or
(0.6 + ym/25)

for centre heave

for edge heave

Equation F2
Equation F3

where
Hs = depth below ground level at which no moisture change occurs (m)
(d)

Mound exponent (m) is a shape factor utilised in the Mitchell Method (1984)
and is defined in Equation F4.

(e)
m =
where

F4

1.5 L
a

Equation F4

L = span of footing (m)

a = Der - De, where
Der =

Hs
7

Equation F5

ym
25

Equation F6

De is the depth of embedment of edge beam from the finished ground

level
(f)

Permissible deflection the level of differential movement that a foundation

can undergo before the building will show unacceptable levels of damage is
specified in AS 2870 and summarised in Table F2 and Figure F2.

Table F2: Maximum design differential footing movement, , for design of footings and rafts (from
AS 2870 Table 4.1)
Type of construction
Clad frame1
Articulated masonry veneer1
Masonry veneer1
Articulated full masonry
Full masonry

Maximum differential footing movement,

As a function of span, mm
Absolute, mm
L/300
40
L/400
30
L/600
20
L/800
15
L/2000
10

Note 1. These construction types are the three most common in Auckland.

Figure F2: Acceptable footing movement as a function of length of footing

F3.6

AS 2870 gravity loads

Table F3 is reproduced from Clause 6.2.2 of HB 28:1997, and is based on a singlestorey building with tiled roof supported by trusses and is suggested as a guide for
preliminary designs.
The distributed loads listed in Table F3 include consideration of the self-weight of the
foundation as full contact between the foundation and the underlying ground is not
always achieved. The distributed loads vary depending on the depth and spacing of
the foundation beams.
From analyses carried out for this report, it is noted that the edge beam depth is
sensitive to the level of edge loading imposed on the building perimeter foundations
for the centre heave case.

Table F3: Typical stiffened slab loads for a single-storey building with tiled roof and trusses (taken from
HB 28:1997 Clause 6.2.2)

F5

Site
class

Type of construction
Clad frame
Articulated masonry veneer
Masonry veneer
Articulated full masonry
Full masonry
Clad frame
Articulated masonry veneer
Masonry veneer
Articulated full masonry

Distributed loads
Short direction (kPa)1 Long direction (kPa)1
4.8
5.3
5.7
6.9
5.8
7.0
9.2
11.2
10.7
12.6
5.3
5.8
6.8
8.0
7.2
8.5
11.6
13.6

Edge load
(kN/m)1
3.3
8.0
8.0
13.1
13.1
3.3
8.0
8.0
13.1

Note 1. HB 28:1997 Clause 6.2.2 explains there is a difference between the short direction ... and the
long direction ... because of the contribution of edge beams to uniform loading. The edge load
is superimposed.

F3.7

Failure definition
In order to guard against unrealistic expectations of foundation performance AS 2870
includes some definitions of failure, including the implications that certain levels (of
damage) should not be regarded as significant failure. These are discussed further in
Appendices B3 and B4 of AS 2870 reproduced in Appendix E.
This failure definition is noted in HB 28:1997 as an honest, though at times
unpopular, aspect of the AS 2870 Standard.
A major intention (of AS 2870) was:
(a)

To discourage claims for failure for what are very minor levels of damage. If
such claims persisted, they would force the cost of footing systems up far
above community accepted levels.

(b)

To provide a reasonable framework for builders, designers and house

guarantee funds to operate. Minor cracking of brick structures on reactive
clays is very difficult to avoid, even by use of good design. Without some
reasonable failure definition, many very competent designers and builders
would have found themselves involved in litigation.

Where significant damage (Category 3 or worse) has occurred, the extent of this is
clearly defined in AS 2870 so that the home owner is protected from poor design or
construction practices.
In addressing the design philosophy concerning load factors for strength, the authors
of HB 28:1997 state:
It would be ideal if house footing systems could be designed to cope with the
worst conceivable loads and circumstances and perform so that even the most
fastidious home owner had no cause for complaint. Unfortunately this is not
possible nor would it be in the communitys interest to waste resources in this
manner. House footings have to be designed to give a balance between cost of
construction and cost of minor repairs.
House footings are not expected to perform in the same way as conventional
concrete or geotechnical structures. In the event of failure there is not risk to life
nor are the failures, generally, expensive to repair.

F6

Thus it is appropriate to seek load factors and performance criteria that are
specific to footing design. These factors should give designs consistent with
known satisfactory past performance.
It is necessary to give guidance to home owners on methods they can use to
minimise the risk of damage to the footing system of their homes, just as they
protect the structural frame with preventative maintenance.

F3.8

Evaluation
Key elements of the approach adopted are summarised below:
(a)

With regard to performance expectation:

The AS 2870 Standard states that houses should be crack-free, but if ground
movement is experienced a low incidence of minor damage may result.
Further, AS 2870 expects the owner to provide some control of moisture
condition on the site so that extreme moisture changes (arising from, for
example, leaking plumbing or large trees) do not have to be considered in
design.
In order to ensure that these expectations can be reached, it is necessary to
rationally consider the load factors and to calibrate designs as calculated with
past practice.

(b)

With regard to assessment of load factor:

A rational method for assessing load factors for unusual design situations
involves a process of cost optimisation.
Using this procedure, the total cost of design plus an allowance for the cost of
failure is minimised, leading to estimates of what are appropriate load factors.

(c)

With respect to serviceability optimisation:

Although the analysis carried out (by the optimisation method) is in terms of
crack width, a conventional deflection/distortion limit expressed in terms of
span/500 or span/1000, or the like, is sought.
If is the crack width expected in the superstructure and is the width at
which repairs will be required, the lack of serviceability occurs when > .
The design problem is to find the average crack width as estimated by the
design method which can be a realistic target for design, given that both and
are subject to statistical distributions which can (in theory) be defined.

(d)

With respect to analysis and data used:

(i)
(ii)
(iii)

Cost definition must be related to, for example, improvements in serviceability

for example, providing deeper beams to increase the stiffness to result in half
the crack width under the same load (or deformation) would (by this model) add
an extra 15% to the structural cost (i.e. (0.5)-0.2 = 1.15)
Failure cost

F7

(iv)
(v)
(vi)
(vii)
(viii)
(ix)
(x)
(e)

actual repair costs to be (typically) no more than 1.5 x cost of original slab
Variability of design
design methods used to predict crack widths included in internal co-efficient
of variation of 50%
Compliant threshold
from records, it was found that average crack width beyond which repairs are
required is 9 mm (being the level at which owners are prepared to undertake
significant investigation, and potentially expensive repairs)
Target crack width
combining all the above factors, gives a nominal target crack width of 4.7
mm, corresponding to a deflection/distortion limit of 1/500.

With respect to yield optimisation:

taking the variability of the design process as 50%, and the cost of failure at three
times the initial cost, it has been found (Walsh 1985) that to minimise the overall
community cost, the average strength should be 2.7 x the average load effect. Converting
this factor to one based on nominal or characteristic values leads to a total load factor of
1.05 (or less) on loads, giving a somewhat surprising result that load factor for strength
can be taken as unity.

(f)

With respect to mound shape:

in the above analysis, the mound shape (ym) was defined so that the chance of being
exceeded during the life of the structure is 5%.

(g)

With respect to individual against global optimum:

The above discussion leads to an overall community optimum position, but some
consideration has to be given to the impact on, and consequences to, the individual who
may suffer a failure, including:
-

the individual benefits from cost savings which can be significant, even compared
with failure costs;
greater reliability in design from (the individuals or communities) improved
understanding actually lessens the risk;
advice to home owners being given to avoid or moderate the extent of damage which
may arise from inadvertent actions by owners;
more conservative designs being selected by owners if they wish to do so.

It must be appreciated that the failure rate of house footings is extremely low. Only one or
two failures per thousand houses are expected, and even then the cost per individual
house would typically be a few hundred to several thousand dollars. The failure rate for
houses is therefore much lower than for other geotechnical structures e.g. large dams
failing at a rate of 10 to 20 per thousand.

F4

AS 2870 PERFORMANCE STATEMENTS

F4.1

Engineering design principles

Paraphrasing AS 2870 requirements for Design by Engineering Principles
(references to AS 2870 clauses e.g. Clause 4.2 as noted).
Cl.4.1 Slabs and footings designed in accordance with engineering principles shall
satisfy:

F8

(a)

Clauses as referenced separately in AS 2870 Section 4.

(b)

Requirement of NZS 3101:2006 Concrete Structures Standard.

Cl.4.2 Slabs and footings and associated superstructure are required to achieve the
performance requirements set out in Cl.1.3 when subject to the loads noted
therein.
This states:
The footing systems complying with [the AS 2870 Standard] are intended to achieve
acceptable probabilities of serviceability and safety of the building during its design life.
Buildings supported by footing systems designed and constructed in accordance with this
Standard on a normal site (see Clause 1.3.2), which is:
(i) not subject to abnormal moisture conditions; and
(ii) maintained such that the original site classification remains valid and abnormal
moisture conditions do not develop
are expected to experience usually no damage, a low incidence of damage category 1 and
an occasional incidence of damage Category 2.

AS 2870 makes reference to Appendix B for information and guidance on the

maintenance of foundation site condition, and to Appendix C for damage
categories, viz Tables C1 and C2 for classification of damage with respect to
walls and concrete floors respectively.
NB: Refer Appendix E reproducing Appendices B and C from the AS 2870
Standard.
Cl.4.3 Footing systems are required to be designed as either:
(a) Raft footing systems, supporting a superstructure that relies entirely on the
raft cantilever to resist cracking.
(b) Footing systems for walls which are (when acting with the foundation
system) able to cantilever without cracking.
F4.2

AS 2870 design philosophy for gravity loads

The Standard includes empirical and section analysis rules for deriving the required
foundation strength and flexural stiffness, when the foundation system is analysed
using the pre-qualified soil structure interaction methods and for the prescribed
design loads as:
(a)

Design Moment M* (being not less than Mu) of the footing system, where
Mu is nominal strength calculated from NZS 3101 and is the strength
reduction factor.

(b)

Flexural Stiffness (EI) of the footing system, when both Ec = 15000 MPa
(fc = 20 MPa) and I is determined from NZS 3101 Table G6.6.

To ensure post-rupture ductility, the flexural cross-section is required to be reinforced

to provide a nominal strength 20% greater than the cracking movement capacity (Mcr)
F9

calculated in accordance with NZS 3101.

F4.3

Soil structure interaction

The following soil structure interaction analysis rules apply:
(a)

Design load and load effects

The requirements of AS 1170 Part 1 are generally similar to those prescribed
in the New Zealand limit state Standard NZS 4203:1992 Code of Practice for
General Structural Design and Design Loadings for Buildings and the related
materials Standards e.g. NZS 3101 Code of Practice for the Design of
Concrete Structures.
Clause 1.4.2 of AS 2870 specifies the following Equation F7 as a means of
determining the design load for the calculation of settlement, while a
reduction factor of 0.3 is applied to the design load for assessing bearing and
uplift failures of the foundations.
Design load = dead load + load combination factor x live load
Equation F7
where
Dead load

= the weight of the slab, foundations and

superstructure of a building

Load combination factor

= 0.5 as shown in AS 2870 Clause 1.4.2

Live load

= as specified in AS 1170.1 Structural Design

Actions Part 1: Permanent, Imposed and
Other Actions e.g. domestic buildings have a
live load of 1.5 kPa.

The load combination factor was included in AS 2870 to take into account
the long-term development of soil pressures induced by the soil heave where
an unfactored dead load and factored live loads are appropriate (Mitchell peer
review comments for Stage I report).
(b)

Foundation movement
Foundation movement is required to be assessed as the level which has less
than a 5% chance of being exceeded over the life of the structure.
Design soil suction profiles (Hs u), e.g. Profile Gamma as derived earlier in
this study using the aforementioned engineering principles, are considered to
satisfy this requirement.
Soil parameters used for this analysis are taken as the mean values of available
results for each soil horizon or particular soil.

(c)

F10

Load effects

The factored design loads for strength and serviceability for the specified
ground movement corresponding to the 1000-year (extreme drought event)
are:
(i)
(ii)

Strength and/or safety (i.e. ultimate limit state to B1.3.1) = 1.0

Deflection and crack control (i.e. serviceability limit state to
B1.3.2) = 1.0

where B1.3.1 etc are performance statements within Clause B1 Structure of

the New Zealand Building Code.
Notably the load factors for strength are significantly less than the values
typically specified in the loading Standard (NZS 4203:1992 or
AS/NZS 1170.0:2004) which typically would be around 1.35 to 1.40 for a
design based on a more frequent (say 50 to 100-year) return period event, and
around 1.2 for the near drought (say 100 to 500-year) return period event.
This low value reflects the relatively low cost of failure as reported by Walsh
(1985 see F3.7), and are consistent with the performance requirements
specified in 1.3.2 of the AS 2870 Standard.
F5

COMPARISON OF NZS 3604 TO AS 2870

F5.1

Introduction
The following is a comparison of the respective geometrical limitations and design
parameters set out in NZS 3604 and AS 2870.

F5.2

Comparison of the geometrical limitations in NZS 3604 and AS 2870

Table F4 summarises the building variables for which both AS 2870 and NZS 3604
provide geometrical limitations for the building structure above foundation level.
Table F4: Comparison of NZS 3604 and AS 2870 building geometry limitations
Factor
Maximum height
Maximum number of storeys
Maximum height of foundation wall
Maximum length of building
one and two-storeys timber framed
Plan floor two-storey other forms of construction
area
three-storey other forms of
construction

AS 2870
8m
2
1.5 m
30 m
1
1

NZS 3604
10 m
3
2m
1
Unlimited
300 m2

No three-storey buildings

250 m2

Note 1. A () indicates that the applicable Standard makes no reference to a limitation on the
applicable factor.

F5.3

Design parameters for Australian and Auckland soils

Based upon the soil testing undertaken for this study and a review of AS 2870, the
foundation design parameters for Australia and those proposed for the Auckland
region are summarised in Table F5.
Table F5: Foundation design parameters for Australia and Auckland

F11

Design parameter
Hs
ys
ym

Span
Average load
Edgeline load

Australian parameter range

(from AS 2870 Clause 4.5.1)
<3 m
>3 m
1070 mm
100 mm
70 mm
749 mm
550 mm
5-0 mm
530 m
530 m
to 15 kPa
to 15 kPa
to 15 kN/m
to 15 kN/m

Auckland parameter range

1.5 m 1
1585 mm 1
1060 mm 1
550 mm 2
530 m 2
to 15 kPa 2
to 15 kN/m 2

Note 1. These proposed parameters have been taken from soil suction change Profile Gamma as
discussed in Section 13.3.
2. These proposed parameters have been adopted from AS 2870 as they have not been a focus of
this particular study.

F5.4

Foundation designs using SLOG software

F5.4.1 Introduction
A desktop review was carried out utilising the DOS-based software programme,
called slab-on-ground (SLOG), provided by Dr Peter Mitchell (Adelaide) as
discussed in Mitchell (1984), to determine the minimum founding depth of footings
for a defined building geometry, loading and proposed foundation reinforcement.
This review is set out in the following sections.
F5.4.2 Structural review methodology
The aim of the review was to determine the area of reinforcement steel required in
each foundation type to achieve the target depths indicated in Table F7.
The target depths are the minimum beam depth that meets the requirements of
NZS 3406 or AS 2870, and the analyses undertaken alter the reinforcing steel to
achieve these beam depths.
Using the design module of SLOG, the depth of edge beam was found using the
foundation design variables discussed in Section F5.4.3.
The results of the analyses are reported in Section F5.4.5.
F5.4.3 Foundation design variables
(a)

F12

Geometry a typical rectangular single-storey house of masonry veneer

construction and of external dimensions, 16 x 8 m, was modelled with the
following foundation layouts, as shown in Table F6:
(i)

Foundation Layouts 1A and 1B are a standard foundation design

from Figure 7.15A of NZS 3604 where the footing is nominally tied to
the slab for placement, but does not have adequate connection for the
footing to act integrally with the slab.

(ii)

Foundation Layouts 2A, 2B, and 2C are a standard foundation design

from Figure 7.14B (right-hand side diagram) of NZS 3604 where the

footing and slab are an integral structural element.

(iii)

Foundation Layout 3 is a standard stiffened slab design from

Figure 3.1 of AS 2870 and is deemed suitable for use on Class M sites
for masonry veneer construction.

(iv)

Foundation Layout 4 is a standard waffle slab design from Figure 3.4

of AS 2870 and is deemed suitable for use on Class M sites for
masonry veneer construction.
Table F6: Design loadings for structural analysis

Load parameter
Edge load on west end
Edge load on east end
Edge load on north side
Edge load on south side
North-south centre load
East-west centre load
Uniform distributed floor load

Foundations 1 and 2
Foundations 3 and 4

Design loads used for study

8 kN/m
8 kN/m
8 kN/m
8 kN/m
0 kNhn
0 kN/m
4.0 kPa
5.8 kPa

(b)

Loadings the SLOG programme allows for seven different loading inputs.
The load parameters and values used for the analyses reported herein are
summarised in Table F6.

(c)

Number of beams this input allows the user to specify the number of beams
parallel to the long and short spans to model both strip footings and raft
foundations.

(d)

Permissible deflection the AS 2870 recommended limits of span/600 and an

absolute value of 20 mm as indicated in Table 21 and Figure 23 for masonry
veneer construction have been used.

(e)

Design characteristic surface movement, ys, of 40 mm has been used, which

corresponds with a site classification of M and a free unloaded mound heave,
ym, of 28 mm.

(f)

Youngs modulus of concrete for the analyses reported herein, an Ec, value of
15000 MPa has been adopted as a default value and is commonly used in
footing designs in Australia to account for the development of shrinkage
cracks, and is specified in Clause 4.4(e) of AS 2870.

(g)

Compressive strength of concrete for the analyses reported herein, an f'c

value of 8 MPa has been adopted for Foundations 1A, 1B, and 1 and an f'c
value of 20 MPa has been adopted for all other foundations.

(h)

Tensile strength of concrete there are two inputs for tensile strength for
calculating the cracking moment capacity for 20 MPa concrete 2.7 MPa for
sagging moments and 1.8 MPa for hogging moments. These values of tensile
strength have also been adopted for foundations constructed using concrete
masonry blockwork.

F13

F5.4.4 Underlying SLOG theory

Mitchell (1984) suggests that the simplest structural analysis for the design of
foundations on expansive soils is the beam-on-mound method, for which the primary
design conditions discussed in Section F5 and shown on Figure F1 are:
(a)

The centre heave condition; and

(b)

The edge heave condition.

SLOG analyses both conditions for each foundation span by calculating the mound
shape factor and utilising a predetermined set of boundary conditions to calculate
either the required bending moment capacity or the minimum founding depth required
for the foundation system to resist both the imposed loads from the building
superstructure and those loads created by the soil structure interaction.
F5.4.5 Results of analyses
Table F8 summarises the results of the analyses undertaken for the foundation
systems discussed in Section F5.4.2 and Table F7.
F5.5

Back analyses of foundations used in the Auckland region

F5.5.1 Introduction
A foundation type similar to the Type 2A foundation from Table F7 and Figure F3 is
commonly used in Auckland. As discussed in Section 4.2, minimum embedment
depths of 300 mm and 450 mm, specified in NZS 3604:1999, and 600 mm, have the
corresponding beam depths of 450 mm, 600 mm and 750 mm respectively, once the
150 mm freeboard to the floor of the habitable space has been incorporated.
Section 2.2 of AS 2870 permits identification of the soil expansivity class of a site by
consideration of established data on the performance of houses on the soil profile.
The commentary of AS 2870 states that the method relies on assessment of damage
(cracking) of houses of masonry (either veneer or full) construction, or the level of
maximum differential movement of clad frame houses. Preferably, the appraisal
should be based on houses with similar wall construction to that which is intended to
be built and which are at least 10 years old. If light footings have been used
satisfactorily in the past, the classification of a site in that area should be Class S or at
worst Class M".

F14

Slab tied into

footings
D12

2-D12

Foundation beams
around perimeter

8m

Beam depth
Embedment depth 150
(300-600mm)
ys

Cladding
type varies
665 mesh

16m

Figure F3: Typical Type 2A foundation

On the basis of the foregoing, the following section examines the likely design
characteristic surface movements, ys, in the Auckland region arising from the
common forms of footing construction, given that there is little anecdotal evidence to
indicate the foundations designed and constructed to a particular detail in the past
10 years have failed.
F5.5.2 Methodology for back analyses of foundations
The foundations were analysed using the foundation design variables set out in
Section F5.5.3 and the NZS 3604 specified reinforcement, as shown in Figure F3, of
665 mesh in the slab, 1-D12 bar at the top of the footing and 2-D12 bars in the base of
the footing.
The design variables discussed in Sections F5.4.3 and F5.4.4 were used in these
analyses. However, the edge loads specified in Section F5.4.4(b) were reduced from
8.0 kN/m to 3.3 kN/m for clad frame construction.
The aim of the back analyses was to ascertain the value of free unloaded mound
heave, ym, and subsequently the design characteristic surface movement, ys, that the
foundation beam could withstand, if the permissible deflection was based on the
AS 2870 recommended limits indicated in Table 21 and Figure 23, for clad frames,
articulated masonry veneer and masonry veneer construction respectively.
F5.5.3 Results of back analyses of foundations
Back analyses of the Type 2A foundations, on the basis of the embedment and beam
depths set out in Section F5.5.2, suggest that the following limits might apply for the
design characteristic surface movement (ys) if cladding damage is to be avoided for a
building with a footprint of 16 x 8 m.

F15

Table F7: Foundation layouts for SLOG analysis (site expansivity Class M)
Reference in
Standard

Minimum specified
reinforcement in
footing

1A

NZS 3604
Fig 7.15A

Top 1-D12
Bottom 2-D12

-2

1B

NZS 3604
Fig 7.15A

Top 1-D12
Bottom 2-D12

550

2A

NZS 3604
Fig 7.13B
(without
DPM slip
layer)

Slab 665 M
Top 1-D12
Bottom 2-D12

NZS 3604
Fig 7.14B
(right-hand
side diagram)

Slab 665 M
Top 1-D12
Bottom 2-D12

600

NZS 3604
Fig 7.14B
(right-hand
side diagram)

Slab 665 M
Top 1-D12
Bottom 2-D12

750

Foundation
reference

2B

2C

Foundation
layout

Foundation
detail

Slab SL72
Bottom 3-L11TM
3

AS 2870
Fig 3.1

AS 2870
Fig 3.4

NZ equivalent
Slab 664 M
Bottom 3-H12
Slab SL72
Bottom 3-L11TM
NZ equivalent
Slab 664 M
Bottom 1-H12

Target
depth1
(mm)

-2

600

600

600

600

750

400
400

310
310

Note 1. Two definitions of target depth apply. The upper value is the structural beam depth and
the lower value is the top of slab to bottom of footing dimension.
2. The analysis of Foundation 1A was undertaken to determine the minimum beam depth
required if the foundation specified in NZS 3604 was constructed on a Class M site. The -"
signifies that the authors had no predefined target depth for this portion of the analyses.

F16

Table F8: Results of analyses of foundations using SLOG programme (site expansivity Class M)

1A
1B

Beam
width
(mm)
240
240

Slab
thickness
(mm)
N/A
N/A

2A

240

2B
2C
3
4

240
240
300
110

Foundation
reference

Reinforcing steel in footings

Bottom

Slab

Top

2-D12
2-D12

NS 1
NS 1

N/A

2-D12

665M

100
100
100
85

2-D12
2-D12
3-L11TM
1-N12

665M
665M
SL72
SL72

D12
2-H30 2
D16 +
D10 2
D16 2
D12
NS 1
N12 3

Minimum target
depth required
1725
600
600
600
750
400
350 3

Note 1. NS indicates where there is no reinforcing steel specified, either in the slab, due to lack of
connection between slab and beam, or in the top of the footing, because AS 2870 specifies
only mesh in the slab for these specific footings.
2. Reinforcing steel specified differs from that required by NZS 3604 to assess the sensitivity of
the minimum beam depth derived by the programme to the amount of top steel in the
footing.
3. There appears to have to be an experience factor used in the design of standard footings
based on the historical knowledge and judgement of Australian engineers in the preparation of
some of the standard designs specified in AS 2870. In this case the AS 2870 specified
reinforcing is adequate for a target depth of 400 mm. An additional N12 bar at the top of the
footing reduces this to 350 mm. However, increasing the bar size beyond 12 mm diameter has
no advantage in decreasing beam depth.
Table F9: Results of back analyses of Type 2A foundations for a 16 x 8 m building using SLOG
programme

Construction type

Clad frame
Articulated
masonry
veneer
Masonry veneer

Beam depth
(mm)

Embedment
depth
(mm)

450
600
750
450
600
750
450
600
750

300
450
600
300
450
600
300
450
600

Calculated maximum
surface movement, Ys, to
avoid cladding damage
(mm)
48
58
68
36
46
56
24
34
44

Expansivity
class
H
H
H
M
H
H
M
M
H

F5.5.4 Interpretation of back analyses

Based on the back analyses results in Section F5.5.3, it can be concluded that the
embedment depth of Type 2A foundations for a building with a footprint of 16 x 8 m
could be determined for a specified site design characteristic surface movement, ys,
using a nomograph as in Figure F4, which reflects the cladding construction type
adopted.
The historical performance of residential building foundations in Auckland indicates
there have been few masonry veneer buildings built in accordance with
NZS 3604:1990, with an embedment depth of 450 mm and in some cases 600 mm,
which have displayed significant levels of cladding damage over reasonably long
F17

periods of time. One conclusion could be drawn from this is that the footing depths
historically provided by NZS 3604:1990, discussed in Section F2, are reasonable for
masonry veneer, articulated masonry veneer and clad frame buildings respectively.
It is the authors recommendation that a building damage survey, as discussed in
Section 2.5, is carried out to validate the performance of residential building
foundations in Auckland, and to provide designers and local TAs with design
nomographs similar to that shown in Figure F4 for varying sizes of buildings.

Figure F4: Surface movement against embedment depth

F6

APPLICATION OF AS 2870 IN AUCKLAND

F6.1

Introduction
Several factors contribute to whether or not AS 2870 is applicable to the Auckland
situation. They include:

F6.2

(a)

Do the geometrical limitations within NZS 3604 align with those described in
AS 2870?

(b)

Do the construction types within NZS 3604 align with those described in
AS 2870?

(c)

Are the design assumptions and parameters from AS 2870 directly transferable
to the New Zealand situation?

Conclusions
As discussed in Section F3.1, AS 2870 provides for two areas of design standard
designs and specific designs. The following summarises the findings of the structural
analysis of foundations on expansivity Class M sites using the SLOG programme:

F18

(a)

Type 1 foundations are unsuitable for construction on expansive soil sites in

the Auckland region without modification.

(b)

Foundations where the floor slab is fully integrated with the block base wall to
provide structural performance similar to that of Type 2A foundations, also
require modification from the details provided in NZS 3604 to be suitable for
construction on expansive soils sites in the Auckland region.

(c)

Foundation Type 2C requires no modification from the detailing specified in

NZS 3604 to be suitable for construction on Class M expansive soils sites in
the Auckland region where the beam depth is at least 750 mm. For reduced
beam depth, additional steel reinforcement will be required as indicated in
Table F7 for Foundation Type 2B.

(d)

Based upon the geometry limitations for AS 2870 summarised in Table F4, it
is concluded that the standard designs from AS 2870 for stiffened slab
construction or waffle slab (similar to Foundations 3 and 4 in Table F7) are
applicable to the following buildings:
(i)

Clad frame construction of up to two-storeys with concrete masonry

blockwork foundation walls less than 1.5 m high.

(ii)

Masonry veneer construction of up to one-storey with concrete

masonry blockwork foundation walls less than 1.5 m high.

All other types of construction need to be subject to specific design in accordance

with Section 4 of AS 2870.

F19