Best practice in medium density housing design

Best practice in medium
density housing design
for
Housing New Zealand Corporation
A report on
Best practice in medium
density housing design
for
Housing New Zealand Corporation
September 2004
David Turner
John Hewitt
Cesar Wagner
Bin Su
Kathryn Davies
i
Contents
Executive Summary
1
Summary of Conclusions
Introduction
5
Context and Research Aims
Legislative Background of Medium Density Housing in New Zealand, and Attitudes Towards
Residential Density
Literature Review
New Zealand
Australia
North America
United Kingdom
Summary and Conclusions
A New Zealand Definition of Medium Density Housing
11
21
Introduction
Density
Density and Privacy
Security and Privacy
Car Parking and Storage
External Style
Summary
Case Studies: Methodology and Criteria
Introduction
Methodology
Site Selection
Location
Multi-development Sites
Methodology
Topographical Criteria
Value and House Types
Refuse Collection
Washing/drying Arrangements
31
Case Studies
Case Study Conventions
Case Study Data
Case Study Evaluation
Glossary
(1) Vinograd Mews, Harbour View, Waitakere City
(2) Adelphi Villas, East Tamaki, Manukau City
(3) Seymour Road, Sunnyvale, Waitakere City
(4) Corban Village, Henderson, Waitakere City
(5) Fairhaven, Glen Eden, Waitakere City
(6) Romola Street, Glendowie, Auckland City
(7) Tuscany Towers, Ambrico Place, New Lynn, Waitakere City
(8) Melview, Ambrico Place, New Lynn, Waitakere City
(9) Albion Vale, Sunnyvale, Waitakere City
(10) Arawa Street, New Lynn, Waitakere City
(11) Oates Road, Glen Eden, Waitakere City
(12) Mt Taylor Drive, Glendowie (Project), Auckland City
(13) St George’s Terrace, Avondale, Auckland City
(14) Gunner Drive, Harbour View, Waitakere City
(15) Rowena Crescent, Glendowie, Auckland City
(16) Tuscany Way, Harbour View, Waitakere City
(17) Sacramento 1A, Botany Downs, Manukau City
(18) Oatlands Development, Pennant Hills Road, Sydney
39
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
ii
Best practice in medium density housing design
(19) Fontenoy Road, Macquarie Park, Sydney
(20) Carolina Place, Albany, North Shore City
(21) Bush Road, Albany, North Shore City
(22) Holly Street, Avondale (Project), Auckland City
(23) Cottontree, Brisbane
(24) Soljak Place, Mount Albert, Auckland City
(25) Ewenton St, Balmain, Sydney
(26) Beaumont Quarter, Auckland City
(27) Sacramento 1B, East Tamaki, Manukau City
(28) Hillsborough Road, Lynfield, Auckland City
(29) 2 Ambrico Place, New Lynn, Waitakere City
(30) Mokoia Road, Birkenhead, North Shore City
(31) Galway Street, Onehunga, Auckland City
(32) Krisley Court, Ambrico Place, New Lynn, Waitakere City
(33) Keeling Road, Henderson, Waitakere City
(34) Eden 1, Mt Eden, Auckland City
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
Case Studies Data Table
77
Discussion and Conclusions
Introduction
Density and Layout Type
Summary
Vehicle Planning and Parking
Mixed Development and Internal Design
Further Research
References
79
87
General Media References
Appendix A Local Authority Intensive Housing Policies in Metropolitan Auckland
North Shore City Council
Manukau City Council
Auckland City Council
Waitakere City Council
95
Acknowledgements
The report was commissioned by the Research and Evaluation Team of Housing New
Zealand Corporation and was prepared by the Housing Research Group of the School of
Architecture and Landscape Architecture at Unitec New Zealand.
The views contained in the report are those of the authors and do not necessarily reflect the
views of Housing New Zealand Corporation.
The authors wish to acknowledge the contribution of Cook Sargisson Pirie, Architects,
JBA Urban Planning Consultants Pty Ltd., (Sydney) and Architectus Ltd, for supply of
data material. All photographs and drawings used in the report were produced by David
Turner and Cesar Wagner, unless otherwise indicated.
Executive
Summary
2
Best practice in medium density housing design
Ÿ development values will be retained
or improved at higher densities if
design techniques are sophisticated;
Ÿ extra development costs of higher
density can be recovered by better
unit values if design improvements
are made.
MEDIUM DENSITY HOUSING:
EXECUTIVE SUMMARY
This study identifies the characteristics and
potential of medium density housing as a
typology suitable for affordable urban
development in the New Zealand context.
The conclusions listed below are based on
three premises:
3.
(i)
Medium density housing has
developed in the last decade as a
common housing typology, but is not
foreign to the urban culture of New
Zealand;
(ii) Research and literature on medium
density housing in New Zealand is
very limited in scope, quality, and
quantity;
(iii) Planning strategies to consolidate
urban growth pre–suppose a higher
density housing form that, at this
stage, lacks any clear definition or
preferred model.
Ÿ density ceilings can be identified for
different layouts (defined primarily
by types of car parking provision);
the trade-offs that occur between
different objectives can be located
on a density scale, as illustrated in
the case studies reviewed;
Ÿ the need for developers and designers to acknowledge that one 'highvalue' compromise often reduces
the quality of the whole living environment for all units;
Ÿ the most successful developments
take detailed account of all design
issues, including the intended resident mix, neighbourhood character,
interface with the public domain,
site specifics (e.g. topography), car
parking, appearance (style), privacy, security, landscaping, low
maintenance, and refuse collection;
Ÿ no single design factor determines
best practice.
Summary of Conclusions
1.
2.
Medium density housing invariably
involves a degree of compromise.
This is a consequence of building at
higher density levels (than traditional
suburban housing) while seeking to
address multiple objectives, including the mix of house types, car access, privacy, security, interface with
the public domain, and construction
costs.
A review of the literature indicates
that:
Ÿ there are numerous ways of calculating density, and the term medium
density housing refers to different
density ranges in different jurisdictions;
Ÿ good design becomes critical above
a density threshold of approximately 30 dwellings per hectare;
The literature on medium density
housing and the case studies reviewed in this report indicate that:
4.
Housing Mix: mix, in this report, refers to a mix of house types, house
sizes, and tenure-types (owner-occupiers and rental), within a development. The case studies in this report
are mainly private sector schemes
that reflect a desire for commercial
certainty of outcome, with few developments catering for a housing
mix. However, the literature review
suggests that where a broader strategy has influenced design a more
mixed development has been
achieved, along with a notably more
Executive Summary
3
diverse, socially active community,
at all levels of density
Medium density housing in New
Zealand needs to identify with the
local traditions of domestic design
(while avoiding a 'compacted
suburbia' approach) and at the same
time establish its own language
without reference to imported 'style'
and expression.
A review of the literature suggests
that a carefully considered mix of
house types, house sizes, and tenure
types
makes
an
important
contribution to the success of many
medium density developments.
5.
The study observes that traditional
housing forms are widely re-employed in New Zealand in modified
forms and in compacted versions,
both inside the house and in the site
layouts, in many new developments.
It is considered that quality medium
density housing environments cannot be achieved by this strategy, and
that the challenges of changing urban
lifestyles, demographic shifts, and
environmental conditions cannot be
adequately met by this 'compacted
suburbia' approach. Best practices in
other comparable countries have developed house types and layouts
specifically suited to medium density
housing.
In addition to the above:
7.
Ÿ occupation by more varied forms of
family and household composition,
as needs change over time;
Ÿ a wider variety of activities to be
more readily undertaken (e.g.
home-based employment).
8.
Public acceptance of medium density
housing is affected by location, and
design. Public and neighbourhood
expectations of new schemes include
their ability to offer economic and
social integration. Good design quality has been identified in Britain, the
United States, and Australia as a key
factor in increasing the degree of
public acceptance of medium density
housing. New projects could follow
the recommendations of Australian
researchers to select architects by
reputation and design skill, as already occurs in New South Wales,
Victoria, and increasingly, in other
centres.
9.
The recognition of the relevance of
urban design principles (e.g. character, legibility, adaptability) in the design of the best examples of medium
density housing is established in the
literature. The principles of high
quality urban design could be applied
Future medium density housing
should avoid a 'compacted suburbia'
approach
and
consider
the
development of climate-responsive,
adaptable house types, including rear
access layouts, and courtyard types,
up to identifiable density 'ceilings'.
6.
Design: In New Zealand, the external
style of medium density housing is a
significant factor in creating both
identity, and compatibility within a
given neighbourhood. Many contemporary medium density developments demonstrate that a wide
variety of styles can contribute to the
critical strategy of disguising the differences between medium density
housing and traditional lower density
suburban housing.
The trend towards more flexible living space in new housing, seen in private sector developments, could
impact on design in all housing
forms, including the medium density
category. More flexible internal
space facilitates:
4
Best practice in medium density housing design
more positively in the medium density typology, in line with urban initiatives currently being considered in
New Zealand and overseas.
10.
Medium density housing in New
Zealand is capable of providing residential environments of excellent
quality. In the best models it offers
identity, security, privacy, proximity
to private vehicles, and ground level
external private space. As a housing
type, it can be designed to achieve affordable and sustainable buildings
and communities, evidenced by
schemes developed in other countries in the 1970s and 1980s, and
earlier.
In future, increasing numbers of New
Zealanders will live in medium
density housing. Improvements in
the design of medium density
housing can enhance the quality of
life for residents, increase public
acceptability of more intensive
housing, and contribute to the
building of more sustainable
communities.
1
Introduction
Context and
Research Aims
6
INTRODUCTION
Context and Research Aims
The purpose of this report is to examine
medium density housing as a typology to
determine best practice in design for an
affordable and durable model for New
Zealand urban conditions. The report
focuses on medium density housing in the
Auckland region but has wider relevance
for other New Zealand urban areas undergoing intensification.
In many other countries, medium density
housing has been recognised as a form of
housing with definitive characteristics, and
offered as an alternative residential form to
low density suburban development. A
study of New Zealand housing in the period
between 1960–1990 reveals a small
number of examples, including the1970
Pitarua Court development, in Wellington,
by Peter Beavan, special housing for the
elderly, and student accommodation. There
are also examples of medium density
housing developments in the supply of
affordable housing. These may be regarded
as prototypes in the genre, and provide
evidence of New Zealand’s capacity to
experiment with different housing models,
without supplying a clear variation identifiably ‘of New Zealand’ in the medium
density typology.
In the period from 1990 to the present,
urban planning in New Zealand has moved
towards growth policies that seek to
consolidate city development in all the
main centres. Although not without opposition, planning strategies to intensify cities
have been widely adopted in international
practice, supporting the theory that
compact urban morphologies can and do
achieve growth through higher densities,
and
produce
sustainable
urban
environments.
These strategies reverse longstanding preferences for suburban expansion at low
density. Similar policies to impose spatial
limits on suburban growth are established
Best practice in medium density housing design
in countries comparable to New Zealand,
including Canada, Australia, and the USA.
Underlying the intensification policies now
in place in New Zealand is the assumption
that a relevant higher density housing
typology can be designed, or evolved, to
meet the needs of many sectors of the urban
community. As part of this process, there
are now many recent medium density
housing developments, particularly in the
private sector, that demonstrate the potential, as well as the problems of evolution, in
an unfamiliar typology.
The report is presented as an extended
summary of research into the relevant
context and literature, followed by a
description of the case study-based methodology for the critique and analysis of
recent medium density examples in the
Auckland area. The analysis is summarised
by a data chart providing an overview of
quantifiable material collated from case
studies. The conclusions drawn from this
are set out in Section 6. Each section of the
report is supplemented by endnotes, where
supporting material relevant to the project
is included.
Legislative Background of Medium
Density Housing in New Zealand, and
Attitudes Towards Residential Density
The debate concerning Auckland’s urban
form, and particularly its low–density
‘sprawl’, is not new. Auckland’s first
comprehensive town planning proposals,
the Outline Development Plan for
Auckland (Auckland Metropolitan Planning Organisation, 1951) specified as an
objective the need “to provide a means of
checking the tendency towards uneconomic and unsatisfactory sprawling development.” It also noted that “if a satisfactory
urban structure is to be developed…
various forms of residential development
will have to be considered.”
The Auckland City Council’s first operative District Planning Scheme (Auckland
City Council, 1958) attempted to foster
Introduction
such variety through the use of residential
1
zoning. However, increasing maximum
density controls in order to stimulate innovative approaches to housing design has, in
the Auckland area, been largely unsuccessful. Subdivision standards, with regulations controlling design decisions
concerning site coverage, setbacks and
height to boundary dimensions, have
stifled much creative endeavour and
favoured the development of ‘standard
solutions’.
One such is the ‘sausage’ flat block, introduced in the 1960s, and associated in the
2
public mind with increased density. Such
attitudes, together with the folk–memory
of the ‘slums’ in Newton Gully (5 room,
single storey cottages at approximately 40
dwellings per hectare, abbreviated to “dph”
in this report) reinforce the public
(mis)conception of what constitutes
medium and high density development,
and of the existence of a causal link
between increased density and decreased
3
environmental standards.
Conversely, the proponents of urban intensification use the concept of density as a
readily identifiable criterion of ‘good quality’ urban environments; with low density
signalling an unsustainable design
4
approach.
In addition, recognition of the interrelationship between housing density and
urban design is evident in local town planning literature, including the City of
Auckland District Scheme (Auckland City
5
Council, 1968), which notes that:
“New concepts of residential design
will be encouraged; e.g. new
concepts of housing and comprehensive developments where a number
of different types of residential buildings are located in a well planned
relationship to one another and to the
adjoining development.”
Similar sentiments were espoused in the
conclusions of the preliminary report into
7
housing, produced as a part of the Regional
Master Plan by the Auckland Regional
6
Authority (1967):
“Higher density housing types
should be located: within or near
main commercial centres…”
“Subdivisional standards for a
variety of residential zones should be
formulated to permit the provision of
a greater range of housing types of
suitable design.”
and
“Medium density housing types
should be designed and built comprehensively and where at all possible
permit separate legal title after
development.”
Since the reorganisation of Local Government in 1989 and the replacement of planning legislation by the Resource
Management Act in 1991 the four new
cities of the Auckland region have developed their own coordinated District Plans.
In the most recent editions, these each
address the issue of higher density housing,
and at the same time engage with matters
relating to sustainability, as the 1991 Act
requires.7
Other cities in New Zealand, particularly
Christchurch and Wellington, have also
recognised the need for higher density
housing design to be regulated separately
from subdivision rules.
The District Plans in all cases are reinforced by Design Guides advising developers and designers on a variety of ‘best
practice’ solutions to an unfamiliar
typology, these often illustrating regional
and local variations. Together, they represent much research effort, and provide an
effective platform for the generality of new
medium density housing. The various
District Plan sections relevant to this report
are summarised in Appendix A.
8
Best practice in medium density housing design
ENDNOTES
1
2
3
Residential zones covered 3963 hectares
(almost 90% of the zoned area of the city) and
were categorised in terms of site density as
Residential B (125 persons per hectare),
Residential C (250 persons per hectare), and
Residential D (500 persons per hectare). At
the 1956 figure of 3.8 persons per dwelling
this produces 33 dwellings per hectare (dph),
65dph and 130 dph respectively. (The
occupancy rate has since declined to 2.8
persons per dwelling in 2003 (Statistics New
Zealand, 2004)). It should be noted, however,
that the lower density zone B accounted for
3611 of the 3963 hectares, zone C 228
hectares, and zone D only 38 hectares (with
the Freeman’s Bay Transitional Zone
occupying the remaining 86 hectares), and that
the District Scheme stated that “it is
unlikely…that this site density will be reached
on more than a small proportion of the total
number of available residential sites”
(Auckland City Council, 1958). Nonetheless,
the figures show a marked correspondence
with those proposed in Sir Patrick
Abercrombie’s County of London Plan of
1944, which recommended net residential
densities of 250–500 persons per hectare for
improved post–war living standards, and
indicate that both the Auckland City Council
and the Auckland Metropolitan Planning
Organisation (who acknowledged their debt to
Abercrombie in the formulation of their
proposals for ‘flexible zoning’) were well
aware of international trends.
A study of housing density in the Auckland
suburb of Sandringham (Auckland Regional
Authority, 1976) notes that the construction of
such blocks has contributed to an increase in
net residential density from 10–15 dph in 1956
to 25–35 dph in 1976, but that “the type of
multi–unit development in the area rejects the
value of open, outdoor living and it is apparent
that many potential occupiers of medium
density housing are rejecting this type of
development because of this deficiency”
(Medium density housing was defined for this
report as 25–40 dph). It also notes that the
response of the (Mt. Albert) Borough Council
was to seek to reduce the maximum permitted
residential density (Auckland Regional
Authority, 1976).
The image of the British slums that the early
European settlers wished to avoid recreating
may be exemplified by the Liverpool ‘courts’
(mainly back–to–back and basement
dwellings) of the early nineteenth century,
which reached a net residential density of
1730 persons per hectare (Muthesius, 1982).
(Muthesius notes that this is only half of the
density of Berlin’s city blocks of the same
period.) At the 1821 figure of 5.75 persons per
dwelling this equals 300 dwellings per hectare
(although
contemporary
reports
of
overcrowding may equate this figure with that
for habitable rooms). London’s late nineteenth
century outer–urban suburbs were built at net
densities of 150–500 persons per hectare
(Muthesius, 1982); at 1881 figures of 5.38 p/d
this provides figures of 28–93 dph. The
Garden Cities of the early twentieth century,
associated in the public mind with the ‘ideal’
of low–density living are, at a net residential
density of 218 persons per hectare—at 1900
figures of 5.20 persons per dwelling
producing 42 dph (Tetlow & Goss, 1965)—
directly comparable with the ‘high–density
slums’ of Newton.
4
The Auckland Regional Authority’s Planning
Division (1967) stated that “present
uneconomic densities of up to 50–60 persons
per hectare cannot be sustained, and in fact do
not produce the choice either of housing type
or environment demanded by a large and
complex urban society.” Noting that
Auckland’s density, “in all sections of the
city”, falls within the definition of low density
at under 54 persons per hectare net, the report
suggested that the optimum range of net
residential density is 100–225 persons per
hectare, where, in terms of land conservation,
capital cost, and flexibility and variety,
“moderate increases in density achieved by
the provision of a variety of dwelling types
would be most economic…” (Auckland
Regional Authority, 1967).
5
This concern with the urban design
implications of Auckland’s ubiquitous low
density sprawl is a restatement of previous
planning policies. Despite a popular
conception of New Zealand as a recently
urbanised society, Johnston (1973), notes that
“as long ago as 1926, just 86 years after the
Treaty of Waitangi was signed, 63% of New
Zealand’s inhabitants lived in its cities and
towns, and of these 62% were in the five
largest urban areas—Auckland, Christchurch,
Dunedin, Hutt and Wellington.” Figures for
the Auckland urban area in 1926 show a
population of 192,000, with an average family
size of 4.2 persons and 92 dwellings per 100
families (Auckland Metropolitan Planning
Organisation, 1951). This net housing
shortage was attributed to the fact that “we
have few if any examples of satisfactorily and
Introduction
9
comprehensively designed housing schemes
other than those incorporating single unit
house development for three or more
persons”, resulting in “a large percentage of
the area being developed for streets with
monotonous similarity in the form of
development.” Thus, from the time of the first
attempts to develop comprehensive town
planning guidelines for Auckland’s projected
growth, it has been recognised that “all types
of residential development have their place in
a large modern urban structure…” (Auckland
Metropolitan Planning Organisation, 1951).
6
Further extracts from this document include:
“Residential development will be
closely related to the availability and
most efficient use of public services
and facilities…”
“The urban and suburban commercial
centres will contain the most widely
used services. Therefore the higher
residential densities will be located
near these centres where services may
be most conveniently obtained.”
“Residential development will be
diversified to provide for a wide range
of different kinds of housing and
physical groupings to meet the varying
needs of the community.”
“A greater variety of housing is
needed…”
“The provision of this greater variety
will result in land savings, due to the
consequent increase in overall
density…”
“The variety of housing needed can be
met with predominantly low rise
construction (i.e. up to 4 storeys.)”
Auckland Regional Authority, 1967
7
A clear pattern may be seen to emerge from
the above synopsis: of repeated attempts by
local planners to instil what they consider to be
essential urban qualities into the amorphous
urban mass of Auckland, only to be repeatedly
rebuffed by an at best apathetic, and at worst
antipathetic, public and its elected
representatives. After a half–century of
reiterating the advantages of vibrant urban and
suburban centres, and of variety and flexibility
in living environments made possible through
residential intensification, the recent inclusion
of the Residential 8 Zone (Strategic Growth
Management Areas) in the Auckland City
District Plan is receiving a predictable public
response. This time, however, the familiar
promotion
of
“sustainable
urban
environments which provide opportunities for
medium to high density housing within
walking distance of town centres…” coincides
with an increase in the status of urban design.
All members of the local building culture—
clients, developers, planners, designers,
builders and managers—are currently
espousing the added–value of design, and the
publication of The Residential Design Guide
for Developments in Residential Zones in
Strategic Growth Management Areas
(Auckland City Council, 2001), is a well–
timed and executed addition to the Council’s
range of persuasive powers.
“The Residential Design Guide is a
statement of what is considered to be
good urban design practice…
(and)…has been introduced to promote
and
encourage
well
designed
residential
developments
within
SGMAs.”
Auckland City Council, 2001
The focus of SGMAs is generally beyond the
levels of density covered in this study. The
acknowledgement that “design quality, rather
than density, is the predominant factor in
maintaining amenity for both residents of a
development and its neighbours” (Auckland
City Council, 2001) has, however,
fundamental significance.
2
Literature
Review
12
Best practice in medium density housing design
LITERATURE REVIEW
The literature review in this section is
selective, and is summarised on a country
by country basis. The material covered
includes social studies, design literature,
and policy publications.
New Zealand
Medium density housing is a product of the
strategic planning policies in place in most
New Zealand cities, and in the Auckland
region particularly, to contain future population growth in an intensified urban form.
Recent research relevant to this report
includes studies conducted by the
1
Auckland Regional Council.
There is evidence from these studies that a
perception of impending ‘slums’ is normal
in public attitudes, and that the most effective process for intensification involves
housing development on a reasonably large
scale to ‘provide a sense of community for
residents’ (Auckland Regional Council,
2000a p12).
After recognising the universal problem of
assembling sites for larger developments,
the Auckland Regional Council’s urban
design review goes on to identify the standard lot dimensions in Auckland (based on
the 55m x 18m quarter acre section) as one
of the impediments to higher density devel2
opment.
“A further barrier to good design is
that in many cases rules and procedures developed for traditional low
density housing are now being
applied
to
medium
density
developments.”
and
“...intensive developments involve a
number of trade–offs. The developer
wants density, the neighbour wants
privacy, the resident wants a good
view and aspect, while the community wants a good relationship to the
street. It is not always possible to
design a solution that overcomes all
these trade–offs.”
Auckland Regional Council, 2000a p21
The report to the Christchurch City
Council entitled The Effects of Infill
Housing on Neighbours in Christchurch
(Vallance, Perkins and Moore, 2002)
generally confirms the widely–held attitude of the New Zealand public to medium
density housing. Statements reflect much
of the social research in the field:
“Over two-thirds also believed that
infill housing would bring social
problems later.” (p5)
“...figures imply that peri-urban, low
density development is still the
popular choice …” (p9)
“...consolidated urban living is not
presenting them [residents of
medium density housing] with any
benefits …” (p43)
Vallance, Perkins and Moore, 2002
The Briefing Paper to the Auckland
Regional Council Forum on Affordable
Housing (Portal Consulting, 2000) identifies the principal demographic trends in
New Zealand as far as they affect the issue
of affordability. These include the growth
of sole–parent families, which tend “to be
embedded within an extended family
household”, rather than independent
households, the trend in Auckland towards
middle–class couples delaying family
formation, and the impact of Asian immi3
gration during the 1990s.
The quality of the built environment in
medium density housing is discussed in the
context of the large-scale development at
Ambrico Place, in New Lynn, in “Urban
Intensification in Auckland, New Zealand:
A Challenge for New Urbanism” (Dixon &
Dupuis, 2003). This study is a rare example
of social and physical planning research
conducted in the field. The paper considers
the relationship of the strategic planning
systems that provide the legislative
Literature Review
13
framework for medium density housing,
and the social and community effects, in a
study of approximately one-fifth of the
Ambrico Place households.
selection along with good practice for site
layout design, using examples from the
history of urban housing to reinforce the
principles discussed.
The characteristics of occupancy of
medium density housing (high levels of
tenanted property, and relatively high
percentages of recent immigrant families)
are confirmed in this study. In the area of
physical planning it comments on the
impact of New Urbanism in this housing
typology.
The New Zealand literature reviewed also
includes reference to regular features on
medium density housing in the general
print media, particularly the New Zealand
Herald, the Dominion Post, and the
Christchurch Press newspapers, and magazines such as Metro and North & South.
These publications normally engage expert
opinion in their feature articles.
In broad terms, the development is
regarded as a success by its residents:
“There were high levels of satisfaction with privacy, with almost all
respondents saying that privacy was
important to them and more than
four-fifths reporting that their indoor
space was private.”
Dixon & Dupuis, 2003
The residents were more critical, however,
of the planning process, which did not
make public the whole strategy for the
development of the scheme.
The Ambrico Place development is the
subject of four case studies in Section 5 of
this report: numbers 7, 8, 29, and 32, and is
described further in Endnotes to Section 4.
The Auckland Regional Affordable
Housing Strategy (Regional Growth
Forum, 2003) emphasises the need for
affordable higher density housing to
achieve high standards of design, to
achieve integration in neighbourhoods, and
to be responsive to cultural and age–related
issues, avoiding at the same time the penalties associated with higher building costs,
which can affect security of settlement.
An important factor in the typology is the
choice of house type, and its relationship to
layout. At different densities this decision
becomes a critical indicator of the residential environment. The HNZC Housing
Design Guide (undated), published for
internal use, deals with house type
Contributions to the debate in the print
media frequently take the form of detailed,
edited summaries of reports of Council
deliberations on changes to development
policy, for instance, the Report to the
Auckland City Council on proposed Residential 8 Zone changes. In others, public
concerns about ‘slums’, ‘ghettos’, and
similar supposed consequences of intensification are discussed.
Typical of such journalism is the feature
article “Security Issues” by Bob Dey
(Metro, May 2003), in part a discussion of
the trend towards gated communities, and
by the same journalist, “Dense City: The
Incredible Shrinking Section”, (Metro,
November 2003), reviewing declining lot
sizes in Manukau City. Useful insights and
comment are often found in this material;
for instance, reservations felt by developers
about the three storey townhouse model
which in one project has been modified (by
raising the rear patio level to the first floor)
to enable direct access to the space for
barbeque use, to suit social habits.
The impact of views expressed in newspaper and magazine journalism is considered to have significant influence on public
attitudes to intensification.
Australia
Medium density housing is a common
form of urban housing in Australia. Literature from 1975 to the present has
14
Best practice in medium density housing design
documented the evolution of the typology
in detail. Professional journals also feature
medium density housing developments at
regular intervals.
Medium Density Housing in Australia
(Judd & Dean, 1983) is a general description of the typology; parts of this text
present the case for medium density
housing as a solution to urban housing in
general (Newman, in Judd & Dean, 1983
p68); others discuss the process of development, effective management systems, and
practical house types for the genre.
This comprehensive study includes a
summary by John Byrne of medium
density housing in the public sector, based
on experience in South Australia. Byrne’s
comments on the public sector deal with
the social, economic, and political issues,
as well as design, which needs to “give
rental housing some of the external trappings of owner–occupied housing, such as
territorial control, some freedom to personalise, and indeed the ability to purchase.”
He observes that:
“The narrower the (street) frontage,
the better the yield, but potentially
the greater the problems of noise
interaction and privacy invasion.”
Byrne, in Judd & Dean 1983 p99
Five detailed case study examples are used
to illustrate public sector housing at densities between 26 and 83 dph (dwellings per
hectare).
A second Australian review, Designed for
Urban Living (Judd, 1993) extended the
relevant design area to include environmental issues, ecologically sustainable
design, and a section dealing with community attitudes. Judd identifies key design
issues as follows: urban and neighbourhood design; environmental fit; pedestrian
access and way–finding; vehicular access
and parking; identity; privacy; security;
dwelling layout; climate control and
energy conservation; and marketability.
Designed for Urban Living includes 21
case studies from all the principal Australian urban centres, illustrating developments that represent good practice in the
period up to 1993, at densities ranging
from 20dph to 67dph.
The study Medium Density Housing 1990
(Victorian Department of Planning and
Urban Growth, 1990) includes nine examples of lower density range developments,
none over 26dph, and all drawn from the
private housing sector; coverage of
consumer and neighbourhood attitudes is
valuable, however. The objective in this
study was to address the issues of declining
interest in Melbourne in medium density
housing as a choice for buyers and developers. Included in the recommendations
are recognition of the potential of the
typology in terms of sustainability, and
affordability.
Site Planning in Australia (King, Rudder,
Prasad and Ballinger, 1996) is a comprehensive summary of good housing layout
planning principles with sustainability,
urban design, and higher density housing
as a focus. The text relates to housing
design in the Commonwealth of Australia
(rather than a particular State) which has,
for more than a decade, been controlled by
the Australian Model for Residential
Development (AMCORD), published in
4
1990, 1992, and revised in 1995.
Following the Victorian Code for Urban
Residential Design (Victoria Department
of Planning and Housing, 1992) more
recent publications refer to the above texts
as primary sources for medium density
housing design. These include the New
South Wales Urban Design Advisory
Service handbooks Better Urban Living
(1998), Residential Densities (1998), Residential Flat Design Pattern Book (2001),
and the Residential Flat Design Code
(2002), between them providing the platform for all new medium and higher
density
development
in
Sydney,
Woolongong, and other urban centres in
New South Wales. Evolution of an urban
Literature Review
housing typology in Melbourne and
Sydney has seen a shift to densities higher
than those in the range considered in this
report. In other cities (Adelaide, Brisbane,
and Perth) low rise housing at medium
density continues to be the preferred form.
In the development of higher density
housing, generally apartments, new regulations do not recognise density in any of the
AMCORD definitions as a primary development control tool. Rather, use is made of
Floor Space Ratios and a building envelope
device (described as a “three dimensional
zone that limits the extent of building in
any direction”) to “inform decisions about
appropriate density for a site and its
context.” Building envelopes, height,
depth, separation, and side and rear
setbacks are of equal importance in the
design and control process to the Floor
Space Ratio. The focus of the New South
Wales Residential Flat Design Code, in
particular, is on the urban design issues
relating to development, and has applications in the New Zealand context for the
Residential 8 Zone category of the
Auckland City Council’s planning
document.
Two further texts are significant contributors to the literature: The Medium Density
Housing Kit (Marcus & Sarkissian, 1983)
and Housing as if People Mattered (Marcus
& Sarkissian, 1986). Both extend the detail
of design advice in the area, with emphasis
on children, domesticity, site planning,
parking, and landscaping.
In addition, the paper entitled “Trends and
Strategies in the Design of Medium
Density Urban Housing” (Radford &
Sarris, 2003), extracted from the Final
Report to AHURI (Southern) on the
subject of affordable medium density
housing solutions for Adelaide, refers to
literature dating from 1983–1993 (covered
above) as the primary research in the field
in Australia. The paper concludes:
“There is essentially little difference
in the design of built form between
15
well designed medium density
housing for low/medium income
families and (that) in the private
sector.”
The paper recognises the fundamentals of
medium density housing set out in Judd
1993, and reviews the principal issues of
parking, internal spatial design and fitting
out, security and privacy, as well as development process and building costs, adding
references to the Melbourne study Medium
Density Housing under the Good Design
Guide (King, 1999), and others. The study
confirms that medium density housing was
defined in earlier research and writings
with relatively minor adjustments necessary for current applications.
North America
Since 1990, housing design in North
America has acknowledged the parallel
needs of containing ‘sprawl’, for economic
and environmental reasons, and the challenge faced by US cities to achieve higher
standards of urban design, in a “search for
meaning in our physical environment”
(Fader, 2000 p2). Density has been at the
core of the debate about city form since
Stein, Mumford, and others, writing in the
1930s, and Jacobs (1961) began a critique
of urban and suburban development and
the consequent deterioration/decline of the
quality of urban life.
Literature is diverse and regional, with a
current emphasis on defeating suburban
sprawl, and with the most valuable contributions tending to be aligned to New
Urbanism. Privatisation philosophies have
lead to a broad literature of critiques of the
standards of housing, particularly for
subsidised accommodation (Garreau,
1991; Plunz & Sheriden, 1999, etc.).
Density by Design (Fader, 2000) is the
second publication by the Urban Land
5
Institute of America under this title. Fader
identifies the issue of ‘urban liveability’ as
a key element in urban housing, seeking
typologies that reverse the trend in the US
16
of fortress–like gated developments, and
that re–engage the street. The selected
examples used in this study “highlight
emerging quantitative standards for the
basic building blocks of housing and
community development: for example, lot
sizes, setback standards, street and alley
dimensions, and parking ratios.”
In a discussion of layout design, Fader
advocates rear access systems, against
what are acknowledged to be additional
costs, for the street-side advantage to
parking and walkability. The study also
deals with mixed housing, pointing to
successful developments where “integrating varying market segments within
small neighbourhood units (single block or
street, for example)” is a traditional urban
pattern that can continue to work in new
schemes.
The book represents the broad theories of
6
the New Urbanist movement. ‘New
Urbanism’ is a planning and urban design
theory that emerged in the 1980s. The
movement has become a major influence in
the planning of new communities, and in
urban regeneration, through the work of
Calthorpe, Duany, and others. New
Urbanism is endorsed by federal agencies
such as the US Department of Housing, and
is adopted as the preferred design approach
by the Urban Land Institute of America and
many real estate organisations and State
housing authorities.
United Kingdom
Housing in the United Kingdom has been
developed at higher densities for many
years: speculative housing in the private
sector is normally built at between 25 and
30 dph in wholly suburban locations. Planning controls are operated in a highly regulated environment in comparison with New
Zealand. Medium density housing generally refers to urban public housing, or
developments carried out by the various
privately managed, state–supported agencies such as Housing Associations. In these
Best practice in medium density housing design
developments density is often much higher
than the density levels of concern to this
report.
Of numerous recent publications, three are
selected here for their relevance to the
study.
Housing Design Quality through Policy,
Guidance and Review (Carmona, 2001) is a
detailed examination of control mechanisms and their effects on the housing
process. The book is divided into three
sections, of which the second deals with
innovations in the control process in relation to design guides, which are commonly
used in the United Kingdom. Relevance to
the development of medium density
housing in New Zealand lies primarily in
the comparisons that can be made with the
land–use policies outlined in Appendix A.
Housing Design in Practice (Colquhoun &
Fauset, 1991) is a broad–based compendium of all aspects of housing design,
including references to Australian (p146)
and New Zealand (p148) examples. The
book is a detailed and illustrated study of
housing in Western Europe and North
America, summarising twentieth century
advances in design at all levels of density. It
establishes the principle that building form
(of housing) is the determining factor in the
development of urban quality.
The authors recognise that social and
cultural differences have a fundamental
impact on choices relating to housing
density, impacts that are illustrated by
comparisons between the numerous countries studied. In their analysis of residential
planning, the authors deal with detailed
strategies: for instance, of the relationship
of density to cost (p175), density to car
parking (p173), and options for layout
7
design (p180–193, and p237). Also
considered and discussed is the relationship between increased densities (and the
consequential increase in development
cost), which is balanced for developers by
decreasing site acquisition costs per unit.
Literature Review
In addition, the RIBA Book of 20th Century
British Housing (Colquhoun, 2000), which
includes a general summary of current
housing finance methods in the United
Kingdom, provides a useful catalogue of
the achievements and processes of housing
in the United Kingdom.
The value of housing design and layout
(Commission for Architecture and the
Built Environment, 2003) is a report which
considers alternative layout and house type
designs in an environment where the
Government’s policies require the private
housing sector to increase residential
densities.8
After establishing the principle of density
as a governing factor, the report also
develops a methodology for assessing the
relationship of density to value. It
concludes that increased density of development, if designed with skill and care, can
both improve development margins and
urban living environments, and maintain
values in the marketplace. A critical
threshold, at 30 dph, is identified as the
point at which high design standards
become an essential factor in the developer’s calculation of density and value.
Car ownership levels are assumed by planning directives and providers of housing to
be acceptable at levels lower than those
applied in New Zealand, particularly in
larger cities, affecting both layout design
and density. Differences between housing
in the private and the public sectors are
identifiable by location, by external form,
including the cost of facing materials, and
by differing standards of maintenance in
the public spaces of the site.
Summary and Conclusions
A consistent feature of the literature is the
agreement that the term ‘medium density
housing’ is characterised by complexity,
and particularity of location and context. In
the literature, density is at once a quantifiable ratio and a condition of quality in
design relating to privacy, security, and
17
identity. The foundations of design theory
in this area have been clarified by the influential writings of Oscar Newman,
Rapaport, Chermayeff, and Habraken,
(dating from the 1960s, and not covered in
this review) dealing with the notion of
territoriality, and of public and private
space, and continue to attract the attention
of contemporary theorists by contributing
to the critique rather than solutions in
practice.
Of the studies in detailed site planning and
internal design in medium density housing,
the British publications are comprehensive, founded on experience in practice,
and have relevance to conditions in New
Zealand, if modified by culture, lifestyle,
building practice, and climate.
Current published material in the United
Kingdom confirms the continuation of a
strongly traditional orientation in housing
design, including medium density housing.
The widespread preference (public, institutional, and political) for traditional design
is reinforced by conservation–based planning controls, particularly affecting the
inner urban areas most likely to be selected
9
for redevelopment. Affordable housing is
generally supplied through rental housing
offered by Local Authorities and Housing
Associations.
The review of North American practice and
literature is abbreviated by the apparent
shortage of relevant material, although
there is a considerable quantity of case
10
study data. It is not thought that solutions
in the North American context contribute
significantly to a better understanding of
medium density housing in New Zealand
conditions, and it is noted that, apart from
relatively recent texts inspired by the Smart
Growth and New Urbanist movements,
neither of which relate directly to low cost
housing design, no distinct body of literature on medium density housing appears to
have emerged in the USA.
The Australian experience is directly relevant to New Zealand, though it requires
18
conversion of building systems, design
traditions, and is based on a more prescriptive regulatory system. There is a large and
well–regarded body of literature dating
from housing developments in the 1970s
and 1980s, including numerous case
study–based texts. This literature has
informed the evolution of the typology, and
combined with a generally more prescriptive planning regime, and the use and application of Design Guides, has contributed
significantly to a good quality standard in
the genre. Regrettably, and probably due to
the same prescriptive system, the diversity
of style seen in New Zealand is not a characteristic of medium density developments
in Australia. Newer urban housing appears
to be shifting towards a different model
characterised by significantly higher
Best practice in medium density housing design
densities than ‘medium’ density housing as
defined in this report, and reliant on a
building form that introduces common
internal spaces, underground parking, and
detachment from ground level access for a
high proportion of units. In the most
common form this housing is between four
and five storeys in height, has been defined
by the most recent literature emanating
from the Department of Urban Affairs and
11
Planning in New South Wales, and affects
all market sectors.
The four and five storey block form is now
the prevailing form for higher density
housing up to 140 dph in Sydney,
Melbourne, and, on a smaller scale, Brisbane. In other Australian cities where
population growth is lower, e.g. Adelaide,
there are fewer examples.
Literature Review
19
ENDNOTES
1
These reports are summarised in Building a
Better Future: Intensification Review–
Summary of Research Findings (Auckland
Regional Council, 2000a) covering the issues
of housing choice, preference, and
demographic profiles most likely to be
affected by higher density development
policy; community attitudes to it, with case
studies prominent in the methodology; an
urban design review in which the impacts of
intensification on the traditional residential
environments of Auckland are assessed, and
an analysis of the implications for the
Regional Growth Strategy.
(Housing New Zealand Corporation, 2002b)
both contribute at the level of house planning
and detail, to the process of developing better
models, particularly but not exclusively for
Maori and Pacific Island families. Neither
report addresses the issue of housing at higher
densities, or the medium density typology.
3
The reports taken together record the
expectation that:
Figures used in this paper are based on the
1996 census; it is noted that other indicators
and current Statistics New Zealand figures do
not fully align with the Portal summary.
“Higher density housing (has) fewer
people per dwelling reflecting the fact
that higher density residents are more
likely to be younger, single and
without children”
Some of the same issues of anticipated social
change are addressed in the paper entitled
“House and Home and their interaction with
changes in New Zealand’s urban system,
households and family structures” (Perkins &
Thorns; 1999). This analysis of demographic
change acknowledges increases in smaller
households and the impacts of lifestyle
choices in a discussion of the nature of place–
making and suburban values in New Zealand.
and that
“There is a common perception
amongst neighbours that medium
density housing attracts ‘transient’
people who are renting and who will
move frequently. The view is not
supported by the evidence.”
Auckland Regional Council. 2000a,
p11
2
The urban design review of this research
recommends
that
developers
should
collaborate with the city councils and the
Auckland Regional Council to promote
innovative ‘best practice’ intensive housing
design and construction practice (Auckland
Regional Council, 2000c p 31); this should
involve comprehensive integrated design
codes with a focus on the encouragement of
sustainable living environments, in accord
with more recent policy statements from all
four councils in the Region (Auckland
Regional Council, 2000c p21).
The research studies from the Auckland
Regional Council’s “Building a Better Future”
programme do not deal in design detail except
at the urban level: the emphasis is on social
attitudes and levels of acceptance, and
business and political decision–making.
Ki Te Hau Kainga, New Perspectives on
Maori Housing Solutions, (Housing New
Zealand Corporation, 2002a), and the Pacific
Housing Design Guide: Guidelines for
Designing Pacific Housing Solutions
The paper concludes with the observation that
demographic changes are driven by
compositional change (ethnicity), and
suggests that declining immigration will
reduce simple growth–driven change. The
‘shift–shares’,
rather
than
numerical
population growth, it is predicted, will affect
household demographics, and impact more
directly on affordability.
4
The AMCORD document, in three parts,
includes
definitions
for
density,
recommending the use of three terms, ‘site
density’, ‘net dwelling density’, and ‘gross
dwelling density’ to describe different
conditions. AMCORD covers all aspects of
urban housing, dealing with the principles of
design for traffic, site selection and layout.
5
The first edition (Wentling, 1988) is a source
quoted in Australian literature.
6
The North American movement followed a
revival of interest in classical origins of
architecture begun in Europe a decade earlier
by Leon Krier, John Simpson and others: this
may be regarded as an extreme reaction to the
inadequacies of modernism, which, until this
intervention, had been the unchallenged
design reference for all but a tiny minority of
housing schemes, particularly those that
aimed to establish medium density housing as
a housing typology.
As an alternative to modernism, classical architecture is unlikely to have any relevance to
20
Best practice in medium density housing design
higher density housing design in New Zealand, but two well regarded developments
based on New Urbanism have been carried out
in Sydney. A mixed scheme of low rise medium density housing combined with a group
of 15 storey apartment blocks, at Raleigh Park,
is the best known development in the genre;
the Oatlands development (case study 18)
draws on some New Urbanist ideas for layout
design, and achieves variety of house type,
unit value, and a variable density across the
site. The example in New Zealand nearest to
New Urbanist design principles is the Harbour
View development in Te Atatu (case studies 1,
14, and 16). These developments have a density of around 40 dph, except Gunner Drive
(case study 14).
A key strategy of New Urbanist theory is a
systematic, structured, and inclusionary methodology for the process of planning new developments, involving the community
affected by a sequence of workshop
'charrettes' to establish a sense of ownership in
the generation of new (usually higher density)
proposals. A frequent objective, based in the
movement's theory, has been to mix housing
tenure in larger projects without making physical or spatial distinctions between social or
economic groups: various design and housing
management techniques are used to achieve
this, as:
“Some of the units are for-sale, some
rental, some market rate, and some
subsidized housing, but the market
segments are not segregated one from
the other. In this case (Crawford
Square, Pittsburgh) the key to success
was that no visual distinctions were
made in the housing designs to signal
the type of housing tenure: a rental
townhouse looks like a for-sale
townhouse. Further, within the pool of
rental units, subsidized units are
rotated periodically, preventing any
stigma from being attached to specific
units.”
floor, or split between the two floors? It
is generally considered that a split …
(is) … the most inconvenient
arrangement. … (the type) particularly
creates difficulties with … washing,
control of small children, and the
disposal of rubbish”
Colquhoun & Fauset, 1991 p284
8
The merits and constraints of all multi–storey
house types are outlined in Chapter 7;
comments on three storey houses with integral
garages, for instance, include the following:
“… is a housing form that has never
been entirely popular in Britain. It is
mainly used in urban areas where high
density is necessary … The problem
relates to the distribution of rooms—
should all the living accommodation be
located on the ground floor or the first
(i)
evidence from research indicates that
there is no penalty attaching to higher
density for developers;
(ii)
good design becomes critical above a
density threshold of 30dph;
(iii)
development values will be retained or
improved at higher densities if design
techniques are sophisticated;
(iv)
extra development costs of higher
density can be recovered by better unit
values if design improvements are made.
9
In other contexts, design choices are
constrained by prescriptive planning systems
and design guides, such as the Essex Design
Guide (Stones, 1997), now extensively used as
a model for design in all southern areas of the
United Kingdom. These design guides are
effective in so far as they ensure compliance
with good practice via prescriptive planning
regimes; they are co–ordinated with nationally
directed practices for road and traffic design,
heritage policies and locally drawn District
Plans.
10
The issue of affordable urban housing appears
to be resolved by continuing use of the various
established mechanisms of low rent private
sector, of varying quality, and very high
density social housing ‘projects’, with small,
one–off developments, often of good quality
and architectural standard and at relatively
high density, providing the most relevant
models. The latter variations most frequently
take the form of apartment blocks with low
parking provision.
11
It is relevant to emphasise the point that
following the moves to consolidate city form,
higher density housing in two and three storey
layouts is undoubtedly successful in a large
number of developments to be seen in
Australian cities, a standard achieved through
the influence of comprehensive studies
researched and published in the period
between 1978 and 1993.
Fader, 2000 p13
7
In case studies, the CABE research team
established findings relevant to this study, as
follows:
3
A New Zealand
Definition of
Medium Density
Housing
22
A NEW ZEALAND DEFINITION OF
MEDIUM DENSITY HOUSING
Introduction
This section discusses medium density
housing in the New Zealand context, in
order to define the typology in contemporary urban residential conditions.
Density is discussed in this section, firstly,
as a system of measurement that references
dwelling units to a given area of land, and
secondly, as a factor that influences perceptions of privacy.
To form a definition of medium density
housing in the New Zealand context it is
also relevant to address the issues of particular concern to developers and designers:
security, car parking, and architectural
style.
Density
The most common definition of medium
density housing in current use in New
Zealand is:
Best practice in medium density housing design
The extended definition that generally
embraces examples in Britain and
Australia would suggest that the following
characteristics are also relevant:
Ÿ Ground level entry from a public
space
Ÿ A dwelling type with private
external
space
within
the
‘curtilage’, or territorial boundary
of ownership
Ÿ A dwelling type with direct or close
proximity to secure parking
Ÿ Separate legal title, including ‘unit
title’ ownership.
The word ‘curtilage’ is used in the British
literature to describe the territorial limits of
identifiable private ownership of a property
within a larger housing development. Separation of titles is also a New Zealand preference. House types that may be included
are detached, attached or terraced, and
apartments in low rise blocks.
According to the Australian Model Code
for Residential Development (AMCORD),
density is:
Housing at densities of more than
150m2/unit and less than 350m2/unit, or
30–66 dwellings per hectare (dph). This
definition is used by the majority of City
Councils and the Housing New Zealand
Corporation.
Ÿ A measure of population or the
number of dwellings per unit of
area;
Ÿ A measure of the form of the built
environment; and
Ÿ A measure of development
potential.
Australian literature further defines the
typology as “small lot subdivision, or
multi–unit development, ... (with the characteristics of) ‘attached, no lifts’” (Victorian Department of Planning and Urban
Growth, 1990 p1), and as “horizontally
attached dwellings which… rarely exceed
three stories above the ground with individual access and private open space at or
near ground level ...” (Judd, 1993 p8).
The AMCORD documents use the term
‘density’ to refer to a ratio describing the
relationship of a given number of household units to an area of land. They refer to
“many different ways in which this relationship can be expressed” (AMCORD,
1992a pp16–17), recommending three
principal definitions, of which the term and
definition ‘site density’ is most relevant to
this report.
A recurring feature of the literature
defining medium density housing is the
view that the concept of ‘density’, and the
nature of ‘medium density housing’, have
no universal or standard application.
The four City Councils in the Auckland
Region, Housing New Zealand Corporation, and both Wellington and Christchurch
City Councils operate a density range on
the basis of site areas of 150m2–350m2 for
A New Zealand Definition of Medium Density Housing
medium density housing, without using
absolute or pre–determined rules to govern
housing development. This practice is typified by Waitakere City Council, which
applies an ‘effects–based’ process to decision–making on medium density housing
proposals. The Auckland Regional Council
identifies “residential intensification as
developments with a net site density of
500m2 or less”, medium density at 350m2
2
or less, and higher density at 200m or less
(Auckland Regional Council, 2000 p 31).
The practice in Britain is to define development capacity, using density as a mechanism alongside other factors, in suburban
locations, with other controls relating to
form and site coverage in others. The
CABE report (2003) takes a different
stance, embracing the developer’s perspective, in defining site area as the area of land
that is required for a given development
(which may include significant public
works).
Australian planning systems recognise that
a density definition relevant to Brisbane or
Darwin is different from one applicable to
inner suburbs in Sydney and Melbourne,
and consequently ‘density’ as a planning
tool is not a sole arbiter of the design
process. The technique of applying a Floor
Space Ratio (FAR) or a Floor Space Index
(FSI) is commonly used in development
control in central and local urban areas in
preference to density.
As a method of setting maximum development limitations, density is therefore used
as a reference or guide rather than a precise
measurement regulator, or a control mechanism that provides certainty of outcome.1
Most recent references to the concept of
density confirm the relevance of two
general points:
(a)
density is not a useful mechanism for
determining quality in residential
design because other factors in
various combinations impact on the
outcome.
23
(b)
density is a human perception,
usually of a sense of ‘crowding’, and
therefore a highly variable factor in
housing design;
It is also the case that in the analysis of built
housing developments, density is usually
the first point of reference in forming bases
for comparisons: an ‘after the event’ position is created by a density calculation,
even where density is not a significant
factor in the design. While other factors
affect the quality of outcome, a density
calculation on some recognised basis is
necessary for valid comparisons to be
made, and comparisons that do not use a
density indicator can function only in terms
of their nominated criteria (for instance,
landscaping, building type or detail, enclosure systems, etc.).
To enable this comparison to be made, this
report will use a simple net site area basis
of calculation. This aligns with the first
AMCORD definition, excludes areas
external to the site (public roads, reserves,
railway lines, and other open space) but
includes public areas within the boundary
of the land predominantly occupied by the
housing itself.
Density and Privacy
Studies have established that density and
privacy are interdependent and that
achieving acceptable standards of privacy
is a key issue in the design of socially
successful higher density housing.
In his seminal study “Towards a redefinition of density”, Rapaport discusses the
nature of ‘density’ in terms of perceptions
of crowding, and the socially complex
issue of privacy.
“It is essential to consider in detail,
and to a high degree of specificity,
the relationship of given socio–
cultural groups to traditional density
figures, the relationship of a particular area to the larger context, … the
detailed layout and design of the
24
Best practice in medium density housing design
setting in terms of privacy, .. the
social rules available and used, and
so on…”
Rapaport, 1975 p153
The connections between density and
privacy are further analysed in the basic
Australian text, Medium Density Housing
2
in Australia (Judd & Dean, 1983), and in
Judd’s later text Designed for Urban Living
(Judd, 1993), which notes that:
“One important way of enabling
control over privacy is to provide a
clearly defined hierarchy of public,
semi–private, and private outdoor
spaces which discourages intrusion
by outsiders and provides necessary
buffer space between dwellings and
associated common access routes
(quoted from Marcus and Sarkissian,
1986 p39). The greater degree of
control that can be given to residents
as to how their private territory is
defined and personalised, the greater
the likelihood that privacy will be
optimised.
In … housing of two or more storeys,
overlooking of the private open space
of adjacent dwellings from upper
level rooms represents one of the
most common privacy problems.”
Judd, 1993 p30
Security and Privacy
Security (or its absence) has been an issue
associated with medium density housing
since the term came into common use, but
with little evidence to support the view that
higher density housing generally, or
medium density housing as a typology is
either less safe or more susceptible to crime
than other housing types. It is typologically
characteristic that greater concentration of
building, and proximity of public open
space can create anonymity (and therefore
lessen the possibility of intruders being
noticed) and equally, by placement of
windows and doors, construct a passive
surveillance environment that discourages
intruders.
As Judd says:
“.. criminal behaviour is related to
broader social problems and their
geographic distribution rather than
housing type or density per se,
medium density housing .. (has
tended to be) .. concentrated in mid
to inner–suburban areas or on public
housing estates, which often have
higher rates of burglary and personal
crime.”
Judd, 1993 p30
Advocates of Smart Growth in the USA
identify security amongst the three highest
priorities in their intensification agendas.
Judd makes reference to the issue of ‘Security’ related to Oscar Newman’s theory of
‘defensible space’ which has direct relevance to site layout design: such spaces
should be “assigned to specific groups of
residents” and “good territorial definition
can help to enhance identity … and
contribute to relieving social conflict
between residents.” (Newman’s work is
open to criticism for over–emphasis on
“design solutions to crime”, and has
limited application to a New Zealand definition for its focus on North American
social housing, in which security and
control are more severe difficulties.)
Defensible space is thus an abstract term
that describes a relationship of private and
public domains in perceived, as well as
spatial, senses. There is some evidence
from the case studies that in pursuit of a
well–lighted ‘defensible’ (in the sense of
‘secure’) common area, the developer’s
determination to remove the possibility of
concealment results in barren, uncomfortable spaces that also discourage
communality. In other projects, where
security is in the form of a physical barrier,
such as a controlled entry gate, there is a
sense that anyone seen ‘inside the fence’ is
A New Zealand Definition of Medium Density Housing
probably entitled to be there, obviating the
3
value of passive or casual surveillance.
In so far as design solutions can achieve
good security in an undefined community,
the ability of owners to view their car is
significant, even when the car itself is also
protected by an alarm system. As in other
housing forms, high standards of security
fittings to doors and window openings, and
electronic intruder alarm systems, are a
normal specification in New Zealand’s
medium density housing schemes.
Car Parking and Storage
Restricted parking and storage space for
privately owned vehicles is inherent in the
typology of medium density housing,
representing one of the most significant
differences between it, and lower density
suburban housing. The loss of security of a
vehicle parked ‘not within the curtilage’,
and consequent loss of amenity
compounds the difference.
From the literature, and the case studies
(Section 5), it is apparent that standards of
parking provisions vary, reflecting a
dilemma at the heart of medium density
layout design. Minimum ratios are required
through District Plans but are commonly
exceeded by developers, particularly in the
private sector. The desire to increase both
proximity and total parking provision is
evident in examples from all countries, and
at all densities.
The vehicular environment has a dominant
role in many examples of the typology.
Low speed internal roadways are regarded
as preferable, with reduced street widths
also possible if measures are taken to
ensure pedestrian safety, (King, Rudder,
Prasad and Ballinger, 1996 p66), and a
maximum of 30 houses are served by the
road. The Dutch Woonerven system (a
‘residential precinct’, in which pedestrian
priority is assumed), although designed for
urban regeneration developments, sets a
relevant standard for medium density
housing by combining landscaping, public
25
space, and traffic in a mixed environment
(Colquhoun and Fauset, 1991). One of the
more successful examples in Australia
(Moverly Green, Coogee, Sydney; not used
as a case study) achieves an acceptable
level of safety with narrow drives and
without footpaths.
In the New Zealand context the parking
issue also reflects differences between the
main urban centres: Auckland has a road–
based transportation system and the typically car–oriented culture of a low density
city, with lower levels of use of public
transport than Wellington or Christchurch.
In addition, climatic differences, particularly Auckland’s high rainfall, encourage
planning that locates the car in close proximity to the house.
Medium density housing has developed in
other countries with localised variations
for parking and traffic design, usually with
less car dependency than observed in the
case studies included in this report.
External Style
Speculative housing development has a
long history of modifying and adapting
existing architectural styles to meet
perceptions of market preferences. The
speculative industry also takes a cautious
approach to all aspects of housing development, including architectural expression or
image, preferring a tried and trusted model
before an innovative one as a matter of
course. Investors as well as developers are
risk averse, placing high value on
achieving the optimum density for the
perceived market, on minimising construc4
tion costs, and on street or ‘kerb’ appeal.
Comment in the literature consistently
refers to the need for affordable housing to
be indistinguishable from other housing. In
particular, rental housing in the public
sector should be as similar as possible to
private sector housing in the same
neighbourhood.
26
With regard to design style, this study
recognises that New Zealand architecture
in medium density housing cannot be fully
represented by examples selected entirely
from the Auckland region; other cities have
developed models in the typology that add
significantly to the body of relevant work.
Within the limited range of examples, and
geography, of this study, the developments
illustrated therefore provide a partial but
not complete picture of the issue of
external design.
Most of the case studies in this report illustrate architectural forms that reflect
commonly held ideas of domestic building.
Both developers and the public seem
prepared to accept imported domestic
vernacular architecture in some form, with
European influences most widely used.
Styles vary widely. The architectural
variety contributes to the strategy, widely
adopted, and regarded as critical to
success, that seeks to disguise the differences between medium density housing
and lower density suburban housing.
Medium density housing is generically a
repetitive typology: stylistic variation
within a general theme (‘Spanish colonial’,
‘French rural’, etc.) conceals repetition by
allowing building detail to be read, by use
of colour to differentiate one house from
the next, and by variation in form, reducing
perceptions of mass. In the best schemes
the perception of anonymity in the ‘mass’
of a large development is replaced by clear
identity of the parts, and the single unit
within the part. A greater mix of dwelling
types is also a perception (but not always
the reality) generated by stylistic variations.5
Stylistic variation occurs across all the
layout classifications: this study found no
apparent correlation between style and
density band, or style and market sector.
More expensive facing materials tend to be
used in the higher priced developments,
which in some cases has led to an architectural style associated with a particular
Best practice in medium density housing design
design ‘school’: Beaumont Quarter (case
study 26) is an example.
A complete catalogue of stylistic influences is beyond the scope of this report, but
a short summary of the principal variations
is considered useful.
The Arawa Road project (case study 10) is
arguably the closest design to a recognisable New Zealand architecture, by form,
simplicity, lack of self–conscious expression or reference to a foreign vernacular,
and choice of materials.
Some of the larger developments in West
Auckland illustrate the high degree of
design licence possible in the typology. At
the Corban Village development (case
study 4) each sub–section of the layout is
architecturally distinctive, including the
following: undecorated modernist externally plastered three storey houses differentiated by colour, with no particular
theme; two storey Breton terraced cottages
with quoins, window architraves and
reveals, and parapets at the party walls;
traditional Dutch decorated curved gables
and party wall profiles; and a group of Art
Deco houses with streamlined curved
corner windows, again using colours of the
style to distinguish one unit from another.
The Harbour View development exhibits,
amongst other styles, many variations
based on the Spanish Colonial style, (eg.
Gunner Drive, case study 14), as do St
Georges Road (case study 12), Sacramento
(case studies 17 and 27), and several of the
North Shore schemes. Others explore
vernacular architecture from England
(Melview Place, case study 8), Italy,
(Tuscany Towers, case study 7) or draw on
late modernism to express complexity,
variety, and difference (Romola Street,
case study 6, and Beaumont Quarter, case
study 24).
In spite of the great variety of style there is
little sense of ‘theme park’ architecture in
these developments, and a strong sense of
free market choice. Medium density
housing design in Australia has not
A New Zealand Definition of Medium Density Housing
generally experimented in a comparable
way, preferring traditional, less exuberant
residential styles that understate rather than
celebrate diversity.
Summary
This section examined key design issues
based on relevant literature and current
27
practice in medium density housing. This
section serves as a platform for the
following case studies which provide a
more detailed review of contemporary
6
medium density housing in New Zealand.
28
Best practice in medium density housing design
ENDNOTES
1
2
contemporary use of domestic space, often
including work from home room(s), storage
for recreational equipment, and with up to date
services for electronic uses as well as
bathrooms and kitchen. A thirty year old
house originally built for a modest market
price will frequently be more cheaply replaced
than modified to meet current lifestyle
requirements.
In the light of the debate regarding the
usefulness or otherwise of ‘density’ as an
indicator, and the numerous differently
defined bases for calculation, it is hardly
surprising that inconsistencies occur in the
literature when density figures are used in
comparisons. For instance, the Auckland
Regional Council publication Urban Area
Intensification (Auckland Regional Council,
2000e) referring to the AMCORD ‘net
residential density’ term, lists and illustrates
several projects of apparent relevance to this
study but states density figures that place them
outside the range of 30–60dph.Using the
AMCORD methodology and revised
calculations, most of these schemes do in fact
coincide with the density range considered
here.
In the process, the opportunity to change the
architectural style is usually taken. For this to
be possible, the house itself has to be
physically and architecturally independent of
it neighbours. In medium density housing this
independent condition is not usually possible.
For medium density housing to be part of the
same housing market in which rapid
redevelopment is a regular market activity,
different, and stylistically indeterminate
models may need to be evolved. This tentative
conclusion may lead to two other issues
relevant in the New Zealand context:
In a chapter entitled “Concepts of Privacy”,
Darroch refers to Altman’s six definitions of
privacy, of which two are quoted here:
“1) privacy is essentially a matter of
person/environment transactions, that
is, it is a dialectic or dynamic system—
it is not a static event or state;
(a)
the custom in New Zealand society of
constant do–it–yourself alteration of the
home;
(b)
the development of a separate housing
type (medium density housing) which does
not lend itself to alteration, either because
of inflexible design, or because of the
controls imposed by a management
structure
representing
community
ownership.
2) complete definitions of privacy need
to take account of the critical role of
‘control’ in the understanding of
privacy.”
Altman, 1975 (quoted by Darroch, in
Judd & Dean, 1983)
3
4
New housing developments in many countries
including New Zealand reflect this concern,
opting for auto–gated compounds, (as in case
studies 13 and 24), electronic alarm systems
fitted during construction, and/or heavily
defended ground floor openings, and
sometimes upper floors also. There is
evidence that insurers, having met a claim,
will demand higher specifications for locks,
doors, and alarm installations, as a condition
of re–insurance.
At the same time, novelty is often welcome in
the marketing process. In some instances,
including the development of medium density
housing,
design
is
architecturally
experimental, evidenced by contemporary
housing design in New Zealand.
There is an increasingly common pattern of
suburban re–development in North American
and Australian cities, where houses are
demolished after 25 years, to be replaced with
a ‘new model’, usually much larger, with a
different plan configuration reflecting
In the process of identifying a design model
for the New Zealand context, both of these
issues may need further consideration.
5
In these circumstances the extreme variety of
external design in medium density housing in
New Zealand is a phenomenon for which
several explanations are offered:
(i)
The generic single storey, suburban
detached house is not adaptable to higher
densities; a different type of building has to
be generated to meet the typological
requirement of medium density housing;
(ii)
There is a pronounced need
distinguish between developments,
order to establish identity, both for
buyers and residents, and for
developers;
(iii)
Stylistic definition can establish
certainty of product for investors and
funding institutions;
(iv)
Cost estimating, adjusted to the new
product, can be consistent and accurate;
to
in
the
the
A New Zealand Definition of Medium Density Housing
6
(v)
A broadly open-minded, or modernminded public encourages more rather
than less experiment with external style
than the industry offers; this is in contrast
to the conservative styles seen in the same
market in the United Kingdom, and is
paralleled by design in Australia;
(vi)
Evolution of the notion that housing is a
commodity governed by the same market
rules that apply to other commodities:
housing is less short term (as a personal
investment)
than
other
domestic
‘durables’, but still a commodity
possession.
The house building industry has been affected
by the problem called ‘leaky buildings’ since
2001 when the consequences of construction
using monolithic plastered cladding systems
fixed to untreated timber framing were first
detected. The movement of framing timber after construction, often due to shrinkage following drying out, causes cracking in the
external wall surface allowing water to enter
the cavity within the wall. If the wall has been
built without the means by which such water
can drain from the cavity the untreated framing starts to rot, leading, eventually to structural failure of the wall. This failure may occur
in a short period: a few months is not uncom-
29
mon. Many schemes included in this review
are affected by the problem, since the cheapest
cladding system able to gain approval from the
central and local building authorities is attractive to developers. The monolithic systems
align readily with the stylistic preferences of
developers and the buying public: various
‘European’ styles in particular the ‘Mediterranean’ styles rely on some form of stucco-like
finish to the external walls, and other details
that suit dry hot climates.
The Building Research Association of New
Zealand (BRANZ) issued appraisal Certificates for numerous proprietary cladding and
finishing systems of this type after 1994; the
Building Industry Authority (BIA) accepted
untreated timber for external and internal construction in 1997. The period of development
of most medium density housing in New Zealand has been subsequent to both these dates,
thus affecting much of the housing built.
The ‘leaky building’ issue is local to the New
Zealand building industry, and is considered
to be a technical matter relating to construction rather than a systemic issue in medium
density housing. It is, however, associated
with this housing type in the press, and therefore in public perceptions of higher density
housing in general.
4
Case Studies:
Methodology
and Criteria
32
CASE STUDIES: METHODOLOGY
AND CRITERIA
Introduction
This case study examines contemporary
medium density housing with particular
reference to the relationship of density to
amenity, internal and external space standards, and access to the private car.
It is acknowledged in the literature and
amongst design professionals in housing
that as density increases, compromises
affecting the quality of the residential environment accumulate.
Density is considered to be a performance
indicator in all ‘after the event’ analyses of
housing developments. In New Zealand, as
elsewhere, the proximity of the private car
(and its security) is regarded as a secondary
performance indicator, for reasons outlined
in Sections 2 and 3. Site layouts that
provide similar car access validate comparisons between schemes at different levels
of density.
House types are also directly affected by
the density scale, reducing options for
frontage widths, access, aspect, and
internal planning as density is increased.
The analysis therefore aims to evaluate
New Zealand examples, to:
(i)
track the pattern of compromise as it
occurs for different levels of density
and layout types;
(ii) identify changes in the quality in the
residential standards achieved,
referring to internal planning and
external space standards, for
different levels of density and layout
types;
(iii) record and establish a database of
quantifiable evidence to represent
key aspects of each scheme relative
to density; and,
(iv) assess the physical environment of
medium density housing relative to
lower density housing.
Best practice in medium density housing design
Methodology
A methodology to select and critique
examples was developed from Australian
and British models, some of which are
referred to in previous sections.
British literature makes frequent use of the
term ‘curtilage’, distinguishing between
‘within’ the territorial boundary of a property (curtilage), and ‘not within’ the
curtilage, specifically in relation to car
storage and parking. To construct a basis
for valid comparisons different site layout
types have been classified, following this
model, to recognise the distinctions
between layout amenity to householders in
terms of car access. The methodology
separates, therefore, layout types by vehicular proximity, to acknowledge the
amenity factors of security, and access in
use; and in the New Zealand context, to
acknowledge the local influence of
climate.
The layout classifications are defined as:
Type 1:
front access to the house, with
the car internally garaged within the
house type, or provided with a
carport or parking space within the
property boundary.
Type 2:
rear access to the house;
secure parking, as defined in Type 1.
In New Zealand’s relatively informal
society rear access is a common
habit: the ‘back’ door does not
represent a high level of social
familiarity.
Type 3:
front or rear access with the
car parked outside the property
boundary: called, for convenience,
‘remote’ parking, and including car
parking adjacent to the house in a
space controlled, and possibly
owned, by the unit but also
accessible from a public area and
therefore not secure.
Type 4:
layouts dependent on the three
storey house type with internal
garaging.
Case Studies: Methodology and Criteria
Site Selection
Initially, 60 examples were listed for
consideration, including schemes reviewed
in other studies, without reference to layout
type or density where known. Sites considered included composite or hybrid layouts,
often combining Types 1 and 4. There are
few ‘pure’ examples of Type 2 in the
Auckland region (where all the New
Zealand examples are located).
Two storey house types dominate in the
density range studied, with the three storey
elevated living area house type used in
some examples; this option is considered
sufficiently common for the case study
selection to include a small number of
examples for comparisons. From the original list of 60 schemes, further criteria were
established to identify representative
schemes covering the principal layout
1
types.
(i)
Size: schemes of less than nineteen
units were discounted: in smaller
projects it was considered that
variables of shape of site, location,
and layout to density characteristics
increase significantly, and affect
validity of comparisons. Case studies
6, 15, 16 and 31 are included to
illustrate a particular layout
characteristic, although smaller than
the preferred lower limit.
(ii) Schemes of interest for reasons of
layout type or density, with resource
consents granted but not yet built,
were included: Holly Street,
Avondale (case study 22), and Mt
Taylor Drive, Glendowie (case study
12).
(iii) Schemes previously included in
other studies are generally excluded,
with the exception of Arawa Street,
New Lynn (case study 10), and part
of the Sacramento development, East
Tamaki (case studies 17 and 27).
(iv) Density: schemes at densities higher
and lower than the range identified as
33
‘medium’ density (30–66 dph) were
included to provide comparisons.
(v) Value: a significant variable
observed, described in the data chart
as a ‘market level’, was seen to
impact on design options. This has
been recorded as a factor influencing
layout design, in some instances
indicating an explanation for the
choice made.
(vi) Affordability: a general preference is
expressed for private sector
developments at low and middle
‘market levels’, in response to the
focus in this report on affordable
housing. Public sector schemes are
also reviewed, for comparisons.
(vii) Quality of environment: schemes
were selected primarily to illustrate
the critical relationship of density to
layout, rather than perceptions of
residential quality. The quality of the
environment achieved is determined
by density conditioned by other
choices made in the scheme
including house types, facing
materials, landscaping, utility and
servicing design, and the provision
of public open space.
(viii) Management: the existence and
effectiveness of Body Corporate
management schemes affects many
of the developments reviewed;
selection has not excluded such
schemes, permitting gated examples
to be included for comparison of
layout types.
The final list yielded 34 examples,
including four Australian schemes,
drawing on the Australian experience in the
typology. It would be misleading to suggest
that these developments are representative
of the average standards achieved in
Australia, because they are not. They do,
however, demonstrate achievable standards, and are selected as examples at two
different density levels, under 40dph (in the
lower range) and above 60dph at the high
end of the medium density range. These
34
four developments illustrate established,
well–regarded, and high quality housing in
which market performance has paralleled
or exceeded similar developments.
Location
Excluding the Australian examples, half of
the case studies are drawn from Waitakere
City. This is partly due to time limitations
on the report, and to the ease of access to
data (and the lower costs of retrieving data)
in the Waitakere City Council procedure.
The process of selection also took into
account the medium density housing study
carried out four years ago by the Auckland
Regional Council (2000c), which documented nine projects, including four on the
North Shore, only one of which (Coroglen)
is located in West Auckland.
A further justification for the use of West
Auckland examples lies in the perception
that many, perhaps the majority, of medium
density housing projects in Waitakere are
set at a low or medium point in the market
scale, and therefore gain relevance to a
study focusing on this typology as an
affordable housing proposition.
Multi-development Sites
Three of the West Auckland examples
(Ambrico Place, Corban Village, and
Harbour View) are large sites that have
been parcelled into smaller sites to attract
commercial development: from the overview of the study, it seems that few
commercial house builders are prepared to
take on a single project of more than 100
units. These three larger developments
have yielded eight examples between them,
providing opportunities for useful comparisons of different layout and house type
options, within a single location, and to
some extent, market.2,3
Methodology
The methodology involved visiting
Council offices to obtain scale plans and
Best practice in medium density housing design
details of the main house types used in each
development. From a pilot exercise it was
found that this data yielded sufficient material to quantify density, total floor space,
floor area ratios (FAR), site coverage footprint, parking ratios, and to identify house
types as percentages of totals. The pilot
study also revealed that small variations—
where extra but numerically insignificant
variations such as modified end unit plans
occurred—had little effect on density or
the FAR, and were therefore not quantified
in the assembled database.
All schemes selected were visited and
photographed. The methodology used
included scanning scale drawings to
provide data by digitally isolating built and
non–built areas, road areas, public open
space, and private gardens and patios. Site
areas given in City Council records, or
taken from dimensions and bearings on
survey drawings were checked by this
method where a simple arithmetic check
suggested the possibility of error, or where
site areas given did not align with the
preferred base data for density calculation.
Topographical Criteria
Since severe slopes tend to distort other
factors, developments have been selected,
as far as possible, to be comparable, with
flat or near flat sites taking priority. Where
slope is significant to the layout design a
note is made in the accompanying description, but is not otherwise indicated on the
thumbnail plan.
Value and House Types
From site observations and, in some cases,
local real estate enquiries, an assessment
was made of market position. House types
are described and discussed in the notes
with each case study to establish a generic
relationship between house type and layout
classification. This is necessary to eliminate—as far as possible—disadvantage to
very low–cost schemes and to identify
high–cost schemes, and is recorded as an
Case Studies: Methodology and Criteria
approximate indicator of market value in
column 19 of the data chart, on a scale from
1 (low market) to 5 (high market). The
number given on this scale is not quantified, but is generated by knowledge of the
original property sale price, where known,
or by the developer’s expectation of sale
prices, estimated by building detail and
location. The purpose of the scale is therefore to indicate the market ‘intention’,
which is regarded as a relevant item of
information in the assessment of the
quality of the environment achieved relative to levels of density and layout type.
Refuse Collection
Refuse collection is referred to in the literature as a significant factor in determining
the acceptability of higher density housing.
It is apparent that some developers take
care with this matter, and others do not. An
unregulated refuse system, in some cases,
severely undermines the development’s
potential, damages the locality beyond the
site itself, and supplies a strong argument
to reinforce public prejudices against
increased densities. In the best schemes the
process of refuse storage and collection is
virtually invisible.
Kerb–side collection from individual properties is the preferred option for a high
quality residential environment. Storage of
refuse inside the unit curtilage needs to be
planned carefully for reasons of hygiene
and practicality in the functioning of the
household, in the best examples taking the
form of an external enclosure with an
external route to the collection point.
Building roads to ‘adoptable’ Local
Authority standards is expensive in
construction costs and in site space at
higher densities, and in some cases is not a
practical option for reasons of access.
Many of the schemes visited have a refuse
enclosure at the site entrance where refuse
is deposited by residents, a workable and
hygienic solution up to a maximum of
about 25–30 houses. Inorganic collections,
35
although infrequent, may have to be tolerated as an annual event, as they are, in less
concentrated forms, in the suburbs.
In larger developments, with few roadside
entrances and collection points, the street
impact of scores of bins, and the difficulty
of identification, is very considerable.
In other schemes where development has
been carried out behind houses on an
existing road frontage, kerb–side collections are sometimes seen to cause unacceptable weekly conditions for those
houses. A minimum requirement for back–
land sites should be a ‘compound’ roofed
enclosure, with subdivisions to avoid an
excessive agglomeration of bags, located
behind the front property boundary and
screened from the street. For soft collection
systems based on polythene refuse sacks a
maximum number of units served should
be established if roadside (not internal)
collection is necessary. This number
should not be more than ten.
Washing/drying Arrangements
Site visits were conducted in good weather
in June and July 2004. Observations
confirmed that external clothes–drying is a
common preference but not always a
straightforward option for householders. In
some instances ad hoc clothes drying
arrangements occupied front gardens,
using various semi–permanent lines sometimes fixed under balconies, while retractable lines and collapsible racks are
common. Such arrangements reflect the
small, and often shadowed rear external
spaces (case studies 10, 13, and 24 provide
some examples).
Open air clothes drying is also a long–
standing tradition in New Zealand households, and should be provided for wherever
possible in all housing with ground level
access to private open space. Site planning
to ensure even small rear yards with orientation to allow some solar access is
possible up to approximately 60–70 dph,
and at higher density levels if the
36
development incorporates underground
car–parking.
The case study commentaries discuss other
factors that affect the overall quality of the
residential environment including management by body corporates, where relevant,
and communal facilities, where provided
4
by the developer.
Best practice in medium density housing design
Case Studies: Methodology and Criteria
37
ENDNOTES
1
from the developer for associated infrastructure costs, but the bond payment took the form
of security against certificates of title on unsold houses in the scheme. The Waitakere City
Council was not the first mortgagee on the titles, effectively making the bond a debt to the
Council alongside other unsecured creditors.
Mr P Brown, Waitakere City Council Resource Management and Buildings Service
manager said the arrangement at Tuscany
Towers was unusual, in that neither a cash
bond nor a bank guarantee was required from
the developer. Because medium density, in
this case on a large scheme of 97 units, normally cannot avoid unit titles (rather than the
standard sub-division freehold title) the developer’s contribution cannot be ‘staged’ across
the financing of the project in smaller increments; the cost of a long–serviced bank guarantee is high for the developer, who is
dependent on sales and contract completions
over a longer period than normal in suburban
sub–division developments.
The Tuscany Way site (case study 16) forms
the southern boundary of the Edgelea block of
43 houses. The block is one model for
greenfield medium density housing, planned
as a perimeter of outward facing (front access)
linked, or detached houses enclosing two
garage courts serving rear accessed units not
The Edgelea block site plan
It would seem that in this instance, in order to
encourage the development (as a landmark
medium density project, amongst the earliest
in West Auckland) the Waitakere City Council took a step back from their usual bond requirements (Western Leader, Thursday 1 Nov
2001 p1 (Tuscany Towers, New Lynn)
“Caught in Collapse”).
located on the block perimeter, in this case a
courtyard type. A small semi-private ‘pocket’
park with no vehicular access but accessible to
emergency traffic is defined by the frontages
of eighteen units occupying the core of the
block in three separate developments.
A similar hybrid layout is used in the Oatlands
development (case study 18), with a similar
intention: to provide variation in house type,
price range, and to gain density.
2
Ambrico Place, New Lynn. The Ambrico
Place development occupies land previously
used for industry, including a brickworks
serving the local district of New Lynn; the site
has been re–built since 1996 as the first larger
scale medium density housing in Waitakere
City. The development now consists of approximately 350 houses. There have been nine
separate developers involved, all except two
using architects for the layout design.
Each parcel is different in architectural style
and there are significant differences in layout
principles, and in relational possibilities, that
are reflected in varying densities (see case
studies 25 and 29). Three of the Ambrico
Place developments have used the narrow
frontage dual aspect three-storey townhouse
plan form. One of these is reviewed (case
study 29).
3
Tuscany Towers (case study 7): Waitakere
City Council required a bond of $485,000
4
It is noted that some Body Corporate
management schemes in higher density
developments ban external clothes drying,
requiring occupiers to use only tumble dryers.
The same restriction is applied in some
medium density developments, to protect
external appearance from the domestic
intrusion of washing. At densities between
30dph and 66dph these restrictions are not
necessary, although at the upper end of the
band, as case studies show, private open space
becomes increasingly difficult to achieve.
5
Case Studies
40
Best practice in medium density housing design
Case Study Conventions
Case Study Data
Each site is illustrated with a thumbnail
sketch plan showing the distribution and
orientation of buildings, the spaces
between them, and the organisation of
roads, access and parking. Where
communal or public open spaces are
significant the area is indicated by a diagonal line. In some instances the sketch
plans are simplified to clarify the layout
type; the authors acknowledge a small
degree of injustice to the designers in such
cases.
On the data table (p77), case studies are
numbered and named in the left hand
columns. Columns 1 and 2 list basic data
describing development size, with other
relevant data for date and place in columns
17 and 18. Column 20, groups schemes
according to the four layout classifications
employed in this study (see p32). Where a
scheme uses more than one layout type, the
secondary type is indicated in brackets in
column 20.
Case studies are presented with the
following conventions:
(i)
Columns 3 and 4 then arrange the schemes
in ascending order of density within the
layout type. Integral garages are included
in the unit floor areas where they occur.
Sketch plans are diagrammatic, to
illustrate the scale and form of the
scheme, with North point to the top
of the sketch. The plans are not to a
given scale.
A summary of statistics is included
with each study.
Architects are credited, where
known.
Case studies 13, 15 and 28 are
Housing New Zealand Corporation
owned developments.
Case studies are presented in
ascending order of density in each
layout type;
Case studies 25, 26 and 28 use
underground
garage
parking,
indicated on the sketch plans by a
broken line.
Parking: the total parking provision is
given as a ratio of car spaces per unit,
including visitor and casual parking. In
some instances there are variations
between approved plans and the development ‘as built’, resulting in under-sized
parking spaces in front of garage doors:
where this has been noted from site visits,
the under-sized space is not counted in the
total. Columns 12–16 all quantify other
aspects of the parking and vehicular access
arrangements.
Where possible, the schemes reviewed
have been selected in groups to minimise
the effects of differences between locations, particularly Glendowie (3 schemes),
Botany Downs (3), and Ambrico Place,
New Lynn. The numerous candidates
around central Auckland were reduced to
one, Beaumont Quarter, to avoid higher–
end examples that may benefit from a
developer’s willingness to invest more in
building costs in anticipation of higher
returns or faster sales.
Exceptional or non–standard figures are
noted as follows:
(ii)
(iii)
(iv)
(vi)
(vii)
The Floor Area Ratio (FAR) (column 5)
indicates the density of the development as
a ratio of total floor space to site area. The
figure generally rises with increasing
density, reflecting increasing footprint, and
balances or off–sets the variations in unit
sizes.
(i)
(ii)
Tuscany Towers, column 6: 2.44*,
and column 13: 143**: figures
include three storey units with four
or five parking spaces available in a
lower
ground
floor
garage/
workshop, slightly raising the total
parking ratio for the development.
Tuscany Towers, column 8: includes
the large lower ground floor in the
total site footprint.
Case Studies
41
(iii) Sacramento 1A, column 9: excludes
carport roofs; as with similar
schemes (case study 24, etc).
7.
Case Study Evaluation
environment and house unit, and by
perceptions of personal territorial
ownership.
method of, and arrangements for,
collection of refuse.
The criteria that determine the quality of
the residential environment in New
Zealand's medium density housing developments are identified and discussed in the
preceding sections. Although the relationships between the criteria are complex, and
vary between developments according to
specific factors, including site topography
and shape, and marketing intentions, the
principal criteria are identified as:
The case studies selected display some
characteristics typical of the typology in
relation to more than one of these summarised criteria. Each study is therefore
accompanied by a table of seven sections
corresponding to the criteria listed above,
indicating positive, negative and neutral
resolutions of the relevant issue.
1.
Terminology or abbreviations used for
convenience in the case study analysis and
the data chart include:
2.
3.
4.
5.
6.
quality of the public environment
within the development, defined by
function,
landscaping,
and
maintenance.
quality of, and provision for private
open space defined by convenience
of access, privacy, and capacity for
extended domestic uses, including
out-door meals, washing, children's
play area, and recreational gardening
activities.
standard of privacy achieved,
defined by overlooking and by
perceptions of crowdedness.
standard of private vehicle parking
achieved, defined by convenience of
proximity and access.
standard of identity achieved,
defined
by
perceptions
of
individuality within the whole
development.
standard of security achieved in the
detailed design of the physical
Glossary
Dph:
dwellings per hectare; also
abbreviated in the literature as
‘dpha’, ‘DpHa’, and ‘Du/Ha’. See
Section 4, defining density for a
fuller explanation of the term.
FAR: Floor Area Ratio: also referred to
elsewhere as Floor Space Index
(FSI), Floor Area Index (FAI); used
here as the reference for the total
floor space built as a ratio of the total
site area.
Parking ratio: the ratio of total car parking
provision to the number of dwellings
in the development: where the ratio
figure is less than 2, there are fewer
than 2 car parking spaces per
dwelling in the overall development.
42
Best practice in medium density housing design
(1) VINOGRAD MEWS, HARBOUR VIEW,
WAITAKERE CITY
Vinograd Mews is a small development of nineteen houses
at a density of 33dph. This development employs terraced
housing but with compromised amenity in comparison
with traditional suburban housing. In this instance the site
itself is also a challenging shape.
Street frontage is an important requirement in the Harbour
View strategy, and is used to advantage in this layout. Two
storey units with integral garages address the street with
small set–backs and vehicle crossings at over–frequent
intervals: however, the street is positively defined, and
without domination by garage doors due to the recessed
plan detail.
To retain the highest possible density, the core of the site is
planned with the balance of units permitted, causing a
sense of crowding.
The ‘Z’ plan unit with a cross wall dimension of 9.7m and
an overall length of nearly 15.0m is not an efficient house
type for this purpose. Internally the planning makes
considerable effort to avoid habitable rooms on both sides
of party walls (built in 200mm concrete blockwork) at both
floor levels. Other details suggest further problems with an
unfamiliar house type: the third bedroom on the ground
floor has to have a separate bathroom, not located
practically for use as a ground floor toilet; upper floors are
windowless on the back, to reduce overlooking and satisfy
height to boundary regulations, making the rear elevation a
featureless wall; and a TV position under the stair cannot
be viewed by any practical arrangement of living room
furniture.
These compromises are reflected in the site planning, in
which the density achieved is not justified by the crowded
environment of the space between the houses. The scheme
illustrates many of the issues confronted by designers
working in the medium density housing field.
Architect: Grant Neill
open
space
(public)
open
space
(private)
-
-
+
no units
parking
ratio
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
19
2.20
privacy
5,742
parking
179
identity
302
refuse
security collection
33
.59
Case Studies
43
(2) ADELPHI VILLAS, EAST TAMAKI,
MANUKAU CITY
For reasons of typicality this project is included to represent a housing form that minimises the value of public
space in order to gain density and private garden area.
There are numerous developments of about this size and
type in New Zealand: at this density (two storey housing in
semi–detached units, 37 dph and a FAR of 0.52), the type
has become a standard product in the market.
The quality of the residential environment is necessarily
compromised in this layout type, a factor most apparent in
the service access. Garage doors dominate, property
boundaries are indicated by concrete strips set into the
(otherwise uniform) tarmac surface, planting is insignificant, and security measures are the dominant feature in
detailing. Parking occupies all available space adjacent to
the internal road, its appearance made more unsightly by
an irregular arrangement which conveys an impression of
haphazard use, lacking ownership or organisation. Refuse
is collected from an enclosure (not roofed) at the site
entrance, fronted by letter boxes.
Overlooking remains a problem in spite of attempts in the
planning of the site to protect privacy. The elevation to the
distributor road to the north of the site presents a wholly
suburban identity.
Architect: Alan Rolston Residential
open
space
(public)
open
space
(private)
-
privacy
parking
identity
refuse
security collection
-
no units
parking
ratio
30
2.30
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
8,135
142
271
37
0.52
44
Best practice in medium density housing design
(3) SEYMOUR ROAD, SUNNYVALE,
WAITAKERE CITY
The development occupies land not previously built on,
close to the Manui rail stop and the Parrs Park recreation
area in West Auckland. The layout is a hybrid, with the
majority of units either detached or linked detached
sharing only the party wall between garages, and front
access, dual aspect houses. This contributes to a density of
37 dph, a level at the lower end of the medium density
range. The internal road is a public street in a compacted
version of a traditional suburban layout. The development
loses most of the benefits of suburban layout design
without gain in any area. As a residential environment, this
approach has little to recommend it. No public open space
has been included, perhaps reflecting the amenities close to
the site. The scheme makes no concessions to recent good
practice in higher density design, or to New Urbanist
theory, or to the potential of urban housing to contribute
lively neighbourhoods as part of the intensification
process. A landscaping scheme, designed by Sinclair
Knight Mentz, which would improve the quality of this
development, has not been implemented.
The development has been built to attract investors in
rental property, providing an explanation for the variety of
separate house types used. To ensure market diversity a
small group of thirteen units has been arranged around a
rear access garage court, locally increasing density and
introducing a variation on the otherwise comprehensively
suburban theme.
The scheme is included in this study to provide evidence of
the need to recognise the difference between suburban (in
the Auckland and New Zealand traditions) and medium
density housing design. A compacted version of low
density housing, such as this, cannot achieve the potentially excellent residential environments of either suburbia
or medium density housing.
Architect: Fuller Design
open
space
(public)
open
space
(private)
no units
parking
ratio
89
2.08
privacy
parking
identity
-
+
-
refuse
security collection
+
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
24,600
118
273
37
0.43
Case Studies
45
(4) CORBAN VILLAGE, HENDERSON,
WAITAKERE CITY
The layout is based on two separate design principles: a
front entry type using two and three storey house types,
and rear entry predominantly with two storey house types,
and, due to adjoining public open space, no provision for
internal communal areas. The plan includes an adopted
road which provides service access and refuse collection.
All units have one secure car space, and the majority have a
second space within view from the house.
The whole development was packaged into approximately
six developments, evidenced by architectural variety that
removes any sense of uniformity or repetition, best illustrated in the central (rear access) group served by a private
access driveway. In this group the north–south orientation
raises the question of the inactive entrance on the south
side where recessed ‘front’doors are not in use in all cases.
Terraced three storey townhouse types as built on the
north–western boundary are not as articulated in plan as
indicated on the original drawings, but succeed in
enclosing this edge of the development and retain a lively
street elevation. The same three storey type used in short
terraces north of the internal road are detached from the
rest of the development by their own paved forecourts.
Although the house type is justified by a south–facing
slope on this site, it also generates a tarmac and car–dominated environment at ground level, excessively so at this
density.
Overall, the development illustrates the quality of a residential urban environment possible at this density without
sacrificing access to and security of the car.
Architect: Various
open
space
(public)
open
space
(private)
no units
parking
ratio
83
2.20
privacy
parking
identity
+
+
refuse
security collection
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
20,686
249
40
46
Best practice in medium density housing design
(5) FAIRHAVEN, GLEN EDEN, WAITAKERE
CITY
This scheme was started in 1999, and is now being
completed with 98 units. The layout design is comparable
to Seymour Road (case study 3), and the Corban Village
(case study 4) development.
Density depends on the use of 26 three storey townhouse
units on the perimeter of the site. These have a ground floor
room behind the garage, accessed, in this variant, by a
corridor alongside the stair leading to living accommodation on the first floor. A narrow plan is further expressed by
dividing the front elevation into two, giving a strong
vertical emphasis.
The core of the site, of two storey dual aspect units, is
developed around a public loop road providing access for
refuse collection and other services, similar to two
previous schemes. Two small areas adjoining the road
provide public open spaces, including a play area, at the
lowest levels on the site, next to a stream. Back to back
dimensions are minimal for the house type, leading to a
rear garden environment of a heavily fenced and enclosed
warren of private spaces where overlooking is a significant
issue.
Short terraces of three units have been used to maximise
side access to rear gardens, and to increase the number of
‘end’ units, seen by the market to have higher value. The
market, in this case, is likely to be sales to investors, a
factor that generally deters the dense planting which would
be necessary to both improve the public side of the development and reduce over looking in the private garden
spaces.
Architect: Harrison Grierson Consultants
open
space
(public)
open
space
(private)
privacy
parking
identity
-
-
+
-
no units
parking
ratio
98
2.30
refuse
security collection
+
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
24,728
141
252
40
.56
Case Studies
47
(6) ROMOLA STREET, GLENDOWIE,
AUCKLAND CITY
This small development is part of the regeneration of the
Madelaine Avenue area of Glendowie, where new residential property is replacing dilapidated housing stock and at
the same time lifting density levels. (See also Mt Taylor
Drive—case study 11). The project stands at the top of the
market value scale in this study.
The site plan illustrates two relevant points:
(a) an architectural intention to propose higher density
housing without loss of a modern tradition to treat each
building as an object of design quality in its own right;
(b) the extreme reduction of space outside the separated
houses, a consequence of pursuing a market goal of
building detached houses, and planning the development at
this density.
Two pairs of houses, shown as attached in the plans, also
appear to have been built as single units. All the buildings
are strongly articulated by form and by facing materials
used, to such a degree that the site plan sketch is highly
simplified. Access is from private culs de sac serving up to
five units, dimensionally minimal so that the shared access
function is only possible if rigorous discipline in use is
maintained; power steering, and medium size rather than
large cars are necessary.
Internal planning of the houses is conventional and also
reflects new domestic uses of space by layout and spatial
diversity. The total average floor area, including the
2
garages, at 176m indicates a relatively large unit size. The
FAR is 0.73, also a high figure for this layout type, contributing to perceptions of a crowded plan. Privacy between
house units is inevitably very poor, particularly between
external private spaces, and upper floor rooms look
directly into opposite units. At this density a different
layout type would resolve these and the access problems,
but would not permit a detached unit design.
Architect: Powley Architects
open
space
(public)
open
space
(private)
-
privacy
parking
-
no units
parking
ratio
13
2.10
identity
refuse
security collection
+
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
3,140
176
242
41
0.73
48
Best practice in medium density housing design
(7) TUSCANY TOWERS, AMBRICO PLACE, NEW
LYNN, WAITAKERE CITY
Tuscany Towers was the first and largest stage of the development in Ambrico Place. The scheme of 97 units includes
a tennis court and a public ‘square’/community space,
marked by a tower, which also houses the communal television aerial. The internal streets are also public spaces,
providing extra non–allocated parking: the public domain
is thus represented at several hierarchical levels. The
public areas are included in the density calculation.
The architecture is uniformly ‘tuscan–suburban’,
including details of ornament, colour palette, and variations of height forming a coherent, consistent, and knowledgeable example of the genre. The majority of houses are
two storey three bedroom terraced units with garages
accessed internally, the layout and the street articulated by
three storey four bedroom houses at corners and junctions,
using a plan form that provides accommodation at ground
floor level for living or business use.
A storey–height step inherited from former use of the site
on the east site boundary introduces a third variation of
larger units on a platform over a four car garage, offering
live/work options to some residents. Access to these units
from street level is via an ornate tiled stair shared by two
adjacent houses, and also from the external public road by
a second ‘front’ door.
The development aims at a high standard of urban public
space, reinforced by controlled rather than abundant landscaping, careful detailing of paths, fences and walls, and
achieves good standards of privacy between units. The
layout also achieves a high level of car proximity and security; consequently, the urban environment is vehicle–
oriented rather than pedestrian–oriented and in this respect
simulates suburban models.
These are standards of amenity that are possible at this
density level, and progressively more difficult to maintain
at higher points in the density range for this layout type.
Architect: not known
open
space
(public)
open
space
(private)
privacy
+
no units
parking
ratio
97
2.44*
parking
identity
+
-
refuse
security collection
+
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
23,017
130*
237
42
0.61
Case Studies
49
(8) MELVIEW, AMBRICO PLACE, NEW LYNN,
WAITAKERE CITY
This variation is the only semi–courtyard house type
included in the survey, and one of the few recent developments of the type in the Auckland region. The layout
design involves agreement from the controlling Territorial
Authority to a high FAR, in this case calculated as 1.0, and
use of a house type generally considered to be expensive
for medium density housing, in this case a wide–frontage
two or three bedroom unit with a double garage connected
to it. In this description there is a clear implication of an
experimental type. Diagonal cross–over garden walls
divide rear gardens, and blank rear walls to the garages
form the ends of the small courtyard gardens which are
also accessible through the house. Essentially, a single
storey design has been used, with two ‘attic’ bedrooms to
reduce roofline heights for minimum back to back
dimensions.
The design identifies by materials and scale with the brick
and tile suburban model bungalow, with an attached
garage.
On the public side the ‘mews’ access ways are shared by
six dwellings, one more than would currently be permitted
under regulations in another part of the Auckland region.
Planting in these accessways succeeds in softening the
otherwise entirely hard surface. The minimised vehicular
space requires disciplined use by residents.
The design achieves a high level of security and privacy—
there is minimal overlooking between units, or into units
from the public side—and consequently little contribution
to the sense of community in the neighbourhood. At a
density of 44dph, however, the layout achieves a higher
standard of privacy than most comparable schemes.
Comparisons can be made with Rowena Crescent (case
study 15).
The high standard of private open space achieved by the
courtyard house type is severely affected by a later development on an adjoining site, illustrated in the bottom
photograph, underlining the need for co–ordination of the
whole site strategy from an early stage if higher density
housing is to be successful.
Architect: not known
open
space
(public)
open
space
(private)
-
privacy
parking
identity
+
no units
parking
ratio
22
2.18
refuse
security collection
+
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
2,768
126
227
44
1.00
50
Best practice in medium density housing design
(9) ALBION VALE, SUNNYVALE, WAITAKERE
CITY
The majority of units in this scheme are two storey three
bedroom houses in terraced or detached type, with front
access and attached garages. In this layout (see also
Seymour Road, case study 3) the influence of the developer’s interest in building for investment is apparent, represented by the variety of detached, semi–detached, and
short terraces built. The scheme is included in the survey to
illustrate the impact this variety can have on the resulting
environment. Overall site density is increased by the inclusion of a three storey, narrow fronted, dual aspect house
type. This interrupts and varies the street terrace and forms
a larger block at the entrance to the site. Numerous materials are used, including metal sheet, board and batten,
plaster finishes, and facing brick. Details of shutters,
screens and entrances introduce variety to the street elevations. This diversity is reinforced by the site planning,
which, unusually in this typology, uses curved roads to
avoid repetitive and tedious views.
Seven two storey detached houses with remote parking in
carports opposite, and space between the units of less than
1.5m, add a further option to the investment market.
The site plan includes two small pocket parks towards the
north end of the site, a better solution to space necessary
for light and privacy distances between buildings than
privately owned back gardens, which are minimised. The
development is not close to shopping or transport other
than bus routes, but is adjacent to the West Auckland
Marae, and the major public recreation space in the district.
The high parking ratio may be partly explained by the
location.
Architect: Powley Architects
open
space
(public)
open
space
(private)
privacy
parking
no units
parking
ratio
94
2.23
identity
refuse
security collection
+
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
20,800
115
221
45
0.52
Case Studies
51
(10) ARAWA STREET, NEW LYNN, WAITAKERE
CITY
The development is one of the early new generation
medium density housing schemes in West Auckland, made
possible by the assembly of under–used land at the rear of
several properties fronting onto Arawa Street. Medium
density classification is justified by proximity to New Lynn
and the Fruitvale rail station, to which it is linked by a footpath at the bottom of the rail embankment on the southern
edge of the site. Proposals to double up the western rail link
and increase rail traffic will affect the quiet environment of
this development in the future.
3m wide driveways in and out of the site operate on a strict
one–way basis. Most units have good orientation, good
access and parking, and a reasonable outlook. The layout is
compromised by the tapered shape of the site at the west
end, reducing the space the road needs clear of building,
and also by the decision to provide a second exit at the east
end. This compresses the site area available for four north
facing units, resulting in private space on the south side
that is unacceptably small at 1.5m wide, and overshadowed. Carports behind this group reduce natural light to
kitchen windows (which could have been placed on the
gable walls) and front gardens tend to be dominated by
washing lines. Bagged refuse is collected from the roadside at the exits onto Arawa Street. The refuse bags form an
unsightly weekly event at the site entrance, and affect the
outlook from other properties on the street. This significant
design flaw could have been avoided by provision of a
boxed compound at each exit.
A slightly smaller development of seventeen units at a
density of 41dph, and a better site services solution would
have relieved most of the problems, in an otherwise
pleasant, quiet residential environment.
All units are clad in timber products and thus avoid association with ‘leaky building’ external finishes; one recent re–
sale suggests that invested values are in line with other
property values in the area, and not stigmatised by the type
of house offered.
Architect: Insite Architecture
open
space
(public)
open
space
(private)
privacy
parking
identity
-
refuse
security collection
+
no units
parking
ratio
19
1.70
-
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
4,168
112
219
46
0.51
52
Best practice in medium density housing design
(11) OATES ROAD, GLEN EDEN, WAITAKERE
CITY
The density at 51dph and a FAR of 0.55 places the layout in
the middle of the density range, and at a high point for the
layout type. This is partly due to the regular shape and
dimensions of the site, and partly to the use of a hybridised
layout design resolving the main street frontage access to
seven units by the use of rear garaging. In other respects it
is an unremarkable scheme of two storey three bedroom
houses, the majority (18 of 25) with attached single
garages not accessed from inside the units.
The Oates Road frontage on the south boundary is established by a terrace of seven houses with front doors and
kitchen windows facing the road (south), and rear access
from open–sided garages within the curtilage approached
from the two–way internal driveway. From observation, it
is apparent that not all the residents use the front door
access onto Oates Road, despite its convenience for access
to local amenities, and to visitors parking on the street.
The site includes a combined park and children’s playground on the west section of the front terrace, apparently
well–used, and adding the important dimension of space to
an otherwise compact development. All properties, and the
small park, have metal ‘pool’ fencing 1.2m high, giving a
slightly defensive impression but also clear definition,
transparency to pedestrians, and excellent security. The
street form is a successful contribution to a more urban
identity in Glen Eden.
Refuse is collected from both Oates Road entrances (no
enclosures: informal on–street arrangement), while mail
delivery is to individual properties. Rear gardens are, at
this density, inevitably small, but adequate for their
purpose, including washing lines. The seven frontage units
are less practical in this respect, the private garden space
dominated by the back wall of the garage, and the small
private area adjacent to the front door not apparently functioning as a garden in all cases. The internal street also
suffers from the garage doors, which diminish the quality
of the streetscape.
Architect: Tse Group Architects [for HNZC]
open
space
(public)
open
space
(private)
privacy
parking
identity
+
no units
parking
ratio
25
2.00
refuse
security collection
+
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
4,941
108
198
51
0.55
Case Studies
53
(12) MT TAYLOR DRIVE, GLENDOWIE
(PROJECT), AUCKLAND CITY
Redevelopment in the area around Madeleine Avenue in
Glendowie has included several experimental housing
schemes, proposed as “innovative” solutions to urban
housing at higher densities. The project (not built) is
included in the survey to illustrate the potential for mixed
housing and architecturally complex design in this process.
The proposal consists of two and three storey terraces in
linear form on a narrow site, with a group of twelve one and
two bedroomed apartment units closing the site plan at the
north end. This group has a local density of 96dph, which,
in two storeys proved not to be capable of providing an
acceptable residential environment: all external space, and
some ground level areas are required for parking and
manoeuvring of cars. The balance of the layout has a
density of 53dph, and a high FAR at 0.78, reflecting
minimal private open space proposed. The FAR figure is
close to that of the Romola Street project designed by the
same architects (case study 6).
In the Mt Taylor layout, two wedge shaped landscaped
spaces articulate the site plan, providing small open park
areas perhaps in compensation for under–sized private
gardens. Some house plans show single width garages with
stacked parking plus one external visitor parking space,
possibly anticipating later conversion to supplementary
living space, which, if effected, would reduce the parking
ratio.
This and the Romola Street scheme represent a distinctive
and lively architecture characteristic of contemporary
Auckland design. They also propose housing at moderately high density with minimal external space, overlooking between houses in this instance controlled by care
in site planning, and variation in unit design to prevent
repetitive streetscape.
Architect: Powley Architects
no units
parking
ratio
30
2.70
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
5,658
148
187
53
0.78
54
Best practice in medium density housing design
(13) ST GEORGE’S TERRACE, AVONDALE,
AUCKLAND CITY
The site lies between the edge of the Western Rail Link
south of the Avondale Town Centre, and the back wall of
the Lansford Crescent industrial area, previously not
developed. This is typical of land in the Auckland region
now being considered for housing use. In this context and
the configuration of the site itself, there would appear to be
minimal opportunity to achieve a reasonable residential
environment.
The accommodation is standardised around a two bedroom
plus study, single bathroom, single garage formula, in an
2
average size of 116m per unit, with some variations. A
version of the three storey townhouse type has been used in
two short groups to screen the 6m high concrete block wall
on the east boundary from the bulk of the site. The twelve
units in these two blocks are penalised by this strategy: rear
patio yards are heavily shadowed, too small to have practical value, with poor natural light on the east side of the
house. Compensation is provided for the units affected by
the use of a modified single aspect plan variant, balconied
decks on the west elevation and 5.0m frontages; the smaller
2
of two bedrooms has 9.2m floor area.
The benefit to the remaining 33 two storey houses is
considerable. These are planned in short terraces following
a curved central access road. Casual parking spaces occur
intermittently along the road without dominating the
space, which is further enhanced by moderately dense and
well maintained landscaping. Garage doors are recessed
behind the front elevation line, and although details such as
meter boxes, refuse bins, and steps to entrances are not all
resolved, the public side of the terraces generally produces
a satisfactory urban housing environment.
Because of the linear site and terraced housing form, overlooking is not a significant problem except for two short
groups in the centre of the plan, where back to back dimensions are too small.
The scheme is entirely built in timber framing with a plastered cladding system. There is a body corporate responsible for maintenance, with a manager resident on site. The
project was built as an open development, and is now
gated.
open
space
(public)
Architect: Tse Group Architects
open
space
(private)
privacy
parking
-
-
+
no units
parking
ratio
45
2.30
identity
refuse
security collection
-
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
8,427
116
187
53
0.62
Case Studies
55
(14) GUNNER DRIVE, HARBOUR VIEW,
WAITAKERE CITY
The third site selected in the Harbour View development
consists of 31 houses in a rectilinear plan form, representing a conventional arrangement of medium density
housing on a straightforward flat rectangular land parcel.
The high density achieved is partly the result of small floor
areas (allowing for integral garages, the net habitable space
2
averages 95m for three bedroom houses), and the use of
narrow fronted deep–planned house types.
These have a single aspect configuration at first floor made
possible by the third bedroom being accessed from the
living room on the ground floor.
Consequently, an internal ground floor bathroom is necessary with access from the living room and headroom partly
restricted by the stair. According to approved drawings a
sliding door unit is used between the garage and the living
room. Sliding doors are not supplied as self–closing, or
air–tight fittings.
Kitchens are placed next to the front entrance in this plan,
maintaining an active street elevation.
Rear gardens are only accessible through the house in most
cases: the site plan, at this density, does not permit rear
footpaths.
Overlooking is contained by the mix of types, the exception being in the use of the ‘C’ variation (a dual aspect
narrow front type) used for the group of three in the centre
of the block—the single aspect unit would have overcome
back to back overlooking, but would also have reduced the
total number of units. The ‘C’ type is a two bedroom plus
‘study’ on the upper floor, with a poorly planned ground
floor internal kitchen and under–sized living room.
This scheme achieves minimum standards of private and
public space without providing any degree of separation
between pedestrian and vehicular space.
Architect: Snell Kaiser Hale Ltd Designers
open
space
(public)
open
space
(private)
privacy
-
-
no units
parking
ratio
31
1.90
parking
identity
refuse
security collection
-
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
5,076
114
164
61
0.70
Type C: Ground floor plan
56
Best practice in medium density housing design
(15) ROWENA CRESCENT, GLENDOWIE,
AUCKLAND CITY
The section of the development reviewed in this study is
limited to the rear access terrace of sixteen units. Low
density housing in the immediate vicinity determines a
relatively low density layout on this site (24 dph), rather
than housing at medium density; the inclusion of this
development in the study is justified by an experimental
site design.
Rear access from semi–private or private rear lanes is
endorsed by many housing designers overseas, particularly
the New Urbanist group in the USA. The rear access layout
type is discussed further in Section 6. The garage is
normally separated from the house, as in this scheme, but
remains within the property curtilage. Private garden or
patio space between the two, and separation of the
extended function of the garage from the house promises a
diversity in practical use that usually cannot be offered by
the attached garage model. The removal of the garage to
the back ‘liberates’ the street frontage by separating the
main public elevation and the front door from the main car
access, thus creating the possibility of an urban street
dominated by active and continuous facades.
In this instance, site dimensions have allowed an extended
garden area and thus a distance between garage and house
that would appear to be too great. At this density other
layout types could have been considered, including courtyard housing with equal private open space and garaging
amenity. This comment does not, however, suggest criticism of the scheme, but recognises the experiment
undertaken.
The street side of the houses, with casual parking, some
landscaping, and front doors, windows, and the stepped
terrace elevation, is a successful and welcome variation in
a typical low–density Auckland suburb, and appears to
satisfy the objectives of the design. Some detail of the units
themselves, such as patio doors to the garden side serving
the rear access determined by the layout, are less
satisfactory.
Architect: Architectus Architects [for HNZC]
open
space
(public)
open
space
(private)
privacy
+
+
no units
parking
ratio
16
2.01
parking
identity
refuse
security collection
+
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
6,570
116
410
24
0.28
Case Studies
57
(16) TUSCANY WAY, HARBOUR VIEW,
WAITAKERE CITY
The first phase of this group, four two storey three
bedroom terraced houses was built as an experiment in
layout typology, adopting the principle of rear access from
a private lane, with the formal house frontage facing a
public street. The site is adjacent to a small commercial
area. The house type used succeeds in bridging the transition to the residential character of new housing to the north
and west.
In this development the front street is the boundary of a
small public park. Orientation places the garden at the back
on the northerly side of the terrace.
Internally the house type used in the first stage is conventional in plan, without significant recognition of the
connection to the garden and garage, or use likely to be
made of the ‘back’ door. In the second stage, not yet
completed, and delayed for several years after the first
block was occupied, the plan arrangement is modified to
form a small courtyard between a large double garage and
the house, with a corridor connection between the two
parts. The internal wall of the garage is fitted with glazed
doors opening into the courtyard. The design has the
potential to offer live/work accommodation.
Ground floor plan: North Block
Benefits to the public street side (both stages) include full
use of the frontage without the interruptions of garage
doors or vehicular pavement crossings, and improved
pedestrian safety creating a wholly pedestrian environment. The rear access lane is entirely hard surfaced, and is
inevitably a low quality space, with insufficient allowance
in the planning for planting or variation to the aesthetic of
continuous metal doors. The lane is separated from the
adjoining garage court (serving another development) by a
robust fence, which reinforces perceptions of high security
but also regrettably doubles the driveway surface.
Architect: Richard Lambourne
open
space
(public)
open
space
(private)
privacy
parking
+
+
+
no units
parking
ratio
13
2.00
identity
refuse
security collection
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
3,539
157
272
37
0.58
58
Best practice in medium density housing design
(17) SACRAMENTO 1A, BOTANY DOWNS,
MANUKAU CITY
Two stages of this scheme are included in this review to
offer a comparison of densities achievable in hybrid
layouts with mixed house types. There are few such examples in recent Auckland medium density developments; in
other respects, particularly for perimeter buildings
enclosing a communal space, and for lower market positioning, this scheme also provides a comparison with case
studies 20 and 21.
The site is close to the Botany Downs shopping centre. A
communal pool with a changing pavilion, and a tennis
court, provide public space in the centre of the layout,
which is also traversed by a public footpath from the
perimeter road.
The development has a density of 34dph, relatively low for
a terraced housing layout. However, this figure includes a
2
large public open space of 2900m . If the public space is
deducted the density calculation increases to 44dph, closer
to a representative figure for two storey mixed housing.
Density is determined by use of two main house types:
(i) dual aspect/dual access two storey two bedroom unit,
of 76m2, with no integral garage. This type can be used in
either principal (north, or east–west) orientation, entered
from either side with parking either side or more distant
(40% of the total). All units have a rear accessed carport or
car parking space, plus visitor space.
(ii) a dual aspect front access three bedroom type, with an
integral single garage, and approximately 138m2 floor area
(60% of the total). The majority of units are accessed from
the site boundary and there are two variations, with off–set
or stepped plans used at corners, some of which do not
have attached garaging.
Private gardens are very small, and overlooking in the
corner sites is pronounced, but the generous central space
provides some compensation, perceptions of spaciousness
and good distance between terraces. The frontage to the
external streets, with reduced numbers of crossings and
few garage doors, succeeds in providing a good interface
with the public realm.
open
space
(public)
open
space
(private)
+
-
-
no units
parking
ratio
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
46
2.40
privacy
parking
identity
refuse
security collection
Architect: Powley Architects
13,440
113
292
34
0.36
Case Studies
59
(18) OATLANDS DEVELOPMENT, PENNANT
HILLS ROAD, SYDNEY
This project is an example of a commercial mixed density
sub–division that combines three separate layout
principles:
(a) a perimeter access road serving a lower density
2
detached unit house type, on approximately 500m lots.
This strategy provides the development as a whole with a
reducing scale of lot sizes to form a boundary to
surrounding low density housing, and also retains 27 high
value houses to sell.
(b) A core area of the site contains 50 detached and
2
terraced units on smaller lots varying from 300m to
2
160m . All units front either the perimeter road or the
central ‘village green’ public open space and are accessed
from a privately owned rear service lane. Partly because of
the narrow lot widths (minimum 4.5m) the lane is dominated by garage doors with variation provided by open
sided carports and seventeen further accommodation units
in the form of studio apartments built over double garages,
the studios accessed from external stairs entered from the
service lane. A three storey block of 24 apartments with
underground parking encloses the central public space, but
with an inactive east façade to the park. The average
density of the development is approximately 22dph, while
the core density reaches approximately 37 dph.
(c) The whole development is planned in accordance
with some of the New Urbanist principles: pedestrian
systems, axes, landscaping organised in geometric
patterns, and formal architectural detail responding to
order in rank and position in the site layout.
The layout includes two small parks, an ‘activity’ area, and
a ‘passive’ recreation space, which includes a children’s
playground.
The development succeeds in generating a strong sense of
communality in the central area, while the character of the
perimeter road is not distinguishable from any average
suburban street in the area. Oatlands is a sophisticated,
carefully planned mixed unit housing scheme that
combines suburban and urban streetscapes.
Architect: Stanton Dahl Haysom Spender
open
space
(public)
open
space
(private)
+
no units
parking
ratio
140
-
privacy
parking
+
+
identity
refuse
security collection
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
64,800
-
277
37
-
60
Best practice in medium density housing design
(19) FONTENOY ROAD, MACQUARIE PARK,
SYDNEY
Fontenoy Road is the oldest example reviewed in the study,
included, along with Ewenton Street (case study 25) to
reflect the greater familiarity of Australian housing
designers with the design of medium density housing, and
to illustrate a twenty year old design that has matured and
improved as a living environment, without alteration or
significant re–investment over the period.
The site was the first of several stages of medium density
development carried out by Lend Lease Homes in the area
around Macquarie Park, all of which have been completed
at similar densities to Fontenoy Road.
On this site, a difficult boundary configuration and a steep
slope to the north east have resulted in some units being
disadvantaged for access from the higher driveway in the
centre of the plan. Others have direct access from parking
or temporary unloading space: garages and carports are
generally a short walking distance from the house unit.
All units are two storey, the majority (25 of 35) with two
large bedrooms and a ground floor bathroom in addition to
a ‘two–way entry’ bathroom on the upper floor. Three
bedroom units are designed with a second bathroom on the
bedroom floor. As in the case of the Ewenton Street
scheme, the house areas are small, and reflect the space
standards applicable at the time rather than current Australian standards.
A high standard of privacy is achieved by pedestrian–only
access to most units, a large central space occupied by the
driveway, and by dense landscaping. The terraces are
connected by steps responding to a sloping site, and the
efforts made by the designers to articulate the otherwise
uniform elevations to the public side. All houses in the
scheme have small private gardens or patios on both sides
of the unit, seen in this analysis as the product of density of
less than 40dph, and the acceptance of separate car storage
and parking.
Architect: Lend Lease Homes (Architects) 1983
open
space
(public)
open
space
(private)
+
no units
parking
ratio
35
1.5
privacy
parking
+
+
identity
refuse
security collection
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
8,900
107
254
39
0.44
Case Studies
61
(20) CAROLINA PLACE, ALBANY, NORTH
SHORE CITY
The scheme designed for this rectangular site consists of a
two stage development, of which the first, with 33 units, is
considered here. The site plan shows both phases of the
development to clarify the planning strategy, of two “U”
shaped courtyards with parking to each side. This is a small
unit development in a two and three storey building form.
All parking is outside the property curtilage, that is, remote
from the dwelling. The central public open space is landscaped, forming a pocket park with a pool and
2
poolhouse/gym of 80m for common use.
To break down the repetitive character of the building
blocks, identity of individual units is sought through the
frequently used device of small pitched gables added decoratively to front and back terrace elevations. Units vary in
size and (presumably) market level, from two bedroom
2
duplexes, to one and two bedroom apartments of 50m .
Internal planning is conventional with all units having
views onto the central space; included in the conventions
of this type of unit are one bedroom units with a combined
entry space/kitchen/bathroom access located on the south
side of the unit. Shared external stairs provide access to
upper units.
Parking is one covered carport for each unit, plus one open
parking space, with each double sided parking block
located adjacent to the dwellings served. Internally, the
layout has consistency, but at the expense of back spaces
heavily dominated by vehicle parking, which create a
barrier between this scheme and adjoining housing. The
cause of the problem is the concentration of similar house
types and a high parking ratio for this size of unit: a mix of
types and a lower density would offer other and better
layout options.
The scheme is compared with the adjoining development
accessed from Bush Road (case study 21).
Architect: Sigma Planners, Architects & Designers
open
space
(public)
open
space
(private)
no units
parking
ratio
33
1.97
privacy
parking
identity
-
-
refuse
security collection
-
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
6,300
59
191
52
.31
Ground floor plan: Typical 1-bedroom unit
62
Best practice in medium density housing design
(21) BUSH ROAD, ALBANY, NORTH SHORE
CITY
The Bush Road development is similar to the Carolina
Place scheme in density, and for the focus on small unit
sizes; it is included to provide a comparison, alongside the
Sacramento development (Stage B). Each of these projects
has been designed to address layout issues generated by
large numbers of small units, market price expectations at
lower levels, and high parking ratios demanded by the
developer.
2
The average unit size in this project is 59m , similar to
2
Carolina Place (60m ). The layout strategy is different,
with a comparable parking standard and FAR (0.31 and
0.33). This scheme has building and active elevations on its
boundaries rather than tarmac and vehicles. This is
achieved without sacrifice of public open space, which, for
projects aiming to provide low cost housing for younger
buyers, has social significance. As with the Carolina
scheme, the central public space has a tennis court and a
pool, surrounded by gardens.
Parking is remote from the dwelling curtilage, reducing
security and the practical value of attached garaging. Inevitably the internal road is entirely dominated by parking and
carports.
The central public space is, curiously, overlooked by no
more than half of all units, because of a cross–over plan
type used in the two storey apartment planning, which
reverses lower to upper plans, placing living rooms at first
floor over bedrooms (in a separate title) below. This
arrangement raises problems of sound transmission, and
orientation on the north–south blocks; it is necessary to
avoid lower floor living rooms on the same side looking
directly into the carports lining the internal road. In addition, on the east–west facing blocks private open space is
possible at ground level, and upper floor balconies overlook rear patio gardens to bedroom windows. Upper units
are arranged in pairs sharing a staircase. The stair, at
ground level, occupies a small courtyard which also
provides access to the two lower units. A three bedroom
ground floor unit (type C) has one bedroom looking into
this courtyard. Density achieved in a two–storey development is 56 dph. A lower parking ratio and a wider mix of
unit types would improve the residential environment.
open
space
(public)
Architect: Powley Architects
open
space
(private)
privacy
parking
+
identity
refuse
security collection
-
no units
parking
ratio
105
1.86
-
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
18,750
60
178
56
0.33
Case Studies
63
(22) HOLLY STREET, AVONDALE (PROJECT),
AUCKLAND CITY
This is a proposed development of 80 houses on disused
industrial land adjacent to Avondale College sports
grounds with access from Holly Street, in an area of low
density quarter acre section housing. The scheme was
granted consent in 2001 as an ‘innovative housing’ development under the terms of the Auckland City Council
District Plan.
The site has a deep hollow in the central section affecting
the outlook of fourteen of the units proposed. The layout
proposes terraced housing spaced 15m apart with parking
on one or both sides of the private access driveway, at a
density of 62 dph and a FAR of .55. Part of the internal
roading is proposed as a public road, as far into the site as
necessary for collection, from a central amenity, of refuse.
Characteristically in this layout type the house entrance is
separated from the parking space or carport by a public
footpath, and the public side of all dwellings is thus dominated by vehicles, despite a parking ratio of only 1.6.
House types vary, the majority proposed two or three
bedroomed two storey terraced type with an average floor
2
area of 89m . At the site entrance five semi–detached units
with double garages add another house type variation.
Public spaces are indicated in three positions on the perimeter of the site plan, without nominated recreational uses or
children’s play areas.
Architect: Andrew J MacGregor Architect
no units
parking
ratio
80
1.60
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
12,870
89
161
62
0.55
64
Best practice in medium density housing design
(23) COTTONTREE, BRISBANE
This development is a type of cluster layout, but can be
classified here as a hybrid remote parking type since six of
the nineteen units are entered directly from the site boundaries rather than from within the site. It is also noted that
the layout is governed by climatic considerations
(emphasising shade, and cross ventilation for summer
cooling) appropriate to the tropical location rather than
Auckland’s temperate climate.
The site planning, however, illustrates a layout that
achieves high standards of privacy, private open space, and
identity of individual units at a high point on the medium
density scale. The project demonstrates the possibilities, in
a small development, that originate in a brief that required
diversity in accommodation, variety in unit value, and
good environmental amenity. Fourteen of the nineteen
units are entered at ground level, have small patio gardens
and secure parking adjacent to their entrances. The
2
smallest units (one bedroom, approx. 55m ) are placed on
the second floor along the southern edge of the site;
building heights taper down from three storeys on this
boundary to single storey units on the northern frontage.
The project has received awards for design.
Architect: Clare Design, Architects
Photographs by Richard Stringer, published in
Architecture Australia, 85(4); 1996.
open
space
(public)
open
space
(private)
no units
parking
ratio
19
1.18
privacy
parking
identity
+
+
refuse
security collection
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
3,572
197
63
Case Studies
65
(24) SOLJAK PLACE, MOUNT ALBERT,
AUCKLAND CITY
The development was completed in 2001, on a site next to
the western rail corridor, one of five such sites in this study.
Apart from a small recreation space in the north corner, the
layout is a dedicated, monotype development of two
bedroomed terraced houses with carport parking adjacent
to the unit, and a small front or rear private garden. All the
terraces are oriented (by the site boundaries) to face northwest or northeast.
The principal variation is the house type used on the
southern boundary for twenty units, which provides a front
patio garden in addition to space behind the terrace. The
internal plan is not, however, modified for this condition,
except for front entry to the living room. The remaining 41
units are planned with entrance through a hallway/kitchen
with side access to the stair, and a laundry space but no
ground floor toilet.
The scheme is included to demonstrate the limitations of
the layout type in which density of development in two
storeys does not permit garaging within the individual
property curtilage. To preserve privacy to ground floor
public side rooms, in this case the kitchen, the front elevation includes a glazed door (the ‘front’ or main entry door)
2
and a ventilating window of 0.25m for the laundry, but no
window to the kitchen area, which is consequently both
dark and unventilated.
The first floor is conventionally planned with two
bedrooms and a central mechanically ventilated bathroom.
Considerable effort has been made in this project to soften
the internal street, by butterfly roofs to carports, small off–
sets in the road layout, and heavy landscaping and
planting, which achieves its objectives to form a pleasant,
but uncompromisingly car dominated public space. Refuse
collection has been well designed and planned, with a
container discreetly located at the front of the development, emptied by vehicle mounted hoist. The scheme is
gated.
Architect: Powley Architects
open
space
(public)
open
space
(private)
privacy
parking
identity
refuse
security collection
no units
parking
ratio
61
2.00
+
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
9,350
75
153
65
0.49
Ground floor plan: Typical unit
66
Best practice in medium density housing design
(25) EWENTON ST, BALMAIN, SYDNEY
This two storey development is located in an inner suburb
where regeneration is occurring by a process of infilling
and small scale redevelopment in accordance with an
intensification policy. The high density figure and the high
quality of the residential environment is achieved by the
inclusion of a naturally ventilated single storey underground garage providing all parking for the 38 units on the
site (shown broken line on plan).
The layout consists of a perimeter two storey terrace of
housing enclosing an internal public courtyard defined by
225mm brick screen walls and planters. The central landscaped areas, in two courtyards are surrounded by private
patio gardens accessed from the units. Houses facing onto
the two street frontages (Ewenton and Darling Streets) are
two storey square plan two bedroomed townhouses with
rooms on the upper floors contained in roof space with
dormer windows, to satisfy the heritage context. Only
seven of the houses in the development have three
2
bedrooms, reflected in the low average size of 107m . Unit
dimensions are not available for a more detailed footprint
calculation.
Judd (1993) comments that the position of the garage
favours some units over others: there is a considerable walk
required for some householders, including a stair. Refuse
collection is from a single point in the development, a
walled, gated compound adjoining the Ewenton Street
footpath entrance. The majority of the mesh–enclosed
garage lock–ups are not used for cars but for storage, workshops, and hobby activities, reflecting the high accessibility of public transport available in Sydney.
Fifteen years after construction it is observed (2004) that
the development has matured into a comfortable, high
quality environment. Evidence of current resale prices
suggests parity with other property in the area.
Architect: Philip Cox Richardson Taylor and Partners
open
space
(public)
open
space
(private)
+
no units
parking
ratio
38
1.50
privacy
parking
identity
+
-
+
refuse
security collection
+
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
5,669
107
150
66
0.71
Case Studies
67
(26) BEAUMONT QUARTER, AUCKLAND CITY
The first stage of this development occupies formerly
commercial land near the city centre. Later stages will alter
the density (achieved in this stage by adopting underground remote parking strategies) for a high class urban
housing solution.
The scheme is included to provide a comparison with
others at a similar density. Density at 69dph is aided by the
inclusion of eight apartments in a central block also
containing a small area of lettable commercial floor space,
and by the use of five main variations in the house type.
Privacy and security, expressed in the external detailing of
louvres, are paramount concerns partly explained by the
scheme being an early intervention in this part of the city.
The louvres also provide solar control. The majority of the
house types used are without internal garaging: cars are
parked in front of units with security provided by surveillance from the house and the street, by individual electronic alarms, or, for the majority, in an underground
garage. Small courtyards, patios and rear yards separate
house fronts from public spaces which vary in character
and planting.
A range of facing materials, including pressed aluminium,
zincalume, painted brickwork and stained timber, provide
diversity and identity.
Finishes are of uniformly high quality, without being
ostentatious or expensive, both internally and externally.
The scheme includes a gym/pool reflecting the market
standard, and has high annual maintenance charges levied
through the body corporate.
The scheme demonstrates some of the potential of medium
density housing by the mixed development strategy, secure
remote parking, and the standard of urban space achieved.
Developer contributions were negotiated against the
benefit to the city of some aspects of the development,
preventing a later move to enclose the public areas.
Architect: Studio of Pacific Architecture & S333
open
space
(public)
open
space
(private)
privacy
parking
identity
refuse
security collection
no units
parking
ratio
70
1.10
+
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
10,150
n/a
145
69
n/a
68
Best practice in medium density housing design
(27) SACRAMENTO 1B, EAST TAMAKI,
MANUKAU CITY
[See general notes: case study 17]
The second phase of this scheme re–uses three of the
previous house types, but adds another, smaller unit type: a
one bedroom apartment in a two storey block: “Arizona”,
2
approximately 44m . This small unit, and the omission of
public space, increases the net density to 72 dph; the
2
average unit size is reduced from 113 to 74m , and the
parking ratio from 2.4 to 1.64; calculating the
bedspace/hectare ratio also increases the effective density
(163 maximum in Stage 1 to 191 maximum in Stage 2).
The apartment unit does, however, add an entry–level
market option for first time buyers.
More critically, in terms of the residential environment
achieved, the second stage necessitates the majority of
parking for the small units to be located off the main access
through the site, rather than contained in the garage court
as in Stage 1. Some ground level enclosed space is used for
covered parking.
The one bedroom units are entered from external stairs
located between units. The stairs lead to balconies shared
between two upper units, which are oriented north or west,
with living room over living room (compare to case studies
20, 21). This causes upper decks to overlook ground level
garden areas (bottom photo, left). Ground floor external
spaces—patios between the apartment and the rear carport
enclosure—generally lack sun and privacy.
The central spaces within the site are entirely car–dominated, with parking on both sides of the roadway, despite
the lower ratio. Façade design provides variation, alongside set–backs, and a colour scheme based on traditional
mexican shades reinforces the chosen style. Identity is
secured by these methods, without offering a convincing
demonstration of the typology at this density.
Architect: Powley Architects
open
space
(public)
open
space
(private)
privacy
parking
identity
refuse
security collection
-
-
no units
parking
ratio
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
50
1.64
6,900
74
138
72
0.53
Case Studies
69
(28) HILLSBOROUGH ROAD, LYNFIELD,
AUCKLAND CITY
This development is located next to the large and
expanding shopping facility serving the Lynfield District.
The site context consists of an arterial road frontage dominated by commercial uses. The project is included because
it illustrates innovative site planning and the quality of a
housing environment possible where larger schemes are
undertaken and are driven by a singular design philosophy,
suggesting that a critical mass factor has potential in the
typology. The scheme has been reviewed in the Architecture NZ journal, and other media, where details have been
fairly widely publicised.
The decision to separate the bulk of parking in a lower level
naturally ventilated garage has had the effect of liberating
internal site space at ground level to produce a landscape–
dominated environment. Topography and access dictated
the position of the garage on the west boundary. This
distances the parking from accommodation on the east side
of the site, necessitating the 12 centrally placed carports
and their access through the centre of the layout.
The west elevation consequently exposes the basement
garage to external views of the development, including
those from Hillsborough Road, and is reminiscent of a
form, typically commercial, of unenclosed, non–secure
parking cavities beneath a larger building mass.
Despite this compromise, and the retaining wall over–
shadowing north facing units on the eastern corner, the
development is able to exploit the low parking ratio
required in housing for the elderly to achieve a high quality
and relaxed example of housing at higher densities. All
units are spacious by comparison with private sector apartments, and planned with care and consideration for elderly
residents.
Architect: Woodhams Meikle Architects [for HNZC]
open
space
(public)
open
space
(private)
privacy
parking
identity
refuse
security collection
+
no units
parking
ratio
51
1.00
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
6,175
79
118
85
0.66
70
Best practice in medium density housing design
(29) 2 AMBRICO PLACE, NEW LYNN,
WAITAKERE CITY
This development is included in the review to provide a
comparison with projects 22 and 24. Unit sizes are similar,
2
in this case a two storey two bedroomed 77m plan, and the
site is classified by parking layout as a comparable scheme.
The density is over 30% higher than the two most similar
schemes, partly because of a regular site boundary, but
mostly the result of ruthlessly efficient use of land.
The scheme borders the Tuscany Towers development,
which it post–dates. Accepting the principle of market
variety, it offers a lower priced alternative to the earlier
scheme but at the cost of a severe reduction in quality of the
residential environment. Density is considered to be the
primary explanation for this low standard.
The two terraces of housing enclose a 16.0m wide
concreted access roadway and parking space (the dimension recommended in the Waitakere City Council design
guide). The blocks are articulated on alternate party walls
by small set-backs and steps which are intended to provide
some visual relief to an otherwise monotonous elevation.
A few of the householders have erected car ports which
contribute a small element of variation in a barren public
space, in which landscaping is entirely absent.
Internal planning of the units is conventional, with a
ground floor kitchen on the street or public side of the
block overlooking parking on both sides of the central
space.
The scheme exceeds the density limits compatible with
good residential design for this layout type. The site dominates the entrance to Ambrico Place, an unfortunate location in that it affects perceptions of quality in the rest of the
development.
Architect: not known
open
space
(public)
open
space
(private)
-
privacy
parking
-
no units
parking
ratio
22
2.00
identity
refuse
security collection
-
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
2,538
77
115
87
0.67
Case Studies
71
(30) MOKOIA ROAD, BIRKENHEAD, NORTH
SHORE CITY
The site is adjacent to commercial developments, justifying higher density housing by location, with a slope of
5m to the south west; there are views of Auckland City and
the upper Waitemata Harbour. The three storey house type
used is a deep plan version aligned in east–west blocks to
maximise solar access; all units have a double stack garage
plus a parking space. Three variations of the townhouse
plan type are used, to offer different unit sizes and accommodation packages. Allowing all parking indicated on the
site plan, the development provides a high ratio of three
spaces per unit and three additional visitor spaces. Two
refuse collection compounds are provided.
Turning and access driveways between the two higher
blocks is landscaped to form an acceptable, but car–dominated area, sufficient to use the dual aspect plan. An unusually high percentage of the site area is not privately owned,
and is maintained to a high standard.
Blocks are stepped and decorated at parapet level to articulate, minimally, the length of the façade, but a repetitive
and unvaried elevation is not significantly affected by this
move. The spectacular prospect of the upper harbour benefits only the lowest rank of the three blocks, since views
from both other blocks are obstructed.
The small park between Blocks B and C is a tapered plan,
and is equipped with a pergola and a petanque court
The centre block (Block B) uses a variation of the three
storey townhouse type that illustrates an aspect of the
evolution of the type in recent local examples: the ground
level plan provides a double length (stacked) garage
connected to an entrance hallway by a sliding door. The
original drawings indicated a rear room at this level,
accessed from the garage, which appears to have been
deleted on the first and third blocks.
Architect: Hornby Architects
open
space
(public)
open
space
(private)
+
-
no units
parking
ratio
24
2.95
privacy
parking
identity
refuse
security collection
+
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
4358
141
182
55
0.77
72
Best practice in medium density housing design
(31) GALWAY STREET, ONEHUNGA,
AUCKLAND CITY
This development is included to represent numerous examples of small housing schemes in this and other parts of the
region. Infill developments, not tied to normal residential
regulations by location on Business zoned sites, have
occurred in a fairly piecemeal pattern. These are understood by most to be typical of the medium density housing
typology.
In this instance, the diagrammatic and barrack–like site
layout seeks no advantage from the slope of the site, and
where mixed uses might have produced a better design for
the developer as well as the wider community, two rows of
more or less identical three storey blocks, the second
looking at the back of the first, have been permitted. The
forward (northern) block at least has a half level connection from the first floor living spaces to the garden, using a
step in the land slope.
Front entrances are adjacent to the garage double door at
road level, approached from the vehicular access, which is
tarmac. The internal planning is extremely confused and
impractical, affected by height to boundary regulations on
the south side. The scheme is an instructive example of the
internal difficulties in planning three storey house types.
Refuse is collected from wheeled bins parked at the site
entrances.
There is little to say in defence of development of this
quality; public doubt about higher density housing is likely
to be reinforced by such schemes.
Architect: Anthony Davis Architects
open
space
(public)
open
space
(private)
privacy
parking
-
identity
refuse
security collection
-
no units
parking
ratio
16
2.00
-
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
2,620
172
163
61
1.05
Case Studies
73
(32) KRISLEY COURT, AMBRICO PLACE, NEW
LYNN, WAITAKERE CITY
This site, part of the Ambrico Place development, uses a
version of the three storey townhouse type and is planned
at a lower density than the two storey project opposite.
Street level entrances with deep north or west facing first
floor decks provide weather protection to one third of the
houses.
The triangular site has a boundary to the Western Rail
Corridor on the northwest side. On–site car parking is
limited to 1.12 vehicles per unit, supplemented by the
public street in front of the development. This strategy
sacrifices higher parking provision to achieve higher
density and results in a congested ground level space
lacking any significant pedestrian domain.
A tilt slab construction system has been used, with internal
structures in timber framing, ensuring good standards of
sound and fire insulation. The floor plans vary between
2
blocks, averaging 134m per unit including a single garage
and a ground floor rear bedroom, similar to others of this
type in recent Auckland developments. With few exceptions (case study 30 is one) these schemes demonstrate the
limitations of the house type: all developments of this type
are characterised by poor standards of privacy, car-dominated access, poor or non-existent public space, and
impractical internal planning.
At the time of development these houses were the lowest
priced new units in the area.
Architect: not known
open
space
(public)
open
space
(private)
privacy
parking
-
identity
refuse
security collection
-
no units
parking
ratio
25
1.12
-
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
3,389
134
135
74
1.01
74
Best practice in medium density housing design
(33) KEELING ROAD, HENDERSON,
WAITAKERE CITY
The Keeling Road development is a variation on the three
storey townhouse type, in this case with a density of 94dph
made possible by a two bedroom top floor plan based on
2
floor plate areas of approximately 31m , and a low parking
ratio of 1.5. Although housing at this density falls outside
the density range, this scheme illustrates a number of
points useful to the study.
The dimension between party walls is 3.85m, producing an
internal garage width of less than 3.0m, requiring a sliding
door between the hall and the garage. All internal habitable
spaces are under–sized for practical or comfortable use:
2
2
the top floor bedrooms are 9.0m and 6.7m respectively,
with short dimensions of under 2.0m. A rear room at
ground floor level, entered through the garage, also has a
minimum dimension of less than 2.0m, while the garage
2
itself is less than 5.2m long. The total floor area of 88m
including the garage is not adequate for a three bedroom
townhouse unit. Market prices at the time of sale were the
lowest for new houses in the area.
Public space on the site is principally the roadway, open on
one side for the western block, and with a 7.6m wide space
between the other two blocks. Overlooking is unavoidable.
Private external space is accessible only through the garage
and the back room, which includes a toilet accessed from
the room. Laundry facilities are on the first floor.
Ground floor plan: Typical unit
This scheme demonstrates both the shortcomings of the
type of house used, and the constraints this type imposes
on site planning.
Architect: ADC Architects
open
space
(public)
open
space
(private)
-
-
-
no units
parking
ratio
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
22
1.50
privacy
2,330
parking
88
identity
106
refuse
security collection
+
94
0.83
Case Studies
75
(34) EDEN 1, MT EDEN, AUCKLAND CITY
Eden 1, on Enfield Street in Mt Eden, is an early example
of the advantage taken of ‘Business’ zoning in Auckland
City Council, to build high density housing without need
of compliance with standard residential design controls.
This loophole has been exploited by several development
companies in the past decade. At 125 dph, the density
places the scheme well outside the remit for this study, but
it is included to illustrate the limitations of the three storey
timber–framed townhouse option for medium density
housing.
Eden 1 also exhibits many of the problems associated with
higher density urban housing: apart from construction
defects relating to the monolithic cladding system, there
has been criticism of errors including balconies overhanging public footpaths on the perimeter, bedroom
windows on the back pavement line at street level—all
typical and symptomatic of detail design issues in the
typology of medium density housing. Internal semi–public
streets are no more than continuous walls of facing garage
doors separated by a 6m wide driveway of tarmac. Use of
this access is necessarily highly disciplined. Entry to units
from this street are unceremoniously industrial in their
presentation. The development would not have been
permitted in any Australian city or in the UK at the time it
was built.
The FAR at Eden 1, at 1.36, is the highest in the survey,
which indicates a need for a building form of at least four
storeys, and with underground parking a necessary corollary of good design for public and/or private open space
within the layout, as well as solar access, privacy distances,
and acceptable relationships to the surrounding
neighbourhood.
Architect: Richard Priest Architects.
open
space
(public)
open
space
(private)
privacy
-
-
-
no units
parking
ratio
83
1.90
parking
identity
-
refuse
security collection
-
-
average
total site unit area density 1: density 2: density 3:
2
2
2
area (m )
(m )
m /unit
dph
FAR
6,641
109
80
125
1.36
Best practice in medium density housing design
No. site and street name
Case Studies
1
2
3
4
5
6
7
8
9
no. units
total site area
(m†)
density 1:
m2/unit
density 2:
dph
density 3:
FAR
parking ratio
total floor area
(m†)
average unit
area
(m†)
site coverage
building
footprint (m†)
10
77
11
12
13
14
15
16
17
18
19
2 storeys (%)
3 storeys (%)
car parks outdoor
car parks indoor
driveway area
incl. footpath
(m†)
outdoor car
parking area
(m†)
hard surface
area (m†)
city
date of
approval
market
indicator
20
site layout
classification
1
Vinograd D r
19
5,742
302
33
0.59
2.20
3,400
179
2,439
100
-
22
19
655
286
941
WCC
1998
4
1
2
Adelphi Villa s
30
8,135
271
37
0.52
2.30
4,248
142
2,770
100
-
13
56
1,360
169
1,516
MCC
2002
4
1
3
Seymour Rd
89
24,600
273
37
0.43
2.08
10,652
118
6,181
100
-
98
77
WCC
2004
2
1
4
Corban Village
83
20,686
249
40
2.20
8,707
4,443
69
31
99
83
5,501
612
6,113
WCC
2003
3
1
5
Fairhaven
98
24,728
252
40
0.56
2.30
13,783
141
6,163
68
32
130
98
9,233
1,717
10,950
WCC
2001
2
1
6
Romola S t
13
3,140
242
41
0.73
2.10
2,292
176
1,252
69
31
2
26
392
26
418
ACC
2002
5
1
7
Tuscany Tower s
97
23,017
237
42
0.61
2.44*
12,651
130*
6,660
72
28
114
143**
n/a
-
-
WCC
1998
3
1
8
Melvie w
22
2,768
227
44
1.00
2.18
2,772
126
1,958
100
-
4
44
52
WCC
1998
3
1
9
Albion Vale
94
20,800
221
45
0.52
2.23
10,850
115
4,881
64
36
116
94
1,508
WCC
2004
3
1
10
Arawa St
19
4,168
219
46
0.51
1.70
2,127
112
1,155
100
-
16
17
WCC
1996
2
1
11
Oates Rd
25
4,941
198
51
0,55
2.00
2,693
108
1,609
100
-
25
25
WCC
2002
3
1
12
Mt Taylor D r
30
5,658
187
53
0.78
2.70
4,433
148
2,086
50
50
26
56
1,394
338
1,732
ACC
2001
5
1
13
St George s Terrac e
45
8,427
187
53
0.62
2.30
5,220
116
2,393
60
40
59
45
2,042
791
2,833
ACC
2000
3
1
14
Gunner Dr
31
5,076
164
61
0.70
1.90
3,536
114
1,790
100
-
27
31
553
351
904
WCC
1999
3
1
15
Rowena Cres
16
6,570
410
24
0.28
2.01
1,856
116
1,072
100
-
17
16
ACC
2001
2
2
16
Tuscany Wa y
13
3,539
272
37
0.58
2.00
2,042
157
1,430
100
-
3
23
WCC
1998
4
2
17
Sacramento 1 A
46
13,440
292
34
0.36
2.40
5,198
113
2,599
100
-
63
47
ACC
2001
3
3
18
Oatland s
140
64,800
277
37
-
-
-
-
83
17
n/a
-
(Aus)
2003
4
3
19
Fontenoy Rd
35
8,900
254
39
0.44
1.5
3,955
107
100
0
(Aus)
1983
20
Carolina Pl
33
6,300
191
52
0.31
1.97
1,936
59
1,452
62
38
33
32
192
NSCC
1999
2
3
21
Bush Rd
105
18,750
178
56
0.33
1.86
6,312
60
3,821
100
-
144
51
1,728
NSCC
1999
2
3
22
Holly S t
80
12,870
161
62
0.55
1.60
7,080
89
4,140
100
-
115
10
ACC
2001
3
3
23
Cottontree
19
3,572
197
63
(Aus)
1995
3
3
24
Soljak Pl
61
9,350
153
65
0.49
2.00
4,574
75
2,287
100
-
ACC
2000
3
3
25
Ewenton S t
38
5,669
150
66
0.71
1.5
4,039
107
2,211
100
0
56
(Aus)
1990
26
Beaumont Quarte r
70
10,150
145
69
n/a
1.10
n/a
n/a
82
18
21
56
ACC
2002
5
3
27
Sacramento 1B
50
6,900
138
72
0.53
1.64
3,675
74
2,418
100
-
30
52
390
MCC
1999
3
3
28
Hillsborough Rd
51
6,175
118
85
0.66
1.00
4,108
79
1,799
72
28
13
38
168
ACC
2001
4
3
29
2 Ambrico Pl
22
2,538
115
87
0.67
2.00
1,690
77
971
100
-
44
0
WCC
1997
1
3
30
Mokoia Rd
24
4,358
182
55
0.77
2.95
3,376
141
1,287
-
100
27
48
351
NSCC
2002
3
4
31
Galway S t
16
2,620
163
61
1.05
2.00
2,752
172
1,166
-
100
-
32
641
-
ACC
1997
1
4
32
Krisley Cour t
25
3,389
135
74
1.01
1.12
3,352
134
1,139
4
96
3
25
1,619
39
1,619
WCC
1997
1
4
33
Keeling Road
22
2,330
106
94
0.83
1.50
1,934
88
839
18
82
12
22
590
162
778
WCC
2000
1
4
34
Eden 1
83
6,641
80
125
1.36
1.90
9,024
109
3,770
12
88
2
159
2,085
26
2,170
ACC
1997
3
4
1,274
1,090
234
1,324
325
221
720
39
759
756
-
-
-
57
3,477
1,495
4,972
1.18
124
-
1,667
631
1,612
572
3,279
1,203
3
3
6
Discussion and
Conclusions
80
DISCUSSION AND CONCLUSIONS
Introduction
The case study review indicates that
medium density housing in New Zealand is
highly varied, wide–ranging in quality, and
evolving within a relatively de–regulated
environment. This section discusses the
issues that emerge from the case studies,
and develops a profile for a New Zealand
model for medium density housing design
at different levels of density.
Three clear points from the context and
literature review (Sections 2 and 3) provide
the platform for this section:
(i)
Medium density housing has
developed in the last decade as a
common housing typology, but is not
foreign to the urban culture of New
Zealand;
(ii) Planning strategies to consolidate
urban growth pre–suppose a higher
density housing form that, at this
stage, lacks any clear definition or
preferred model; and
(iii) Research and other literature on
medium density housing in New
Zealand is limited in scope, quality,
and quantity.
In all medium density housing there is an
element of compromise relating to house
type, car access, external private space, the
public domain, and construction costs. For
this reason the case study comments and
some of the discussion in this section
necessarily focus on areas in which the
most significant compromises are identified, thereby constructing a critical
template for the analysis. Most of the
examples included in the study have some
merits in at least one area.
Best practice in medium density housing design
crowding in some of the case studies generates the perception that privacy is reduced
or lost altogether, in turn suggesting that
personal security and individual identity
are also reduced. This leads to a progressively higher level of discomfort in the
environment as a whole, which is associated with ‘density’.
Different developments exhibit these characteristics of medium density housing to
different degrees, sometimes at very
similar (quantified) densities. Variations in
the type of layout used are considered to
partly account for such differences.
Secondary indicators are landscaping,
building detail, diversity (apparent or
actual) in built form and mix, arrangements
for refuse collection, exposure of private
open space to overlooking, and evidence of
care taken in the maintenance of public
space.
By classifying all case studies according to
the four principal layout types, comparisons can be made between developments
with the same layout type at similar densities, and different layout types at similar
densities (where overlaps occur in density
levels). The layout types are therefore
discussed in order, and as summarised in
the data chart.
Layout Type 1: Case studies 1–14
Density and Layout Type
At the lower end of the density scale two
schemes of similar size and layout type,
Seymour Rd (caser study 3) and Fairhaven
(case study 5) offer a comparison based on
density: Seymour Rd uses rear access
parking for part of the layout, while Fairhaven achieves a slightly higher density by
use of the three storey house type. Both
devices are trade–offs against the stand
alone house type preferred by most of the
housing market.
Density has been taken in this study as the
principal quantifiable ‘indicator of difference’ between housing developments that
are similar in other respects. The sense of
A larger and more positive use of the same
devices occurs in the Fontenoy Rd scheme
(case study 19; a Type 2 layout), at about
the same density, where most units do not
Discussion and Conclusions
have attached garages. The amenity loss
(represented by refuse collected from the
site entrance, remote parking affecting
55% of the residents and a relatively low
parking ratio) is balanced by a quiet, high
quality internal site environment with very
high standards of privacy, and good landscaping in the public areas.
At the same density, case studies 3 and 5
achieve high levels of direct car access, and
public service refuse collection from about
80% of the units, but at the expense of poor
public space, and low standards of individual unit privacy. The strong sense of
crowding in both these developments is
entirely absent at Fontenoy Road.
Detached and attached (i.e. paired) house
types occur at the lower ends of the density
scale; they generally fail to provide workable layouts at any point in the range
between 30dph and 66dph; a not unexpected conclusion. Romola Street (case
study 6) demonstrates the limitations, as a
development that only succeeds at all by
voluntary restrictions on car size, power
steering, and permanently curtained
windows. However, this project, and case
studies 1 and 2, suggest that density is a
market choice and not an index of social
standing.
Short two storey terraces are often
favoured by developers. In a block of three
(for example, Fairhaven (case study 5),
only the centre unit has two inter–tenancy
walls, which are expensive to build, and
end units are perceived in the private sector
market to be worth more than middle
terrace units.
Three storey house types are used in
schemes with densities as low as 40dph for
the amenity value of front access. The mix
of two and three storeys helps to produce
diversity in built form, and identity in the
street, (Albion Vale, case study 9), a quality
less easily achieved in two storey terraced
layouts. This strategy, however, results in
loss of privacy to adjoining two storey
units, and loss of the active street frontage,
81
and therefore casual surveillance, with
living rooms and kitchens at first floor
level. To reduce overlooking some of the
hybrid mixed type schemes locate the three
storey type on the boundary (Corban
Village (case study 4), Fairhaven (case
study 5), Albion Vale (case study 9)) effectively enclosing the development, although
the higher buildings on the perimeter may
sometimes adversely affect neighbourhood
relationships.
At the high end of the scale, examples of
two storey terraced housing include Oates
Road (case study 11; 51dph.) and Gunner
Drive (case study 14; 61dph.).
The Gunner Drive project was revisited
twice to observe different conditions in
use. A large public park adjoins the development of 31 terraced houses, justifying
the lack of public space within the site. The
very high density, approximately 20%
more than the next highest figure in the
layout type, is reflected in the high FAR
figure (0.7) and partly explained by the
parking ratio of 1.9, one of only two examples in this layout type with a ratio of less
than 2. On street parking adjacent to the
site is used by residents at night and weekends, but is not included in the data
recorded. Privacy between units is
minimal, with overlooking from first floor
windows affecting all houses and particularly the three in the centre of the layout;
there is much evidence of domestic activity
over–flowing to the public side of houses
due to small or shadowed rear gardens.
Comments on the internal planning of
house types, which also explain the trade–
offs involved in achieving the density, have
been made in the case study notes (p55).
In the Oates Road scheme (case study 11),
a small park is important spatially, along
with street frontages varied in height by
single storey garages alternating with two
storey houses, in that it relieves the sense of
crowding, and reduces bulk and
perceptions of density. The row of six units
on the north side are over–shadowed by a
5m high back wall to adjacent retail
82
Best practice in medium density housing design
buildings, and face towards garage doors
on the internal street. This contributes to
the car–dominated environment that
establishes the street, not as a community
space (able to serve unspecified but
implied communal activities) but as the
service
conduit
between
them:
relationships of houses are based on tarmac
rather than a shared public space, offset
only by the park area.
The most unusual, and perhaps experimental, scheme in the lower range of
density is the terraced front access development at Melview Place (case study 8;
44dph) based on a courtyard house type,
one of only two examples in the study
(Tuscany Way, case study 16, is the other).
This design prioritises privacy and security,
as well as preferred front access and close
connections between the car and house; the
only apparent trade–off is in the under–
sized access ways on the public side.
Layout Type 2 (Rear Access): Case
studies 15–19
Several examples of hybrid layouts use the
rear access system to provide parking
within the curtilage, including Seymour
Road (case study 3), Sacramento 1A (case
study 17), Oatlands (case study18), Oates
Road (case study 11), and Corban Village
(case study 4), in each case to locally
increase density and resolve site planning
problems caused by the preferred front
access type.
Only two examples are planned to exploit
the full advantages of the rear access type,
Rowena Crescent (case study 15), and
Tuscany Way (case study 16). Rear access
alters the relationship of the car and garage
to the house, placing the working entrance
on the ‘back’, and removing the vehicular
access from the front, thus relieving the
street of traffic crossings for each house.
The layout type is strongly endorsed by
New Urbanist planners, and many others,
and attracts equal criticism from some
medium density housing advocates.
Tuscany Way (case study 16) in the
Edgelea block context.
The principal contemporary merit in the
New Zealand context is that the garage can
function independently of the house and
the public street for domestic or other
purposes, including home business. An
example is provided by the Oatlands development, where the ‘mews’ rear access
private roadway is lined with garage doors,
relieved, as an urban space, by studio units
built above double garages (see text, p59).
The additional studio unit that is sold with
the house provides passive surveillance of
the mews, and offers a live–work option, or
separately lettable accommodation. (The
seventeen studios in this scheme are not
recorded as separate household units in the
density figure of 37dph).
The 32 units in two groups at the Edgelea
development are accessed from three separate lanes linking garage courts to the
public street network. The rear access
system is developed to the most sophisticated standard seen in the case studies
reviewed. Reasons for the relative absence
of this type in medium density developments in New Zealand include:
i)
ii)
iii)
density over about 40dph is difficult
to achieve because of the site space
required for the rear lane;
the cost of construction and
maintenance of the rear lane;
the house type, which is expensive to
build because of the necessarily high
external wall to floor ratio and
Discussion and Conclusions
iv)
v)
additional internal space required for
dual entry planning;
the non–traditional ground floor in
which the back door serves as the
principal entrance from the garage;
and
the unresolved dilemma of locating
the kitchen and laundry. Examples
include all possible variations of
kitchen location, the preference
generally being for a location on the
garage side of the house for direct
access to refuse disposal and use of
the private rear garden for washing.
Where the formal front elevation faces onto
a pedestrian–oriented public space as at
Oatlands, the front door, which tends to be
redundant in layouts without access to a
public space, (for instance, Corban Village
(case study 4) and Oates Road (case study
11)) can come into more frequent use.
Short walking distances, and pedestrian
routes to shops, schools and other services,
also help to justify the arrangement. The
advantages are in the formal relationship of
the house to the wider community, the
urban qualities achievable, and the flexibility of the house type.
Three schemes reviewed, (case studies 17,
18 and 19) all at densities of less than
40dph, are hybrid layouts mixing front and
rear access with integral and remote
parking.
Layout Type 3: Case studies 20–29
Parking and car storage detached from the
curtilage is regarded by developers and
householders as a less convenient and less
secure arrangement. It is apparent from this
study that such sacrifices are justified by
the developer as a trade–off against the
higher density achieved, and is acceptable
to some purchasers in terms of value and
the quality of the housing environment
offered.
From the data chart it is apparent that the
Type 3 (dedicated remote parking) layouts
83
range in density from 52dph to 87dph (2
Ambrico Place, case study 29). Type 1
layouts are displaced at a density of about
50dph, with the exception of Gunner Drive
(case study 14; 61dph), which is an unsatisfactory housing environment in numerous
respects.
The characteristics of Type 3 layouts are
relatively low Floor Area Ratios, lower
parking ratios, and significantly smaller
average unit sizes. With Type 3 layouts it is
common to find moderately large projects
with little or no variation of house type, for
example, Soljak Place (case study 24) and
Holly Street (case study 22).
The layout type is therefore considered to
be an option that suits higher density development in the private sector, where lower
market expectations are established by
location, and where little variation is
intended, and where the urban potential of
the typology is not a priority.
Layout Type 4: Case studies 30–34
The three storey house type, and its effect
on site planning, has been discussed in the
context of its role in predominantly two
storey front access layouts, as a device to
increase density, and variation in built
form. The five examples included here are
a small representative selection from a
large number of similar developments in
the Auckland area.
From densities listed in the data chart it is
clear that this house type relates to high
density rather than medium density
housing. The internal limitations of the
type have been considered in Section 3 and
commented on in case study notes, particularly when the type has been used to
increase density. In all such examples the
position of first floor living spaces imposes
overlooking and reduced privacy on adjacent two storey units. The ground level
environment is invariably car–oriented,
unless the layout and density objectives
allow enough space for separate pedestrian
movement, as at Mokoia Road (case study
84
30); at this density (55dph) however, other
layout and house types are also options.
On sloping sites the construction of
retaining walls for garaging, as at Mokoia
Road and Galway Street (case study 31),
introduces the principle of underground, or
part underground parking. At the highest
density in the schemes reviewed, (Eden 1,
case study 34; 125dph), also on a sloping
site, access and garaging are in effect
entirely underground, but not enclosed.
Type 4 layouts are not considered useful in
site planning for affordable housing at
densities of less than 66dph. It is also
apparent from the examples of this layout
type reviewed that establishing any significant public open space—the prerequisite
for the development of a community—is
not achievable at any level of density.
Summary
The most successful developments with
Type 1 layouts are all at densities of less
than 46dph. The terraced housing form in
New Zealand is an acceptable house type in
this density range, but works most effectively in shorter terrace lengths of 6–8
units, beyond which the tradition (in New
Zealand) of greater individual identity is
difficult to retain. The long straight blocks
at Tuscany Towers (case study 7) and
Sacramento (case study 17) reflect European and British design rather than the
developing local custom. The shorter rows
at St Georges Terrace (case study 13), or
Arawa St (case study 10), and elevations
varied in detail are examples of this recommended local practice.
Small, secure garage courts are justified at
all densities, and particularly above 45dph,
in two storey layouts where density begins
to require remote parking if good residential standards are to be retained. This may
suggest rear access for some units.
Progressive undergrounding of parking is a
consideration at densities over 55dph, and
for the most satisfactory environments, a
requirement at densities over 60dph, unless
Best practice in medium density housing design
house types include duplex or vertically
arranged units, in a multi–storey development with low parking ratios.
Vehicle Planning and Parking
The distinction between ‘building–dominant’ and ‘landscape–dominant’ design
made originally by the Essex Design Guide
(Stones, 1997) needs to be revisited in the
context of New Zealand and Australian
medium density housing to include a third
category, that of a ‘car–dominated’ environment. By observations from case
studies, planning for the manoeuvring and
storage of, and access to the car, and the
domestic value of the garage as an extension of routine household activity are
considered to be central to the analysis of
the typology.
Many aspects of this issue have been dealt
with in previous sections and the case study
commentaries. Further points, of general
value, are made as follows:
i)
ii)
Underground garaging: cars in
underground parking spaces relieve
the ground level environment of the
presence of the car, and can
transform the quality of the
residential
environment.
Case
studies 25, 26, 28 and 30 illustrate
this; all have achieved good
standards of public space and privacy
at densities between 55dph and
85dph. Progressive under–grounding
as density increases is shown, by
these developments, to be necessary,
from approximately 60dph in two
storey housing.
The Dutch ‘Woonerven’ integrated
traffic and pedestrian design system
may have some application in two
storey housing where a Type 2 layout
is used. There is a moderately high
maintenance penalty to consider with
this design. The nearest examples
found in this study (to the model
developed in Holland) are Fontenoy
Discussion and Conclusions
iii)
Road (case study 19), and
Hillsborough Road (case study 28).
Four wheel drive vehicles are in
evidence on many sites investigated.
These vehicles, which cannot usually
be accommodated in standard height
garages, tend to be parked outside
houses, where they block views, and
dominate by bulky ‘presence’, and
by noise. Four wheel drive vehicles
now represent 8% of private cars in
New Zealand (198,000 in total), are
increasing in popularity, and present
a particular problem in medium
density housing design. Similar
remarks apply to ‘people–mover’
vehicles based on one–tonne vans
(Toyota Hiace, etc.), and the vans
themselves, owned and used for
commercial purposes.
Mixed Development and Internal Design
A broad preference for a development
monoculture is evidenced by a large
majority of the schemes reviewed in the
private sector: there is an apparent reluctance to experiment with mixing of household sizes or types. In developments where
a mix of types has applied, there is a
tendency to restrict the range to adjacent
socio–economic groups, with no more than
one or two steps between groups. Of the
minority in the mixed category, Sacramento (case studies 17 and 27), and
Tuscany Towers (case study 7) are typical
of schemes offering housing to a narrow
social range. In smaller schemes the monoculture of a single house type is more
pronounced (Soljak Place (case study 24),
2 Ambrico Place (case study 29), Rowena
Crescent (case study 15), etc.). Repetition
of house types, creating monotonous environments in some of these schemes, helps
to build the sense of crowdedness that characterises the typology in the public mind.
This perception occurs at all density levels,
for instance Adelphi Villas (case study 2:
33dph) and 2 Ambrico Place (case study
29; 87dph), but is diverted by the constant
85
variation of building style at Corban
Village (case study 4), the Harbour View
development (case studies 1, 14 and 16),
and others.
Some of the schemes that embrace diversity of household type, and variations in
external design, also, perhaps predictably,
generate a lively, vibrant community,
visible even from relatively brief site visits.
Comments on internal details are limited to
a small number of examples where access
was available, and cannot be regarded as
comprehensive in this study. Comments
noted in the case studies are summarised as
follows:
i)
garages and ground floor toilets
should not be accessed from living
rooms or kitchens;
ii) internal routes for laundry and refuse
need to be planned to avoid passing
through living rooms;
iii) kitchens should be ventilated and
able to receive natural light by
location on external walls;
iv) ground floor toilets in two–
bedroomed four person units are
desirable;
v)
rear garden access should include
external
pathways
wherever
possible;
vi) more use should be made of first
floor single aspect house types to
control overlooking;
vii) single aspect two storey house types
based on courtyard front access plans
should be considered;
viii) the actual higher building cost of
medium density housing needs to be
recognised,
particularly
where
density exceeds 45dph.
In addition, the regular appearance of small
extra spaces within a house plan for
‘study/office/sitting’ uses, including power
points, telephone connections and
television aerials often in quiet corners or
first floor landing areas, is noted as a
86
reflection
of
requirements.
Best practice in medium density housing design
changing
domestic
In several developments, semi–commercial activities were observed during visits.
These generally consist of garages in use as
workshops with doors open for light and
air, and in one case several people working
at sewing machines on tables and benches
set up for out–work or ‘work from home’
business operations. Sub–letting of rooms
or garages is another common form of use,
reflecting demographic change and new
patterns of work in New Zealand’s urban
centres. Activities of this type are invisible
in the suburbs, but are often conspicuous in
medium density housing.
the upper end of the present scale: research
is needed to examine the costs and benefits
of this option.
iii)
Technical aspects of sustainable
design, particularly energy consumption,
but also water services, in medium density
developments, needs further research to
establish criteria for cost–effective insulation methods, orientation, and construction
materials.
Further Research
iv)
Retained capital value is considered to be a vital indicator in sustainable
medium density housing; a long–term
study that tracks re–sale prices relative to
local property values is needed to establish
similarities and differences.
This study has been restricted to the density
band between 30 and 66dph, in which the
standard form of construction is two and
three storey housing using timber frames as
the primary structure. Medium density
housing in other countries is now moving
towards multi–storey development at
densities up to 120dph. In the context of
these two observations, the report has identified several areas that need further study:
v)
Public acceptance of the typology
is known to be linked to the widespread
‘leaky homes’ problem, originating from
the housing construction industry generally. A study to identify this issue in the
context of medium density housing would
aim to recommend design practices to
overcome the effects of association with
this problem.
i)
Research is needed to relate costs
of construction to density to determine
steps in the density scale that are critical in
the process of medium density affordable
housing design.
vi)
A more detailed study of internal
design of components and fittings is
needed to identify durable specifications in
the context of medium density housing.
Access to as–built plans and construction
details would be necessary for such a study
to be effective.
ii)
Increasing density will require
consideration of underground garaging at
References
88
Auckland City Council (1958). City of
Auckland District Scheme: Code of
Ordinances and Scheme Statement
(Recommended
for
Approval),
Auckland: Auckland City Council.
Auckland City Council (1968). City of
Auckland District Scheme: Scheme
Statement
(Recommended
for
Approval), Auckland: Auckland City
Council.
Auckland City Council (1976). Housing: A
Challenge for Council, Auckland:
Auckland City Council.
Auckland City Council (1977). City of
Auckland District Scheme: Scheme
Statement (Public Notification),
Auckland: Auckland City Council.
Auckland City Council (1999). Growing
Our City through Liveable Communities 2050, Auckland: Auckland
City Council.
Auckland City Council (1999). The Urban
Design Code for Liveable Communities 2050, Auckland: Auckland City
Council.
Auckland City Council (1999). The Liveable Communities 2050 Strategy
Draft, Auckland: Auckland City
Council.
Auckland City Council (2001). The Residential Design Guide for Developments in Residential Zones in
Strategic Growth Management
Areas, Auckland: Auckland City
Council.
Auckland City Council (2003). Auckland
City District Plan: Plan Change 58,
Auckland: Auckland City Council.
Auckland Metropolitan Planning Organisation (1951). Outline Development
Plan for Auckland, Auckland:
Auckland Metropolitan Planning
Organisation.
Best practice in medium density housing design
Auckland Regional Authority (1967).
Regional Master Plan: Housing
(Preliminary Report), Auckland:
Auckland Regional Authority.
Auckland Regional Authority (1975).
Auckland: Population, Employment
and Land Use Patterns, Auckland:
Auckland Regional Authority.
Auckland Regional Authority (1987).
Residential Consolidation Study,
Auckland:
Auckland
Regional
Authority.
Auckland Regional Authority and Mt.
Albert Borough Council (1976).
Housing Density Study, Auckland:
Auckland Regional Authority.
Auckland Regional Council (2000a).
Building a Better Future: Intensification Review—Summary of Research
Findings, Auckland: Auckland
Regional Council.
Auckland Regional Council (2000b).
Building a Better Future: Intensification Review—Community Perceptions & Attitudes, by Research
Solutions. Auckland: Auckland
Regional Council.
Auckland Regional Council (2000c).
Building a Better Future Intensification Review—Urban Design Review,
by Boffa Miskell. Auckland:
Auckland Regional Council.
Auckland Regional Council (2000d).
Building a Better Future: Intensification Review—Review of Literature on
Housing Preferences and Choices,
by Synchro Consulting & P.R.I.S.M.
Auckland:
Auckland
Regional
Council.
Auckland Regional Council (2000e).
Urban
Area
Intensification:
Regional Practice and Resource
Guide,
Auckland:
Auckland
Regional Council.
References
Australian Model Code for Residential
Development (1992a). Amcord
Urban: Guidelines for Urban
Housing Vol 1, Canberra: Commonwealth Department of Housing and
Regional Development.
Australian Model Code for Residential
Development (1992b). Amcord
Urban: Guidelines for Urban
Housing Vol 2, Canberra: Commonwealth Department of Housing and
Regional Development.
Australian Model Code for Residential
Development (1992c). Amcord
Urban: Guidelines for Urban
Housing Vol 3, Canberra: Commonwealth Department of Housing and
Regional Development.
89
Reviewed: Urban Housing, London:
Commission for Architecture and the
Built Environment.
Commonwealth Information Services
(1989). Australian Model Code for
Residential Development. Canberra:
Australian Government Publishing
Service.
Darroch, R. (1983). “Concepts of Privacy”.
In B. Judd and J. Dean (eds) Medium
Density Housing in Australia. RAIA
Education
Division:
Canberra.
pp81–83.
Dixon, J. & Dupuis, A. (2003) Urban
Intensification in Auckland New
Zealand: A Challenge for New
Urbanism, Housing Studies Vol 18
No.3; pp 361-376.
Carmona, M. (2001). Housing Design
Quality through Policy, Guidance
and Review. London: Spon Press.
Droege, P. (1999). The Design Dividend.
Sydney: Property Council of
Australia.
Centre for Housing Research Aotearoa
New Zealand (2003). Review of
Statistical Housing Data, by J.
Leung–Wai & Dr G. Nana.
Wellington: Centre for Housing
Research Aotearoa New Zealand.
English Partnerships and The Housing
Corporation (2000). The Urban
Design Compendium. London:
English Partnerships.
Christiansen, W. K. S., Ed. (1991).
Mahoney’s Urban Land Economics.
Wellington: New Zealand Institute of
Valuers.
Colquhoun, I. (1999). RIBA Book of 20th
Century British Housing. Oxford:
Butterworth–Heinemann.
Colquhoun, I. and Fauset, P. G. (1991).
Housing Design in Practice. Harlow:
Longman Scientific and Technical.
Commission for Architecture and the Built
Environment (2003). The Value of
Housing Design and Layout,
London: Commission for Architecture and the Built Environment.
Commission for Architecture and the Built
Environment
(2004).
Design
Fader, S. (2000). Density by Design: New
Directions in Residential Development. Washington D.C.: Urban Land
Institute.
Forsyth, A. (2003). Measuring Density:
Working Definitions for Residential
Density and Building Intensity.
Minneapolis, MA: Metropolitan
Design Centre, University of
Minnesota.
Garreau, J. (1991). Edge City: Life on the
New Frontier. New York: Doubleday.
Hoque, A (2001) Urban Design Intervention for Intensive Housing, Planning
Quarterly 141, June 2001 pp 19-22.
Housing New Zealand Corporation
(2002a). Ki Te Hau Kainga: New
Perspectives on Maori Housing Solutions, by R. Hoskins, R. Te Nana, P.
90
Best practice in medium density housing design
Rhodes, P. Guy, C. Sage. Wellington:
Housing New Zealand Corporation.
Housing New Zealand Corporation
(2002b). Pacific Housing Design
Guide: Guidelines for Designing
Pacific
Housing
Solutions,
Wellington: Housing New Zealand
Corporation.
Kunstler, J. H. (1998). Home from
Nowhere. New York: Touchstone
(Simon & Schuster).
Leccese, M. and McCormick, K., Eds.
(2000). The Charter of the New
Urbanism. New York: McGraw–
Hill.
Jacobs, J. (1961). The Death and Life of
Great American Cities. New York:
Modern Library.
Marcus, C. C. and Sarkissian, W. (1983).
The Medium–Density Housing Kit.
Milsons Point, N.S.W.: Social
Impacts Publications.
Jain, U. (2001). “Effects of Population
Density and Resources on the
Feeling of Crowding and Personal
Space”.
Journal
of
Social
Psychology (127(3)), pp331–338.
Marcus, C. C. and Sarkissian, W. (1986).
Housing as If People Mattered.
Berkeley & Los Angeles, CA:
University of California Press.
Johnston, R. J. (1973). Urbanisation in
New Zealand. Wellington: Reed.
Judd, B. (1993). Designed for Urban
Living: Recent Medium Density
Group Housing in Australia.
Canberra: The Royal Australian
Institute of Architects.
Judd, B. and Dean, J., Eds. (1983). Medium
Density Housing in Australia.
Canberra: RAIA Education Division.
King, S., Rudder, D., Prasad, D. and
Ballinger, J. (1996). Site Planning in
Australia.
Sydney:
Australian
Government Publishing Service.
Kohler, A. (2004). “Land of Confusion”.
Progressive
Building,
February/March 2004, pp24–25.
Krieger, A., Ed. (1991). Andres Duany and
Elizabeth Plater Zyberk: Towns and
Town Making Principles. New York:
Rizzoli.
Kunstler, J. H. (1994). The Geography of
Nowhere; the Rise & Decline of
America’s Man–Made Landscape.
New York: Touchston, Simon &
Schuster.
Mt. Eden Borough Council (1980). Third
Review of the District Scheme: Background Report on Residential Zoning
in Mt. Eden, Auckland: Mt. Eden
Borough Council.
Mumford, L. (1938). The Culture of Cities.
London: Marin Secker & Warburg
Ltd.
Muthesius, S. (1982). The English
Terraced House. New Haven: Yale
University Press.
New South Wales Urban Design Advisory
Service (1998). Better Urban Living,
Guidelines for Urban Housing in
NSW, Sydney: Department of Urban
Affairs and Planning.
New South Wales Urban Design Advisory
Service (1998). Residential Densities, Sydney: New South Wales
Department of Urban Affairs and
Planning.
New South Wales Urban Design Advisory
Service (2001). The Residential Flat
Design Pattern Book, Sydney: New
South Wales Department of Urban
Affairs and Planning.
New South Wales Urban Design Advisory
Service (2002). Residential Flat
Design Code, Sydney: New South
References
Wales Department of Urban Affairs
and Planning.
North Shore City Council (2001). Good
Solutions Guide for Intensive Residential Developments, North Shore
City: North Shore City Council.
Perkins, H. and Thorns, D. (1999). “House
and Home and Their Interaction with
Changes in New Zealand’s Urban
System, Households and Family
Structures”. Housing, Theory and
Society 16 pp124–135.
Plunz, R. and Sheridan, M. (1999). “Deadlock Plus 50; on Public Housing in
New York”. Harvard Design Magazine, Summer 1999.
Radford, A. and Sarris, T. (2003). “Trends
and Strategies in the Design of
Medium Density Urban Housing”.
Radford, A. and Sarris, T. (2003). Affordable Medium Density Housing Solutions for Adelaide, Adelaide:
Australian Housing and Urban
Research
Institute,
Southern
Research Centre.
Rapoport, A. (1975). “Towards a Redefinition of Density”. Environment and
Behaviour 7(2), pp7–32.
Rapoport, A. (1985). “Designing for Diversity”. In D. Brown, B. Judd, and J.
Dean, (eds) Design for Diversification. RAIA, Education Division:
Canberra; Portland, OR. pp30–42.
Regional Growth Forum (2003). Auckland
Regional
Affordable
Housing
Strategy,
Auckland:
Auckland
Regional Council.
Stones, A., Ed. (1997). Essex Design Guide
for Residential and Mixed Use Areas.
Chelmsford: Essex County Council;
see also editions of this Guide from
the original publication dated 1973.
91
Tetlow, J. and Goss, A. (1965). Homes,
Towns and Traffic. London: Faber
and Faber.
Vallance, S., Perkins, H. and Moore, K.
(2003). The Effects of Infill Housing
on Neighbours in Christchurch,
Christchurch: Lincoln University.
Victoria Department of Planning and
Housing (1992). Victorian Code for
Residential Development—Multi–
Dwellings, Melbourne: Victorian
Department of Planning and
Housing.
Victorian Department of Planning and
Urban Growth (1990). Medium
Density Housing 1990, by Tract
Consultants, Swinburne Centre for
Urban and Social Research and
Sarkissian Associates. Melbourne:
Victorian Department of Planning &
Urban Growth.
Yates, J. (2003). A Distributional Analysis
of the Impact of Indirect Housing
Assistance, Sydney: Australian
Housing and Urban Research
Institute.
92
Best practice in medium density housing design
General Media References
Barton, C. (2004). “Escape from the
Ghetto”. New Zealand Herald.
Auckland.
Berry, R. (2004). “Govt Calls for Councils
to Introduce Low–Cost Zones”. New
Zealand Herald. Auckland.
Beston, A. (2003). “Snapshot of Life in the
Big City”. New Zealand Herald.
Auckland.
Chapman–Smith (1993). “The Future Style
of Infill Housing”. Architecture NZ,
May–June 1993, pp77–78.
Chapple, I. (2001). “Developers Have Cafe
Set Firmly in Their Sights”. New
Zealand Herald. Auckland.
Gibson, A. (2003). “Dream of an Urban
Paradise”. New Zealand Herald.
Auckland.
Gibson, A. (2004). “Housing Market under
Scrutiny”. New Zealand Herald.
Auckland.
Gibson, A. (2004). “End of the Quarter
Acre Paradise”. New Zealand
Herald. Auckland.
Gibson, A. (2004). “Auckland’s Growing
Pains”. New Zealand Herald.
Auckland.
Hawley, J. (2003). “Humungous”. Good
Weekend. Sydney.
Jacques, R. (1999). “Co–Housing—the
New Suburbia?” Build. Wellington.
Cumming, G. (2003). “The Big Squeeze”.
New Zealand Herald. Auckland. 01–
02/11/2003.
Johnston, A. (2003). “Setting up Your
Body Corporate”. Build, April/May
2003, pp60–61.
Dey, B. (2003). “Security Issues: The Rise
of the Gated Community”. Metro
(NZ). Auckland.
Kohler, A. (2004). “Land of Confusion”.
Progressive
Building,
February/March 2004, pp24–25.
du Chateau, C. (2001). “Little Boxes”. New
Zealand Herald. Auckland.
Ledbury, M. (2003). “Reviving Our Urban
Centres”. NZ Environment, 39, pp6–
8.
Editorial (2003). “Act Now to Halt Urban
Free–for–All”. New Zealand Herald.
Auckland.
Editorial (2003). “Our View”.
Zealand Herald. Auckland.
New
English, P. (1998). “Living in Urban
Denial”. New Zealand Herald.
Auckland.
Gaynor, B. (2003). “Aggressive Banks
Drive Housing Boom”. New Zealand
Herald. Auckland.
Gibson, A. (2003). “$1m to Fix Unit Leaks
at Albany”. New Zealand Herald,
Auckland.
Gibson, A. (2003). “Tiny City Apartment
Shockers: Developer”. New Zealand
Herald. Auckland.
McLeister, D. (1999). “Consumers Favour
Growth, Oppose Density”. Professional Builder, July 1999, p42.
McShane, O. (2003). “Pollution, Congestion Will Leave City Empty”. New
Zealand Herald. Auckland.
Orsman, B. (2001). “High Density Bends
the Rules”. New Zealand Herald.
Auckland.
Reid, G. and Watkin, T. (2003). “Blank
Faces a Blot ...” New Zealand
Herald. Auckland.
Rudman, B. (2003). “Quick Look at What
$1.74m Buys”. New Zealand Herald.
Auckland.
References
Turner, D. (2001). “Better Design, Layout,
Vital for High–Density Housing”.
New Zealand Herald. Auckland.
Turner, D. (2003). “In Defence of
Density”. Property Business (28),
pp28–30.
Ward, P. (2000). “Save Our Suburbs”. The
Australian. Melbourne.
93
Appendix A
Local Authority
Intensive
Housing Policies
in Metropolitan
Auckland
96
Best practice in medium density housing design
NORTH SHORE CITY COUNCIL
District Plan & Plan Change 1
Plan Change 1 updates policies dealing
with intensive housing. The location of
medium density housing is required to be
within easy walking distance of shops
including proposed shops, (defined as
offering a “wide range of goods and
services”); public transport (defined as
four or more trips/hour at peak periods);
and public open space for recreation. Easy
walking distance includes some recognition of topography, while public open
spaces may vary in character and not all
reserves provide all recreational opportunities. It is noted that existing retail benefits
from greater density of population.
Roads serving new medium density
housing must provide the opportunity for
“significant” visitor parking, community
services, and essential public infrastructure, including refuse collection. There are,
in addition, a series of urban design objectives listed in the Plan Change, item 1.1.3.
Amongst these, and continued in paragraph
1.1.4, is the stipulation that medium
density housing sites should enable “all
residential units to face or relate closely to
public streets”, and be capable of forming
relationships with “nearby properties and
public areas.” This requirement is to
prevent spatial separation of new neighbourhoods, even if built at higher densities,
from existing ones; to “facilitate the integration of the development.”
Plan Change 1 observes that:
“Quite significant adverse effects,
both immediate and cumulative, can
arise, and accordingly intensive residential development warrants a
distinct objective and associated
policies.”
Plan Change 1 defines “intensive housing”
as terraced housing and other forms of
multi–unit
development
generally
involving more than five units on a site.
Generally it is two–storeyed, though three
storeys are also possible. Densities will not
2
exceed one unit per 150m of land area.
Intensive housing developments must be
on sites that are capable of providing the
desired environmental outcomes, bearing
in mind their shape, size and location relative to other public facilities. Because less
open space is usually a concomitant of
intensive housing, the design detail is of
greater significance, as well as the shape
and size of the site proposed for
development.
Section 2 of Plan Change 1 deals with
improving subdivision processes.
Referring to Area ‘D’ applications, (for
‘varied residential and mixed use overlay
areas’) Plan Change 1 requires a Concept
Plan that addresses all aspects of macro
planning of roads, parking, and public open
space in relation to new housing proposals.
On–street parking must be provided at the
rate of 0.5 per unit proposed, and by the
arrangement of indented bays that do not
effectively widen the street when not in
use. The following section (including
amended Table 17A.1) provides for variations to previous District Plan(s) to revise
categories for sizes of unit lots and recognising smaller areas. This results in a revision to Table 17A.3 which now stipulates
“Density” defined as minimum net site
2
area per residential unit of 1 unit per 250m
2
in Area D, and 1 unit per 250m in the
previous category of mixed use overlay
area, in place of nil. The definition of
density is determined as “the net site area
of the site being developed divided by the
number of units proposed.”
Further amendments allow for redistribution of public and private open space, to
permit communal space in lieu of, but not
wholly in place of, private open space: this
allows for and encourages the development
of medium density housing with communally owned, and useful, park areas separate from traffic spaces. A minimum area
2
of 200m is accepted, at a minimum rate per
Appendix A Local Authority Intensive Housing Policies in Metropolitan Auckland
2
97
dwelling of 25m ; design and management
of the public space is to be approved in the
process of consent.
and 1999 would meet the criteria now
established in this Plan Change 1
document.
Regarding vehicle access, Plan Change 1
replaces previous regulations with a rule
that not more than five units will be
permitted access from a single private
driveway, defining at the same time (4.2:
rule 17A.5.1.10(d)) requirements for
pedestrian access, street frontages (at least
two habitable rooms facing directly onto
the street), and reinforcing urban design
intentions inherent in the intensive housing
development process.
Privacy, in particular, is relevant to any
study of medium density housing design.
Other authorities, and commentators,
measure privacy by fixed minimum
distance, as well as by other means. North
Shore City Council define acceptable standards of privacy in general terms, referring
only to “acceptable” levels, as in
Thus, in explanation,
“...units fronting the street contribute
to the liveability of the principal
public space in a residential area.
They foster a sense of ownership of
the street, and where doors and
windows face or front the street, residents can observe and overlook the
street, thereby enhancing the
personal security of people in the
street.”
Additional criteria for assessing all intensive housing development include the
following:
Ÿ Streetscape and neighbourhood
character and amenity
Ÿ Building form
Ÿ Outlook and outdoor spaces
Ÿ Privacy
Ÿ Landform,
vegetation
and
landscaping
Ÿ Traffic, parking, access and
pedestrian amenity
Each of these headings is explained in
greater detail. Together they demonstrate
North Shore City Council’s determination
to bring high standards to the environment
of future medium density housing or intensive housing proposals.
Testing some of the individual developments studied in the case study section of
this Report, it can be stated that few of the
schemes built in the period between 1995
“...private outdoor spaces should be
located, designed and screened to
maximise
privacy
for
unit
occupants.”
All designers will acknowledge the value
of this intention, and the difficulty of
achieving a good standard in these terms
when the most common block form is a
terraced plan of connected units, of whatever height, and at whatever density.
North Shore City Council also notes the
significance of refuse and recycling collection systems, and requires a “well integrated” provision, which is readily
accessible by service vehicles, and which
will not “detract visually or generate health
risks in the area.”
MANUKAU CITY COUNCIL
Operative District Plan 2002: Chapter
13: Residential Areas
Six issues relating to Manukau City Council’s residential areas have been identified
for further discussion, including:
Issue 13.2.2:
“Intensified residential development
can enhance the efficient use of the
City’s infrastructure … and create
energy savings, but it also has the
potential to cause adverse effects on
residential amenity values.”
Discussion advises that it is difficult to
determine the cumulative effect of
98
Best practice in medium density housing design
intensification on residential amenity.
Referring to the Victorian Code for Multi–
dwellings (Nov. 1993), the discussion
concludes that reasonable levels of amenity
can be provided by appropriate design
input: design quality is the critical factor.
Density rules are also used in the Main
Residential Zone (MR) so that “residents
have certainty about the potential of development on any adjoining site.”
Intensification also alters the existing character of an area, and is widely resisted by
existing communities.
(a)
Neighbourhood design, street
layout, street frontages, vehicle access,
public open space, and parking and landscaping provision.
Issue 13.2.6
This relates to the (current) lack of diversity in residential environments, which
limits different lifestyle options for current
and future generations. Manukau City
Council’s population is becoming increasingly diverse. The present patterns of traditional subdivision, albeit at slightly higher
densities, may be market lead, but may also
be encouraged by current development
standards and policies. These policies
appear to limit choice in housing, as well as
limiting choices for a culturally and
socially diversified population, and choices
in terms of transport options other than the
private car.
In outlining the Residential Strategy, the
document quotes from studies referred to
in the AMCORD Urban 1 1992 in which:
“...no single form can achieve all
environmental,
social
justice,
economic
and
lifestyle
requirements...”
and that
“...the most acceptable approach
seems to be selectively making cities
more compact, to increase housing
variety (and) access.” (p13).
In a strategy to moderate the impact of
higher density housing on existing low
density suburbs, Manukau City Council
uses a “special policy zone” applied to very
small pockets of land around the Botany
Centre (only), requiring net site areas of
2
400m . A special design code applies to
these areas.
In higher density areas, the Council
reserves rights over the following:
(b)
Site design including front yards,
front doors, back yards, balconies, building
envelope and frontages, landscape &
vehicular access.
(c)
Servicing.
In all these developments the Council will
have regard to all elements of the intensive
housing code (App. 1)(p51).
Appendix 1 covers two sections, A and B,
dealing with Neighbourhood Design, and
Site Design respectively (as listed above,
and p80–102). The Design Code outlines a
comprehensive set of urban design principles including street design, street frontages, development interfaces (with
existing: height to boundary regulations),
and general rules for traffic management
within larger sites. Typically, Public Open
Space is described as needing to protect
significant landscape features, to protect
privacy of dwellings, and to be designed to
ensure a high degree of public surveillance
of the space proposed (p90). Public Open
Space should therefore be adjacent to
public streets (rather than tucked behind
housing), and should avoid back yards
adjacent to it. In the same vein, public
parking is endorsed on secondary streets,
and in positions where security is provided
by overlooking.
Streets must not be dominated by parking
or by garage doors, “to conserve and
enhance neighbourhood landscape visual
amenity values.” (p92). Privacy is determined by sections B3 and B4, where details
such as fences and balconies are recommended, and privacy distances between
Appendix A Local Authority Intensive Housing Policies in Metropolitan Auckland
buildings are proposed: 12m between
facing frontages, and 20m between backs
of dwellings set around a 10m radius circle
determining acceptable relationships.
Where two separate dwellings meet at a
corner and at an angle of 135 deg. or less, a
4m distance is required between windows.
Side yards are reduced to nil metres in this
code.
Section B7 deals with car parking,
including dimensions for shared driveways
for horizontal and vertical standards of
amenity.
AUCKLAND CITY COUNCIL
District Plan operative 1999
The Plan acknowledges that there are few
sites left in the city for traditional subdivision, and therefore addresses the need for
residential growth in terms of infrastructure limitation, and the concerns of the
community to preserve and enhance the
existing character of residential areas.
(pA4).
Residential zones 1–5 are either special
character or low density areas, although
some higher density development may be
permitted in Zone 5. Zones 6–8 are
medium and high density areas. Zone 6(b)
2
provides for sites down to 300m per residential unit, with 2 height limits applied to
differing contexts.
Residential 7 is called “High Intensity”, the
policy providing for
“minimal development controls …
while affording appropriate protection on the interface with lower
intensity .. zones.”
Two height limits, of 10m and 12.5m
presuppose 3 and 4 storey developments,
respectively, in order to “facilitate more
intensive development in areas near major
public transport routes, … commercial
centres, ” (etc.).
99
Residential 8 zone, introduced in 2003, is
applicable to Strategic Growth Management Areas (SGMA’s). (pA17). This zone
applies to sites of 1 ha. or more, and anticipates 2 and 3 storey buildings in the 8(a)
areas, and 3–4 storeys in the 8(b) areas
which will generally be within five minutes
walking distance of a town centre or major
transport centre. Development controls are
limited to overshadowing, overlooking,
visual domination and loss of privacy. The
Residential Design Guide applies to this
zone in order to achieve quality medium to
high density development.
Density limitations included in zone 8(a)
and (b) propose a minimum of 150m2 of
gross site area per unit (8a) and 100m2 in
zone 8(b); in all areas a minimum of 40m2
of floor space is required per unit. (pB25).
This lower limit is supported by regulations
dealing
with
Maximum
Building
Coverage, para. 7.8.1.4 (pC8); allowing
55% coverage for sites up to 200m2, and a
sliding scale reducing to 35% for sites
between 200– 499m2. Further controls are
exercised through the Maximum Height
regulations, para 7.8.2.2 (pC13), as
outlined above.
Visual privacy is ensured at a minimal level
of operational usefulness by off set dimensions for windows facing each other less
than 6m apart by 1m vertical or horizontal
re–alignment, or other devices including
cill heights and glazing options. Acoustic
privacy is addressed in similarly minimal
regulations, including external traffic
noise. (pC18–19). Parking standards are
similar to those elsewhere in the Auckland
region, allowing 1 space per unit up to
75m2, and 2 spaces + 0.2 visitor spaces per
unit for all larger dwellings. Up to 100m
distance is permitted for visitor spaces.
Innovative housing development is anticipated on large sites, and in combination
with mixed uses including housing other
than standard use types (elderly persons
housing, etc). (7.7.4., pB11).
100
Interpretations and Definitions explains
Floor Area Ratio (FAR), as the gross floor
area of building proposed, divided by the
site area defined as exclusive of adjoining
roads (that is, area to site boundary rather
than the net residential calculation used by
AMCORD), and Gross Floor Area details.
WAITAKERE CITY COUNCIL
The principal documents relating to the
Waitakere City Council’s policy on
Medium Density Housing are:
1) WCC medium density housing criteria:
an 18 page sub–section of the District Plan
setting out criteria designed to ensure that
such housing developments “provide a
positive contribution to the character and
amenity of residential areas”, and dealing
with 8 separately headed areas of design;
2) WCC developers’ design guide for residential subdivision and medium density
housing, (1998), a 59 page illustrated
recommended practice guide intended to
advise developers, residents, and designers
on matters relating both to subdivision and
the urban qualities attainable through the
process of higher density housing.
This Guide has 3 sections covering (i)
subdivision design; (ii) design elements for
medium density housing; (iii) house types.
The Guide makes recommendations in
considerable detail for narrow lot widths,
living room surveillance of the street, and
active street frontages. It advises lot widths
for single and double garages, and allows
upper level living rooms where views are
possible, and minimum lot depths (22m)
for north and west facing sites, and 18m for
south or east facing sites (entrance side).
The Guide refers, with the same diagrammatic control detail as used in the North
Shore City Council’s Guide (see above), to
height to boundary requirements for
adjoining pre–existing developments,
anticipating the difficulty of stitching
medium density housing into the existing
suburban landscape. Privacy is addressed
Best practice in medium density housing design
with the proposal that “a reasonable degree
of privacy in … dwellings” can be achieved
by back to back dimensions of 16m
between upper level windows and 10m
between ground floor windows; in other
details, the Waitakere City Council guide
adopts identical separating dimensions
between adjacent houses as the Manukau
City Council guide.
In addition, the Waitakere City Council
recommendations include the advice that
windows of kitchens and living rooms
should not overlook adjacent private open
space, which is also defined as minimum
areas for different sized houses and units.
Parking and garaging are advised with a
view to ensuring safety of vehicle movement and to enhance street quality. In this
section, the Guide considers rear service
lanes as a “last resort”, particularly
“through” lanes which are seen as a security hazard, and are required to have front
door access and visitor parking on the
street side of the dwelling if used in
medium density housing layouts. The clear
objection in the Guide to rear access
reflects a legitimate concern for street
design where access is reversed in such a
way that the street itself is a back lane space
lacking interest, casual surveillance, and
active frontage.
In the third section the Guide House Types
are outlined in detail, covering varying
orientation of types, mixed use types, and
corner lot design preferences. There is no
attempt to relate house types to layout variations, or to density.
The Waitakere City Council approach to
medium density housing is an “effects–
based” one, in accordance with the intentions and principles of the Resource
Management Act. Density is not used by
the policy–making group of the City Council’s planning section as a regulating tool
for judgements or guidance in the
processing of housing developments, other
than as a rough estimating device at an
early stage, here based on square metres
Appendix A Local Authority Intensive Housing Policies in Metropolitan Auckland
per unit proposed rather than in dwellings
per hectare.
The guide is a comprehensive and detailed
handbook for good design in this housing
typology, recognising the essentially urban
character of higher density housing, and
addressing the principal differences
between medium density housing and
traditional suburban layout.
101