1 Chapter 1 Outline of Thesis 1.1 Introduction The level of energy consumption in modern urban environments is currently subject to considerable attention. It is apparent that modifications to energy provision and usage are needed to accommodate future demands and constraints. This requires a thorough understanding of the different components and patterns of urban energy consumption. A significant portion of the energy usage of cities and towns arises from houses and other dwellings in the form of the embodied and operational energy of these buildings and associated infrastructure. In order that more informed decisions can be made about the design, construction, operation and redevelopment of residential urban areas, a comprehensive understanding of energy consumption is required. This thesis identifies a gap in existing knowledge which renders the current understanding as incomplete. It proposes the development and application of a general model which spatially depicts the embodied energy of residential areas and provides the means for a more holistic approach to urban energy analysis. The model is developed on the Adelaide metropolitan area although the principles underlying it can be adapted to other urban locations. 1.2 Background 1.2.1 Energy consumption in the urban environment Australia has become one of the most urbanised societies in the world with some 64% of its population living in the capital cities and a further 19% in the regional cities and large towns (ABS, 2005). It is mainly in the cities of Australia that the activities of manufacturing, commerce, construction and the provision of services occur. In addition, cities are where the greater proportion of the population are engaged in work and leisure pursuits as well as where a significant part of the national wealth is created. The construction, operation and maintenance of these urban areas require resources, materials and energy. Currently, the consumption of energy in the urban environments of Australia is mainly dependent on the supply of fossil fuels. This can present a number of disadvantages and the solution to this problem is subject to much research (Energy Efficiency and Greenhouse Working Group, 2003). A dependence on fossil fuels has associated risks in terms of environmental effects, supply disruption and projected economic costs. Possible energy supply scenarios in the future, such as a greater proportion of renewable energy sources and nuclear power also have certain drawbacks, risks and probable higher financial
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
Chapter 1 Outline of Thesis 1.1 Introduction
The level of energy consumption in modern urban environments is currently subject to
considerable attention. It is apparent that modifications to energy provision and usage are
needed to accommodate future demands and constraints. This requires a thorough
understanding of the different components and patterns of urban energy consumption. A
significant portion of the energy usage of cities and towns arises from houses and other
dwellings in the form of the embodied and operational energy of these buildings and
associated infrastructure. In order that more informed decisions can be made about the
design, construction, operation and redevelopment of residential urban areas, a
comprehensive understanding of energy consumption is required. This thesis identifies a
gap in existing knowledge which renders the current understanding as incomplete. It
proposes the development and application of a general model which spatially depicts the
embodied energy of residential areas and provides the means for a more holistic approach
to urban energy analysis. The model is developed on the Adelaide metropolitan area
although the principles underlying it can be adapted to other urban locations.
1.2 Background
1.2.1 Energy consumption in the urban environment
Australia has become one of the most urbanised societies in the world with some 64% of
its population living in the capital cities and a further 19% in the regional cities and large
towns (ABS, 2005). It is mainly in the cities of Australia that the activities of
manufacturing, commerce, construction and the provision of services occur. In addition,
cities are where the greater proportion of the population are engaged in work and leisure
pursuits as well as where a significant part of the national wealth is created. The
construction, operation and maintenance of these urban areas require resources, materials
and energy.
Currently, the consumption of energy in the urban environments of Australia is mainly
dependent on the supply of fossil fuels. This can present a number of disadvantages and
the solution to this problem is subject to much research (Energy Efficiency and
Greenhouse Working Group, 2003). A dependence on fossil fuels has associated risks in
terms of environmental effects, supply disruption and projected economic costs. Possible
energy supply scenarios in the future, such as a greater proportion of renewable energy
sources and nuclear power also have certain drawbacks, risks and probable higher financial
2
costs (Enkvist et al, 2007). The concept of minimising energy consumption and increasing
energy efficiency is recognised as being a significant contributor to energy management
both now and in the future (McLennan Magasinik, 2004). In order that a suitable emphasis
can be placed on minimising energy usage in the urban environment, a thorough
understanding of current consumption patterns is required.
Achieving long term strategic modifications in energy consumption requires a holistic
approach which takes into account all forms of energy expenditure. According to Troy et
al (2003), the principal components of energy expenditure in the urban environment are:
• embodied energy of the built form (including buildings and infrastructure),
• operational energy consumed by buildings, and
• transport energy used by private and public vehicles.
There is likely to be a degree of inter-relatedness between these three components and
comprehensive analysis offers the possibility of achieving improved outcomes in
minimising energy consumption. In reality, such analyses are rarely applied to urban
developments (Perkins, 2001) except in a broadly qualitative manner. The concept of
comprehensively monitoring energy consumption as a means of achieving the sustainable
operation, maintenance and renewal of cities has been suggested by Troy and Smith
(2000). It was argued that the development of energy ‘profiles’, based on existing datasets
from various government agencies and utility providers, would allow urban planners to
consider different development scenarios for city infrastructure services and the built
environment. The ‘profiles’ would relate data not previously assembled providing a more
holistic understanding of urban energy consumption patterns. A further feature of this
proposal was that urban infrastructure should be re-used and refurbished to minimise
further energy expenditure embodied in new building materials.
1.2.2 Embodied energy of the built form
The embodied energy of a material or component is the energy consumed in its production
including upstream activities such as raw material extraction, conveyance, manufacturing
and assembly. The urban environment is composed of very large quantities of materials
and components as these are used to construct the buildings and infrastructure which create
the built form. Of the three components of energy expenditure in the urban environment,
embodied energy is normally the least considered.
3
There has been some attention paid to the embodied energy of buildings as part of life
cycle and environmental impact assessment in building research. Life cycle analysis takes
into account all of the inputs to a product and in the case of life cycle energy, the largest
components are embodied energy and operational energy. Cole (1998) states that life cycle
assessment is the only legitimate method to consider all of the impact over the life cycle of
the component or building concerned. In the life cycle energy analysis of a Melbourne
commercial building, Treloar (1995) indicated that the embodied energy was of a similar
significance to the operational energy when the additional embodied energy of periodic
refurbishments was taken into account. In recognition of this, some designers have begun
to consider the embodied energy of the building materials so that a more holistic approach
can be made to the overall energy consumption of buildings (Maitland, 2005; City of
Melbourne, 2006).
Despite advances in building research, the consideration of embodied energy in a
quantitative manner has not yet become normal practice in the design and assessment of
urban developments. Furthermore, the inclusion of the embodied energy of infrastructure
such as roads and reticulated services is rarely considered. This renders what little
information there is on embodied energy incomplete on the broader scale of urban
planning and development of built areas. Hence, there are significant gaps in the
knowledge available for assessing total energy consumption of the built environment.
It is proposed that the analysis of different existing urban design configurations, which
takes into account embodied energy, could be used to better inform decisions about new
developments. Furthermore, in the case of the redevelopment of existing urban areas,
information on the embodied energy of the built environment would offer the potential for
valuing the existing built form and accounting for saved energy in the re-use or recycling
of construction materials. This would provide a more comprehensive analysis of energy
consumption not currently undertaken and provide a broader understanding of the
significance of embodied energy. Therefore, this research is aimed primarily at the
development of new applications and insights of embodied energy at the urban scale rather
than the methods for deriving embodied energy values for materials.
1.2.3 Focus on residential areas
The built environment is often classified into residential and non-residential forms for the
purposes of regulation, planning, economic and other analyses. This research gives
4
priority to the residential sector. The reason for this is that it can be surmised that the
residential sector is more significant from the life cycle energy perspective and
consequently has a greater potential for the modification of overall energy consumption. A
number of observations support this view.
Firstly, embodied energy is related to the quantities of materials in the built fabric of the
urban environment and, taking total floor areas as a preliminary indicator, dwellings
dominate compared with all other building types (Kellett and Pullen, 2007). Additionally,
from an analysis of energy consumption of various industrial sectors in Australia, Dickson
et al (2003) report that the residential sector consumes more operational energy than the
commercial and services (non-residential) sector which comprises wholesale and retail
trade, communications, finance, government, community services and recreational
services. In the 2005-06 period, the operational energy consumption of the residential
sector in Australia was 447 PJ (petajoules) which was 12% of total final energy use,
whereas the corresponding figure for the commercial and services sector was 258 PJ or 7%
of the total.
Furthermore, any changes to the design and operation of residential areas are likely to offer
the opportunity for the modification of overall energy consumption. Such changes could
arise from the construction of new compact dwelling forms, the densification of older
residential suburbs or the retrofitting of existing dwellings (Bilsborough, 2006). The
ability to analyse both embodied end operational energy consumption would inform urban
planning decisions with regard to total energy consumption.
1.2.4 Spatial dimension
Where relevant information on energy consumption in the built environment is available, it
is normally in the form of data at the urban scale of individual buildings or aggregated
totals for the residential and commercial building sectors at a state or national scale. There
is a paucity of information at the neighbourhood, suburb or urban area scale which might
assist in comparing different built forms. Geographical information systems (GIS) offer
the potential for storing large amounts of information in the form of databases. These
databases can provide a means of aggregating data into larger geographic areas. Where
there are various data types to be considered, GIS can present the information with a
spatial dimension which allows the overlaying and accumulation of these data and a
5
greater appreciation of their significance. Hence, this method of data collection is likely to
be suitable for displaying the various components of urban energy consumption.
Supporting this concept is some research work conducted in the UK using GIS to depict
the operational energy consumption of residential areas as well as the associated
greenhouse gas emissions (Jones et al, 2001; Gupta, 2005). However, in terms of a
comprehensive energy analysis, which considers the whole life cycle of residential
buildings, neither research project considered the embodied energy of the dwellings.
A pilot study was carried out in Adelaide at the end of 2001 which considered both
operational energy and embodied energy consumption of buildings in six selected
metropolitan areas (Troy et al, 2003). The addition of transport energy of residents added
a further dimension. An objective of the study was to establish the feasibility of obtaining
the relevant information and of presenting it in a spatial format. The project achieved this
objective within the limitations of the relatively small sample size. It demonstrated that
more comprehensive research was required using a larger proportion of the urban
environment.
1.3 Problem statement
There is a lack of comprehensive information on the life cycle energy consumption of
residential areas particularly with respect to embodied energy. This is because the
derivation of embodied energy of the built environment is very difficult to achieve and its
significance has not been fully explored. Furthermore, there is no convenient way of
linking embodied, operational and transport energy consumption in residential areas. As a
result of this, planning and development decisions related to residential areas are
insufficiently informed to provide outcomes that minimise total energy consumption.
The contribution to knowledge described in this research is the construction of a model that
provides information which has not previously been available on the embodied energy of
residential areas. It is proposed that such a model could be used as a tool to inform life
cycle energy assessment of the built environment in the pursuit of more energy efficient
towns and cities. This problem statement leads to the following hypothesis.
6
1.4 Hypothesis
The embodied energy of residential urban areas can be estimated and represented
spatially with sufficient accuracy to usefully contribute to life cycle energy analyses of the
urban environment. Such analyses can better inform urban planning decisions regarding
alternative residential configurations and the redevelopment of existing residential areas.
1.5 Overall aim of the research
The overall aim of the research is to fill a knowledge gap by developing a model which
spatially depicts the embodied energy of residential urban areas as a contribution to the
mapping of total energy consumption in the built environment. This would enable the
model to be used as a tool which can assist in decision making in the urban planning and
development of residential areas. The model will be based on the Adelaide metropolitan
area but the underlying principles will be applicable to other cities.
1.6 Objectives of research
In order to investigate and determine the validity of the hypothesis, the overall aim can be
sub-divided into the following objectives.
Objective 1.
Determine the instrumental value of a model of embodied energy in the built environment
presented in a spatial format as part of a broader analysis of energy consumption.
Objective 2.
Demonstrate a method for determining the embodied energy of buildings and
infrastructure, particularly houses, in the urban environment based on input-output analysis
and other datasets.
Objective 3.
Construct the model of embodied energy based on three knowledge areas of embodied
energy theory, property data of the Adelaide metropolitan area and geographical
information software, in such a way that the model can be applied as a tool for urban
planning and development purposes. The model should offer the possibility of linking
with other components of residential energy expenditure.
Objective 4.
Consider an alternative urban centre to determine which parts of the model are transferable
and to what extent adaptation would be required.
7
Objective 5.
Show how the embodied energy maps can be combined with other energy consumption
data to provide a tool which provides more comprehensive and useful information for
decision making in urban development.
1.7 Methodology
The methodology consists of three elements which are aimed at achieving the research
objectives and determining support for the hypothesis:
(a) establishing the instrumental value of the model (Objective 1),
(b) the development of the model (Objectives 2 and 3) and
(c) the application of the model (Objectives 4 and 5).
1.7.1 Establishing the instrumental value of the model
Research work relevant to the proposed model will be reviewed to determine the current
state of knowledge and determine the instrumental value of the model in that context. The
issues to be explored are:
• the case for minimising overall energy consumption with regard to enhanced
greenhouse gas emissions, energy supply disruptions and financial costs.
• the existing knowledge relating to the design of the built form and its possible
influence on energy consumption.
• the significance of embodied energy and its consideration as a ‘sunk’cost.
• the relevance of the existing embodied energy of both newer and older residential
areas for informing decisions about future residential developments.
• the advantages of depicting embodied energy as a baseline for the analysis of other
components of urban energy consumption.
The review will form the basis for constructing the model and establish its value for energy
analysis in urban residential areas.
1.7.2 Development of the model
The method for developing the model will draw upon and synthesise three areas of applied
knowledge which are:
• The emerging body of research on the methods and techniques for estimating the
embodied energy of building materials and components.
• The collection of information on the characteristics of buildings compiled in the
property register of an urban area and additional synthetic data.
8
• The techniques developed to depict data in a spatial format by geographical
information systems (GIS).
This represents a cross disciplinary study which includes architectural science, urban
planning and modern surveying. A schematic representation of the development of the
model and its application is shown in Figure 1.1
An ideal model would depend upon the availability of comprehensive information of the
design and materials used for all dwellings and infrastructure in residential areas.
Although property registers may contain a substantial proportion of this information, there
will be a requirement for novel techniques to derive missing data. The property register for
the Adelaide metropolitan area will be used to construct a model of embodied energy
consumption and test the hypothesis that residential buildings and infrastructure can be
estimated and spatially represented. Research based on the study of single exemplars has
Figure 1.1 Development and application of the model for spatially
representing embodied energy
General under-
standing of energy
consumption in the urban environment
Application of model as
a tool for analysing residential
urban development
Depiction of the model in the form of maps and databases using GIS software
Derivation of embodied
energy model
based on a GIS platform
by aggregation of data units
Embodied energy of materials
Embodied energy of dwellings
Property Register synthetic
data
Property register
actual data
Data units Model
development Model
representation Model
application General
understanding
9
been rigorously supported by Flyvberg (2004) as a valid method of testing hypotheses.
The model will make provision for the ultimate inclusion of all building types but for this
research will focus on residential areas for reasons previously described.
This part of the methodology will demonstrate the feasibility of depicting embodied energy
in a spatial format by providing examples of maps of suburban locations. It will also show
that the model can provide links with other components of residential energy expenditure.
1.7.3 Application of the model
The application of the model as a tool will be demonstrated using three case studies
relating to different dwelling configurations and the redevelopment of existing residential
areas.
The first case study will be based on an alternative urban centre where the information on
the characteristics of buildings is not as comprehensive as in the Adelaide model. This will
determine the extent to which the model can be adapted and applied elsewhere. This case
study will also provide insights to different residential configurations.
The second case study will use the model as a tool to address the question of urban
consolidation and the preferable form for new dwellings. This will contribute to the
debate about whether low density outer suburbs or higher density inner suburbs are more
environmentally sustainable.
The third case study will relate to older established suburbs and the issues of improving the
life cycle energy efficiency of existing dwellings as well as the redevelopment of older
stock. These processes require various energy inputs such as additional embodied energy
for retrofitting older dwellings and for the construction of new dwellings. At the same
time, the demolition of older buildings liberates materials for re-use and recycling. The
tool will be used to study the energy balance of these alternative strategies.
The question of sufficient accuracy as stated in the hypothesis will also be addressed
during the three case studies described.
10
1.8 Structure of thesis
The following provides a brief summary of the thesis structure.
Chapter 2. Literature review.
This chapter establishes the instrumental value of the model proposed. The significance of
energy consumption in the urban environment is described as well as the desirability of
taking a life cycle approach to energy consumption which includes embodied energy. The
relevance of the embodied energy of the existing built form in relation to future
developments is established.
Chapter 3. Embodied Energy.
An overview of the methods of deriving the embodied energy of materials and products
including input-output based hybrid analysis is given in this chapter and comparisons are
made with data from other sources. Embodied energy and carbon dioxide coefficients are
evaluated and presented.
Chapter 4. Estimation of the Embodied Energy of Houses in the Adelaide Metropolitan
Area.
This chapter describes the method for estimating the embodied energy of houses based on
records from the State Property Valuation Register. It details the development of
techniques required to derive information not included in the property register. By
combining derived and property register information, the embodied energy of residential
areas is evaluated.
Chapter 5. Spatial Representation of Results.
The representation of the results in GIS format is shown in this chapter for selected parts of
residential areas in the metropolitan area of Adelaide. This includes the spatial
representation of as-built embodied energy and maintenance embodied energy. Examples
are also provided of greenhouse gas emissions arising from embodied energy and links to
other components of residential energy consumption. The verification of the model and
assessment of potential error of embodied energy estimations are discussed here.
Chapter 6. Case Study 1 - An Alternative Urban Centre.
This chapter summarises a study of operational energy and embodied energy in the Sydney
area which compares inner and outer suburbs. It also represents an application of the
techniques described in this research to a region where the property register does not
provide the type of information available in the Adelaide case study.
11
Chapter 7. Case Study 2 – Central Business District Apartments in Adelaide.
An example is provided in this chapter of the application of the model as a planning tool
for new dwellings. This is to guide decision making in the design and planning of different
forms of dwellings with regard to energy consumption.
Chapter 8. Case Study 3 – Redevelopment of Older Established Suburbs.
This chapter provides a further example of the application of the model as a planning tool
and draws on themes of improvement and redevelopment of the existing housing stock. It
uses the model to monitor embodied and operational consumption under different
scenarios.
Chapter 9. Summary and Conclusions.
The research is reviewed in the context of the hypothesis and objectives as stated in
Chapter 1. The significance of the model as a planning tool is summarized as is its
instrumental value as part of decision making processes in urban design. Other potential
uses of the model are described relating to the monitoring of materials resources.
Conclusions and recommendations for further research are given.
12
Chapter 2 Review of Literature 2.1 Introduction
The purpose of this chapter is to elaborate on the context for this research and to explore
gaps in the current knowledge on energy consumption in the urban environment. This will
provide the basis for establishing the need for the proposed model which depicts the
embodied energy of residential areas (Objective 1 stated in the outline of this thesis).
Initially, a case is made for the minimisation of overall energy consumption in the urban
environment. The influence of built form on energy consumption is addressed leading to
an appreciation of where a more comprehensive understanding is required. This is
followed by an explanation of the significance of embodied energy. The expected
contribution provided by the proposed model is then defined. Factors relevant to the
model are described from both national and international sources where appropriate, and
more local information relating to South Australia is provided consistent with the initial
focus on the metropolitan area of Adelaide.
2.2 The case for minimizing overall energy consumption
The desirability of minimizing overall energy consumption in the urban environment is
initially canvassed in this section. An ethical context is provided and the possible negative
consequences of energy usage are considered. This provides a context to explore the need
for a more comprehensive understanding of overall energy consumption (including
embodied energy) to assist decision making in urban planning and development.
2.2.1 Energy consumption in Australia
Energy consumption in Australia on a per capita basis is high compared with other nations.
The International Energy Agency (IEA, 2006) reports this to be 5.7 tonnes of oil equivalent
(toe/capita) in 2004. This compares with 7.9 toe/capita for the United States, 4.7 toe/capita
for countries in the Organisation for Economic Co-operation and Development (OECD)
and 1.8 toe/capita for the world. The distribution of this energy consumption in Australia
according to industrial sectors is shown in Figure 2.1 (Cuevas-Cubria and Riwoe, 2006).
NOTE: This figure is included on page 13 in the print copy of the thesis held in the University of Adelaide Library.
Figure 2.1 Energy consumption projections for industrial sectors in
Australia for 2004-05 (source of data: Cuevas-Cubria and Riwoe, 2006)
The combined residential and commercial sectors account for one fifth of the total
energy consumption in the form of direct energy used for the operation of buildings.
In addition, a significant proportion of the energy consumed by the transport sector
can be attributed to the operation of the urban environment as can a proportion of the
manufacturing sector in the fabrication and supply of construction materials (ie
embodied energy of materials manufactured in that period).
Most energy consumed in Australia is from non-renewable sources of fossil fuels ie
coal, oil and gas. Figure 2.2 shows the mix of energy sources and this has tended to
have a
NOTE: This figure is included on page 13 in the print copy of the thesis held in the University of Adelaide Library.
Figure 2.2 Australian primary energy consumption by fuel for 2004-05
(source of data: Cuevas-Cubria and Riwoe, 2006)
13
14
greater reliance on coal and gas over the last 20-30 years. Although the diversification of
primary energy sources to more renewable (MRET Review Panel, 2003) and nuclear
(Commonwealth of Australia, 2006) is subject to considerable debate, it is likely that fossil
fuels will remain significant in satisfying total energy demand in Australia in at least the
short to medium term future.
The dependence of modern towns and cities on fossil fuel sources can be placed in an
ethical context which can be used to guide urban design and development. In addition, the
reliance on a constant supply of energy has some disadvantages (Droege, 2004) and these
support the case for minimizing overall energy consumption. These disadvantages include
environmental risks, a necessity for an uninterrupted energy supply and the future burden
of higher energy costs.
2.2.2 The ethical view
An ethical perspective provides the prime reason for minimizing overall energy
consumption in the urban environment. The conflict between individual interests and the
common good in the use of natural resources is exemplified in the parable of the ‘tragedy
of the commons’1 which was popularised by Hardin (1968). At a simple level, the use of
fossil fuel energy in cities is a modern example of this dilemma. Any resolution of this or
related predicaments would benefit from an ethical framework as a reference for
formulating priorities and actions.
In discussing ethics in the context of the built environment, Fox (2006) states that:
…achieving a sustainable way of living is clearly not just a technical issue
(although it is often discussed as if it were) but also (and fundamentally) an
ethical one.
He proposes the ethical theory of responsive cohesion (Fox, 2000) which is the most
valuable form of organization in any area considered by informed judges and this can
include practical applications such as the design and operation of buildings and cities.
Responsive cohesion represents an approach where elements are responsive to each other
1 The tragedy of the commons refers to the use of a common pasture by herders. It is in the interest of each herder to
increase the number of their herd as they will personally gain the proceeds from each extra animal. The loss is the
reduction in available pasture but this is shared by all herders and is outweighed by the personal gain. With all herders
engaged in the same process, the pasture eventually becomes degraded.
15
and provides a foundational value which should be used to guide ideas and judgements. It
falls between the extremes of fixed cohesion, where elements are held together in a fixed
and rigid manner, and discohesion, where there is no logic to the combination of the
elements. In responding to conflicts in the organization of elements, Fox argues that
responsive cohesion provides a clear guide to prioritization. The biophysical realm is more
important than the human social realm which, in turn, is more important than the human-
constructed realm. This approach is consistent with that of Lemon (1999) who, in the
context of design and planning decisions, criticized conventional solutions which work
towards a deterministic ‘end point’ in favour of those which are more adaptable to
changing circumstances.
Significant in this debate is the issue of intergenerational equity (Brundtland, 1987) and it
can be argued that there is an increasing acceptance of moral responsibility towards
immediate future generations but a more ambiguous response towards distant future
generations (D’Amato, 1990). This creates a dilemma as the continuation of energy
consumption based on fossil fuels may increase the wellbeing of the current and next
generation but diminish that of more distant future generations. Responsive cohesion
provides an underpinning on which to make decisions and judgements which is cognisant
of this dilemma and which can respond to changing circumstances. This approach can be
used when forming policies dealing with issues such as the diminution of limited
resources, adverse effects of resource extraction and environmental risks associated with
fossil fuel consumption which are a result of activities in the human-constructed realm.
2.2.3 Environmental risks
The greater proportion of energy consumption in the urban environment currently
originates from fossil fuels and an important consideration of this is the contribution to
climate change risks.
Climate change and the link with anthropogenic greenhouse gas emissions from the
combustion of fossil fuels has been investigated by the Intergovernmental Panel on
Climate Change (IPCC) since the late 1980s. A tentative relationship between atmospheric
gases and the absorption of solar radiation through the earth’s atmosphere was first
proposed by Tyndall in 1861 (cited in Perkins, 2001). A more specific link between the
artificial production of carbon dioxide (CO2) and an influence on climate was
controversially made much later by Callender (1938). Over the last twenty years, the rate
16
of increase of atmospheric CO2 concentration has been about 1.5 ppm per year to the
current level of about 380 ppm (IPCC, 2001). As result of these increases, an enhanced
greenhouse effect is believed to be occurring and is thought to have given rise to an
increase in the earth’s surface temperature of 0.6±0.2ºC in the twentieth century (IPCC,
2001).
The IPCC has used models of future increases in greenhouse gases to predict changes in
global climate over the 21st century. The Special Report on Emissions Scenarios (IPCC,
2000) showed a range of scenarios based on assumptions about population growth,
economic development, technological changes and consumption of energy. Modelling of
these scenarios predicts atmospheric CO2 concentrations to rise to between 540 and
970ppm by the end of the century. It is predicted that this would result in a global
warming effect of between 1.4 to 5.8ºC over the period of 1990 to 2100 and a sea level rise
of between 90 and 880mm. Weather patterns would be more variable but with more
precipitation overall in the mid and high latitudes.
The Summary for Policymakers of the Fourth Assessment Report by the Intergovernmental
Panel on Climate Change (IPCC, 2007) has broadly confirmed the findings of the earlier
(third) report but with increased certainty regarding the link between human activities and
global warming. The Summary claims that the rise in globally averaged temperatures over
the last 60 years is now believed to be very likely due to an increase in anthropogenic
greenhouse gas emissions. A warming of about 0.2ºC per decade is predicted for the next
20 years and there is now higher confidence in projected patterns of warming including
wind patterns and precipitation.
Applying similar climate models to the Australian region has enabled the CSIRO (2001) to
predict increases in temperature for the main population centres. Predictions for rainfall in
most locations in summer and autumn vary from -35% to +35% by 2070 or tend towards
an overall increase. In winter and spring, most locations tend towards decreased rainfall
(-35% to +10%). A further CSIRO report (Suppiah et al, 2006) has focused on risk
assessment studies relating to temperature increases, coastal impacts and water resources.
The concept of minimising energy consumption as a contribution to reducing greenhouse
gas emissions has been adopted by the Council for Australian Governments in a national
17
Framework for Energy Efficiency (Energy Efficiency and Greenhouse Working Group,
2003).
2.2.4 Uninterrupted energy supply
The uninterrupted supply of energy is considered to be essential for the operation of
modern towns and cities and disruptions to this situation can result in social and economic
disturbances.
Disruptions to the supply of energy can occur at a various levels in the complex
international and national supply lines. The most notable international disruptions to
energy supply took place in the decade of the 1970s and arose from political and economic
conflicts (Mather and Chapman, 1995) as a result of the Arab Israeli, Yom Kippur war in
1973 and the revolution in Iran later in the decade.
There are also risks associated with the supply of energy when the sources are within
national boundaries from accidents, unusual weather conditions or major breakdowns.
Among these are the severe ice storms in Montreal in January 1998 when 150,000 people
were still not reconnected to the electricity supply after four weeks (BBC, 1998) and the
five weeks of power outages in Auckland commencing on 20th February 1998 causing
direct costs to business of $60 million a week (CNN, 1998). At the end of August 2005,
Hurricane Katrina wreaked such havoc that a month later, 400,000 people in Texas and
Louisiana were without power and 90% of crude oil production and 70% of natural gas
production in the region were shut down (Congressional Budget Office, 2005).
Another type of supply disruption that is less serious but which may well become more
common is when supply can not meet high demand such as that caused by air conditioning
use during hot summer conditions. In Australia, peak electricity demands are substantially
above normal base loads (up to double) and have been rising steadily over recent years.
The gap between base load and peak load is a key driver of internal policy within the
electricity supply industry (ESIPC, 2006) and demand side management is seen to play a
part in the amelioration of the risks of supply disruptions (CSIRO, 2002; Pareto Assoc.
P/L, 2004). The relationship between summer peak load and the design of dwellings is
currently subject to research in South Australia where some of the highest relative peaks
are experienced (Denlay, 2007).
18
The minimisation of energy demand and diversification of energy supplies are two
methods of reducing the susceptibility of towns and cities to energy supply disruptions.
Furthermore, analysis of the interactions between energy components (for example, the
embodied energy of energy efficient houses and their operational energy) may offer
potential for energy reduction. A more comprehensive understanding of the energy inputs
to urban areas would assist in formulating strategies to minimize energy usage.
2.2.5 Higher energy costs
Although Australia has a high per capita energy consumption, it benefits from low energy
costs compared with other developed countries (Australian Government, 2005). In the
urban environment, costs to the energy customer can be either direct, such as the charges
for electricity and gas consumed to operate buildings, or indirect as reflected in the price of
purchased goods including materials for the construction and maintenance of buildings.
Increases in energy costs have disadvantages both economically and socially and there
have been changes to the structure of the energy supply industry to mitigate these
(Anderson, 2003; Willett, 2006). There has also been a focus on improving energy
efficiency in various sectors of the economy including the operation of buildings (Energy
Efficiency and Greenhouse Working Group, 2003; ABCB, 2006).
Of the many factors that may influence future energy costs, global reserves of crude oil and
possible measures to curb greenhouse gas emissions are significant. The debate about oil
reserves carries on between the ‘depletionist’ and ‘anti-depletionist’ advocates (Bureau of
Transport and Regional Economics, 2005) and attention is now focussed on predicting if
and when the loss of a cheap and abundant supply will occur.
Measures to reduce greenhouse gas emissions have been economically modeled
(Ahammad et al, 2006) in Australia through to the year 2050 by the Australian Bureau of
Agricultural and Resource Economics (ABARE). The modeling process was aimed at
achieving a target for a stabilized atmospheric carbon dioxide concentration of 575 ppm by
2100 based on the ABARE’s global trade and environment model (GTEM). A range of
scenarios were adopted including the availability of carbon storage technology and the use
of nuclear power. Depending on the level of international cooperation proposed in the
scenarios, carbon taxes of A$77 to A$157/tonne of CO2 equivalent gases were used as well
as more extreme tax scenarios. It was recognised that, in reality, these tax levels would
require frequent adjustment. At A$100/tonne and at an approximate emissions rate of
19
1kg/kWh for electricity generation, this would amount to a tax of 10c/kWh which is a
substantial increase on current tariffs. It was considered that the economic outcomes
would be similar with either a taxation or trading scheme instrument. Economic modeling
of the substantial increases in costs associated with severe cuts to greenhouse gas
emissions have also been carried out by Allen Consulting (2006).
In contrast to the methodology adopted by Ahammad et al (2006) using stringent emissions
targets, the Prime Ministerial Task Group on Emissions Trading (Shergold, 2007)
proposed a ‘pledge and review’ approach to limiting greenhouse gas emissions. This
suggested aspirational targets and favoured the emissions trading approach following the
setting of emissions caps on industrial sectors. Emissions trading was preferred as it was
seen to be more focused on the environmental objective of reducing greenhouse gas
emissions in the most efficient manner. Furthermore, this instrument allows the market to
select the best low emissions technologies, encourages the use of carbon ‘offsets’ (eg tree
planting) and is compatible with emerging schemes around the world. Some economic
modeling was undertaken at lower prices of around A$45/tonne of CO2 equivalent gases.
However, the report recognised that both households and business will pay more for their
energy.
The effects of both supply factors and demand side economic instruments on energy
consumption and greenhouse gas emissions will be related to the price elasticity (Case &
Fair, 2007) of the various forms of energy1. In the short term, energy price elasticity is
normally low (ie inelastic) as consumers have few alternatives or are unwilling to modify
consumption patterns whereas, in the longer term, greater price elasticity occurs as
alternative consumption strategies are adopted (Oxera, 2006).
The minimization of overall energy consumption offers the possibility of moderating the
impacts arising from increases in energy costs due to supply restrictions or greenhouse gas
mitigation instruments. These costs will be incurred directly in the form of operational
energy, and indirectly in the form of embodied energy of materials used for maintenance,
and for new buildings and infrastructure. Since energy use in the urban environment is
1 The price elasticity of demand (PED) for goods is the ratio of the change in demand to the change in price. Where the
demand is highly responsive to price, the PED is greater than 1 and the goods are price elastic. Conversely, where
demand is not greatly affected by price, the PED is less than 1 and the goods are said to be price inelastic.
20
significant, its analysis over a broad front would contribute to this effort. This requires a
holistic approach which considers all energy inputs to determine lower energy outcomes.
2.3 The built form and energy consumption
2.3.1 Possible relationship
The possible relationship between energy consumption and the built form in the urban
environment has, and will continue to be, subject to research (Owens, 1986; Perkins, 2001;
Steemers, 2003). Analysis of these relationships can occur at the building scale, such as
comparing different types of dwelling, through to the examination of whole
neighbourhoods or regions of towns and cities.
At the dwelling scale, a significant factor over the last two decades in Australia has been
the substantial increase in average floor areas of new houses from approximately 160m2 to
230m2 (RMIT, 2006). This trend has contributed to greater energy consumption both in
terms of the embodied energy of building materials and the operational energy required to
provide air conditioned spaces.
To a degree, this trend has run counter to the newer planning concepts for residential areas
that are denser in terms of numbers of dwellings per hectare and which have some mixed
use that provides services and employment to minimize travel. Newman and Kenworthy
(1989) expressed a prescription for denser, more compact urban designs following a study
of Australian and world cities in the 1980s. The purpose of this was to avoid highly
dispersed cities which involved high travel energy usage, pollution and social costs.
Sustainable residential development in Australia which minimises the use of resources and
focuses on environmental performance was described in detail in the AMCORD Practice
Notes (Department of Housing and Regional Planning, 1995). More recently, the House of
Representatives Standing Committee on Environment and Heritage conducted an inquiry
into Sustainable Cities in Australia (Washer, 2005) which revisited urban transport and
densification issues and called for more leadership from the Federal Government in this
area. The Committee’s first recommendation was for the establishment of an Australian
Sustainability Charter that sets national targets across a number of areas including energy,
building design and planning.
There are few studies comparing the operational energy consumption of different forms
and densities of urban dwellings and even fewer comprehensive analyses involving other
energy inputs such as embodied and transport energy. A study of energy consumption
in housing forms of various density in Sydney was commenced in 2003 (Myors et al,
2005) by Energy Australia and the New South Wales Department of Planning. The
types of dwellings considered were high-rise apartments (9 or more storeys), mid-rise
apartments (4-8 storeys), low rise apartments (up to 3 storeys), townhouses and
detached houses. The study reported on total annual greenhouse gas emissions arising
from operational energy consumption as determined from electricity and gas accounts
of residents. The results of the study are shown in Table 2.1 and the relevant points
arising are as follows:
• significantly increasing emissions with increase in number of storeys (from
5.1 to 10.4 tonnes CO2/dwelling/year) for multi-unit dwellings.
• large emissions from detached dwellings (9.0 tonnes CO2/dwelling/year).
• a substantial moderation of the emissions of detached dwellings compared
with other types of dwelling when the number of occupants per dwelling is
considered. Indeed, on this basis of comparison, only townhouse and villa
dwelling forms produce lower emissions.
• the emissions per person in high-rise, mid-rise and low-rise dwellings are all
above the benchmark of 3.39 tonnes CO2/dwelling/year which has been
adopted by BASIX (Building Sustainability Index) as the average residential
greenhouse gas emissions level for New South Wales. The BASIX target
requires a 25% reduction on this benchmark.
• Sites equipped with central air conditioning systems created an average peak
demand for electricity in hot weather that was 111% higher than those sites
without.
Table 2.1 Results of a study by Energy Australia and NSW Department of Planning
into the total annual greenhouse gas emissions of different dwelling types (source:
Myors et al, 2005).
NOTE: This table is included on page 21 of the print copy of the thesis held in the University of Adelaide Library.
21
22
Overall, the study suggested that occupants of apartments caused greater greenhouse gas
emissions than those in detached houses. Of the various dwelling forms, only townhouses
and villas appeared to offer advantages in terms of reducing emissions on a per occupant
basis.
The complexities of designing urban areas within cities to minimize energy consumption
has been discussed by Steemers (2003) who showed that high density cities such as Hong
Kong consumed far less energy on transport than lower density cities in the USA or
Australia. However, Steemers questioned whether increasing urban density would reduce
transport energy in the short term. Furthermore, focussing on housing in the UK, the
arguments for and against densification were finely balanced when operational and
transport energy consumption were considered.
Despite the adoption of compact city concepts by many in the urban planning profession,
there remain opponents to urban densification. Whilst referring to the proponents of the
compact city, Troy (1996, p167) states:
Solutions to pollution problems have been proffered and adopted with scant
regard for scientific evidence either about the extent of the problems and their
sources or any understanding of the history of cities and why they take their
form. Moreover, there is little scientific evidence that the solutions proposed to
cope with environmental stress can or will have the beneficial effects claimed.
As part of the research project at the Warren Centre for Advanced Engineering at Sydney
University called ‘Sustainable Transport for Sustainable Cities, Troy and Smith (2000)
outlined a number of objectives in the quest for a transition to sustainability for the city of
Sydney which included reduced consumption of energy from non-renewable sources and
increased energy efficiency in the operation of buildings and structures.
The need for a better understanding of the relationship between the built form and energy
consumption is echoed by Droege (2004). In the context of changing cities to match
energy supply constraints, he comments:
Another challenge is the energy-blind nature of contemporary physical
planning, typically very short planning horizons and political uncertainties that
prevail on the local level.
23
In the existing built form, an example of improving energy consumption and greenhouse
gas emissions has been undertaken by the City of Newcastle using the measuring tool
known as ClimateCam (City of Newcastle, 2002) and this has resulted in a reduction in
energy consumption of 40%. ClimateCam is an initiative under the umbrella of the Cities
for Climate Protection (CCPTM) campaign which was initiated by the International Council
for Local Environmental Initiatives in the early 1990s (Brugmann and Jessup, 1993). The
City of Newcastle is one of the founding members of the CCPTM campaign and in the
subsequent 10 years over 200 local councils in Australia have contributed to greenhouse
gas reductions of 8.8 million tonnes (Turnbull, 2007). This tool can create a greater
awareness of operational energy consumption and greenhouse gas emissions within the
community although it is not intended as a planning tool for guiding the design of new
developments.
2.3.2 Future issues
A challenge for urban development in many cities in the future is to reconcile increases in
the population of residents with commitments to what is broadly termed ‘sustainability’ in
the local and greater environment. An example of this is the city of Adelaide where the
Strategic Plan of the South Australian Government has set an increased population target
whilst committing to limits on greenhouse gas emissions consistent with the Kyoto
Protocol (Department of the Premier and Cabinet, 2004). The population target for the
State of 2 million by 2050 is ambitious compared with the relatively low growth rate in the
past compared with other Australian cities (Nicolson et al, 2003; ABS, 2003). Recent
trends indicate that much of this growth could occur in the Adelaide metropolitan area
(ABS, 2006) as the State Government aims to prevent urban sprawl. Urban consolidation,
energy efficient buildings and the retro-fitting of existing dwellings (Bilsborough, 2006)
are possible means of mitigating increased energy consumption. This is compounded by
any requirements to adapt the performance of buildings to probable climate change
(Shimoda, 2003). The comprehensive analysis of different mitigation and adaptation
strategies is required to determine optimum outcomes for overall energy consumption.
A further target nominated by the South Australia Strategic Plan with regard to
sustainability is to limit the amount of waste going to landfill by 25% within 10 years.
More than 1000 dwellings are demolished each year in the Adelaide metropolitan area
(Burrows & McQueen, 2002) and this is a substantial proportion of all building and
24
demolition waste (EPA, 2001). The re-use and recycling of demolition waste as substitutes
for new materials represents a potential saving of embodied energy consumption.
With regard to the risks associated with climate change the Government of South Australia
has embarked on a program to reduce the State’s greenhouse gas emissions by 60 per cent
of 1990 levels by 2050. It has been the first State in Australia to introduce a Climate
Change and Greenhouse Reduction Bill (SA Govt, 2006b) which sets targets to achieve
this reduction in emissions. The strategy for achieving these targets includes goals for
buildings, transport and planning in the urban environment. Since climate change is a
global phenomenon and South Australian greenhouse gas emissions are only a very small
proportion of the world’s output, this program by itself will obviously have a negligible
effect on climate change. (Australia contributes less than 2% to global greenhouse gas
emissions (Campbell, 2006) of which a fraction is generated in South Australia).
However, it sets an example for other communities, enables South Australia to participate
in international mitigation efforts and positions the State to exploit possible opportunities
arising from climate change.
In summary, the investigation of the relationship between urban energy consumption and
the built form has been mainly at the scale of whole cities or city areas, but this has been
confined to the analysis of operational energy and transport energy. There has been little
analysis of this relationship at the smaller scale of neighbourhoods or groups of buildings,
and this has been compounded by the lack of attention given to embodied energy. A
consequence of this is that possible interactions between embodied energy and other
energy components have not been considered. For future changes in the urban
environment, the inclusion of embodied energy would provide a more comprehensive
understanding of energy consumption which could be used for informing planning and
development decisions.
2.4 The significance of embodied energy
The significance of embodied energy as part of overall energy consumption can be
assessed at the large scale of the national economy and at the smaller scale of urban areas
or individual buildings. Embodied energy is also of importance when considering the re-
use of the existing built fabric which is an area of research not previously addressed in a
quantitative manner.
25
2.4.1 Energy flows in the national economy
Focusing on the residential sector of the national economy, it is possible to compare the
total annual energy expended in the construction of new dwellings with the energy used to
operate all existing dwellings. The primary energy flows into the residential construction
sector have been calculated by Foran et al (2005) in the Balancing Act report based on
economic statistics for the national economy for 1994-95 and amount to 149PJ. A major
contributor is the energy consumed in the manufacture and supply of building materials ie
embodied energy. This compares with the primary energy consumed in operating the
existing residential dwelling stock for the same annual period of 360PJ estimated by the
Australian Bureau of Agricultural and Research Economics (Bush et al, 1997). The
primary energy flows into the non-residential construction sector (Foran et al, 2005) for
1994-95 are similar to that of the residential construction sector at 147PJ. These data
provide an indication of the significance of embodied energy. They also suggest the extent
to which the energy consumed by the built environment, as represented by the Residential
and Commercial sectors of the national economy in Figure 2.1, can be increased to include
indirect (embodied) energy inputs. Earlier research indicated that direct and indirect input
to the whole Australian construction sector in terms of embodied energy was 19.5% of
national energy consumption (Tucker et al, 1993).
2.4.2 Buildings and urban areas
At the smaller scale of individual buildings or urban areas, embodied energy can be
usefully compared with operational energy using a life cycle approach. Life cycle
assessment in a broad sense requires a detailed inventory to be made of a wide range of
impacts arising from the manufacture and supply of a product including effects from all of
the upstream processes in handling raw materials and other pre-manufactured parts (Fava
et al, 1993; Todd & Curran, 1999; Guinee, 2002; ISO 14040, 2006). This form of
assessment has been recommended to study the environmental impacts of building
materials so that the various aspects of manufacture, usage, re-use, recycling and disposal
are all considered (RMIT, 2006).
Life cycle energy analysis is a form of life cycle assessment which focuses specifically on
energy consumption, often in conjunction with an analysis of the associated greenhouse
gas emissions (Treloar et al, 2000). If applied to a building, this type of analysis adopts a
cradle-to-grave approach and considers the total energy inputs over the life cycle including
the embodied energy of construction materials and process, the operational energy over the
26
lifetime of the building and, sometimes, the demolition energy. It provides a greater
understanding of the energy inputs to the urban environment and more scope for modifying
energy consumption.
In identifying the critical factors influencing the total energy consumption of low and
medium density dwellings in Melbourne, Fay (1999) took a life cycle approach and
considered both operational and embodied energy. In the study of a small house also in
Melbourne, Treloar et al (2000) found that the embodied energy was approximately half of
the operational energy over a 30 year life cycle. This proportion increased to nearly two
thirds when the additional embodied energy arising from maintenance and repair was taken
into account. The time required for the operational energy to equal the initial embodied
energy was 14 years.
At the urban scale of neighbourhoods, Perkins (2001) compared two different styles of
housing development in Adelaide by studying samples of houses in an outer suburb at
Seaford about 40km south of the city centre and in an inner suburb at Norwood,
approximately 4km east of the city centre. In considering just the built form, it was found
that in Norwood, the split between embodied energy and operational energy was 37% and
63%, respectively. The corresponding figures for Seaford were 33% and 67% with the
higher figure for operational energy arising from higher lighting, heating and cooling
loads. Taking the two sets of results together, it can be seen that embodied energy is more
than half of the operational of the houses studied over the assumed lifetime of 50 years.
A further study in Adelaide (Troy et al, 2003) studied six small urban areas for embodied
and total operational energy consumption where the latter included estimates for the
transport energy expended by residents. The size of the embodied energy components
varied from 22.5% to 45% of the total operational energy.
The role of embodied energy in energy modelling of the built environment has been
referred to in a report by the Australian Greenhouse Office (AGO, 1999a) on greenhouse
gas emissions arising from the residential sector. Although embodied energy was not
considered in the modelling, it was mentioned as being significant being the equivalent of
as much as 20 years of operational energy. For a building with a relatively long life of 100
years, it was stated that the embodied energy for refurbishments and renovations could
amount to the equivalent of a further 20 years of operational energy ie total embodied
27
energy is 40% of operational energy. Furthermore, the report recommended a baseline
study for the embodied energy of residential buildings.
2.4.3 Embodied energy as a sunk cost
(a) The concept of embodied energy as a sunk cost
The concept of treating embodied energy as a sunk cost has been referred to by Stein
(1979) in discussing the energy value of existing stocks of buildings. He suggested that the
built environment represented an enormous investment in both dollars and energy. Stein
proposed that these sunk costs should be evaluated on the basis that urban settlements
allow the full range of human activities to take place.
In the context of cultural heritage, the World Bank (2001) relates sunk costs to cultural
patrimony:
…the key economic reason for the cultural patrimony case is the vast body of
assets, for which sunk costs have already been paid by prior generation, is
available. It is a waste to overlook such assets …
Rypkema (2005) has directly linked the argument by the World Bank to the concept of
embodied energy in the valuing of buildings as part of the stewardship and preservation of
the built environment.
These observations reflect the capital theory approach to sustainable development which
describes the interrelation of man-made, human, social and environmental capital and the
potential for substitution of one type by another (Kohler 2006, Kohler and Wang, 2007).
Pearce (2003) has commented on the large, but as yet largely unquantified, benefits of a
well designed and maintained built environment to human wellbeing and cultural identity.
(b) Comparison with sunk costs in economic models
Reference to the use of the term sunk costs in economics theory provides some background
to a comparison with embodied energy. The significance of sunk costs in economic
models is subject to some interpretation. In microeconomics theory, sunk costs are
conventionally defined as costs that have been irrevocably committed and cannot be
recovered (Wang and Yang, 2001). Hölzl (2005) distinguishes between tangible sunk
costs, that is physical capital such as buildings and machinery, and intangible sunk costs,
such as advertising and technical knowledge. In macroeconomics theory, the conventional
definitions of sunk costs are strongly disputed by Owen (2007) who claims that they are
28
inadequate and dismissive of the significance of these costs. His main criticism is that the
definitions do not consider how sunk cost evaluations might evolve over time due to
unforeseen events and contingencies.
These observations are consistent with the treatment of the embodied energy of the
existing built environment as a sunk cost. Hence, sunk embodied energy includes tangible
assets such as buildings and infrastructure as well as the associated aesthetic, cultural and
social benefits. In recognition of the possible future recovery of some construction
materials from demolished buildings, a portion of sunk embodied energy may eventually
be liberated in the form of construction materials for re-use or recycling. The
transformation of this portion of sunk embodied energy to recoverable embodied energy at
the end of the life cycle of a building is consistent with the views of Owen (2007) and the
evolutionary changes of sunk costs with time. Future costs of construction materials
(influenced by higher energy costs) and probable higher landfill disposal costs are likely to
create the conditions favourable for increased recoverable embodied energy.
(c) Valuing the existing urban environment
The use of sunk embodied energy as a proxy for the buildings, infrastructure and
associated cultural benefits offers a potentially useful method in valuing the existing urban
environment. Since these combined benefits may be difficult to define in terms of
conventional capital accounting, the sunk embodied energy would have a floating value
beyond a quantifiable minimum, where the minimum value was the future recoverable
embodied energy of construction materials destined for reuse and recycling. The
associated cultural benefits within the sunk embodied energy proxy would be represented
by the difference between the embodied energy value for buildings and infrastructure
minimum and the floating value as shown in Figure 2.3.
29
Figure 2.3 Representation of the sunk embodied energy concept
This concept is relevant to the use of embodied energy knowledge of the existing urban
environment in the selection of alternative urban development proposals. Three examples
are discussed which explain this in the context of life cycle energy consumption.
Example 1. Green field site.
With the example of a green field site development, there is no direct input of embodied
energy from the existing urban environment. Total life cycle energy consumption consists
of the operational and embodied energy components as shown in Figure 2.4. However, the
study of energy consumption in the various types of existing residential urban form can
assist in the selection and design of green field developments eg high or low density, low
rise or high rise dwelling forms.
Increasing Embodied Energy
Sunk embodied energy
Associated cultural benefits
Embodied energy of buildings and
infrastructure
Future recoverable embodied energy of building materials
Minimum value
Floating value
30
Figure 2.4 Comparison of the energy consumption of three development proposals
Example 2. Redeveloped site with recycled materials.
This example consists of an older urban area which is redeveloped as a new residential
suburb. Some of the materials generated from demolition activities are re-used or recycled
in the new buildings and infrastructure. This is shown as a lowering of embodied energy
in the form of an energy credit and a subsequent reduction in total life cycle energy
consumption. A knowledge of the embodied energy of materials in older buildings and
infrastructure is required for this analysis.
Example 3. Redeveloped site with the use of infrastructure and recycled materials.
This development is similar to example 2 except that it retains some of the existing
buildings and infrastructure and uses some infill construction. This increases the energy
credit for reused materials and infrastructure. Furthermore, it provides an additional
dimension of the cultural value of utilising existing built form and this is indicated as a
further benefit. For convenience, this benefit is shown as the downward dotted arrow as it
contributes to making the particular redevelopment a more attractive proposition from both
energy consumption and cultural benefit perspectives. A knowledge of the embodied
energy of materials in older buildings and infrastructure is also required for this analysis.
Life
cyc
le e
nerg
y co
nsum
ptio
n of
dev
elop
men
t
Redeveloped site with recycled
materials
Embo
died
ene
rgy
Ope
ratio
nal e
nerg
y
Green field site
Energy credit due to recycled materials
Tota
l ene
rgy
Ope
ratio
nal e
nerg
y
Tota
l ene
rgy
Ope
ratio
nal e
nerg
y
Redeveloped site with use of infrastructure
and recycled materials
Tota
l ene
rgy
Energy credit due to recycled materials and use of infrastructure
Cultural benefits
31
The use of embodied energy as a proxy for sunk costs and associated cultural benefits has
not previously been considered in quantifying the merits of urban development proposals
and this research makes an advance in this concept.
2.4.4 Embodied energy in re-used and recycled materials
The re-use and recycling of construction materials from the demolition of buildings and
infrastructure represents an energy flow in the built environment. By using these materials
as substitutes for new materials, a significant reduction in the embodied energy input to
new urban developments in Australia is possible (Tucker et al, 1993). As part of a
scoping study for the Commonwealth Government into improving the environmental
sustainability of building materials, the importance of re-use and recycling activities has
been emphasised (RMIT, 2006). The cradle-to-cradle concept (Ness & Field, 2003) based
on cyclical flows of materials and products is considered to be a more sustainable model
for manufacturing and construction.
Research into mass and energy flows in the building stock has been suggested by Kohler et
al (1997) in a German context. The value of such an analysis is to predict the
environmental impact of new and recycled building materials. This type of research is
particularly relevant for rapidly developing economies where environmental degradation is
an unfortunate consequence of intense construction activity (Yang and Kohler, 2005).
It has been estimated that between 150 and 250 tonnes of waste materials are generated
during the demolition of a typical house in Australia and part of this contributes to all
waste from building and demolition activities which creates more than half of landfill
waste (EPA, 2001). Tucker et al (1993) found that the ratio of houses demolished to
houses constructed in Australia over the 1954-1991 period was 1:8 and that an average
recovery rate of building materials of up to 52% was indicated. For all demolished
buildings, the rate of recovery was found to be higher for metals although substantial
proportions of lower value materials such as concrete are ‘down-cycled’ into uses such as
roadbase (RMIT, 2006).
Information on the embodied energy of the built form would provide a means of
accounting for energy flows from re-used and recycled materials as part of overall energy
analysis in the urban environment. In addition, the process of deriving the embodied
energy would require a knowledge of the materials quantities in existing buildings and
32
infrastructure, and this could be used to form an inventory of materials stocks to predict
their future availability for re-use, recycling and landfill.
2.5 Knowledge gaps
There are identifiable gaps in the knowledge required to comprehensively analyse energy
consumption in the urban environment and this is an impediment to efforts aimed at
minimising overall energy expenditure. Previous research on energy consumption in the
urban environment has mainly focussed on operational energy consumption of buildings
and the improvement of their energy efficiency as well as transport energy. In recent
years, studies have been carried out on embodied energy as part of life cycle energy
consumption but this has mainly related to individual buildings. There is currently no
convenient way to represent the embodied energy of residential areas (including
infrastructure) in the urban environment which demonstrates the significance of this
component of energy consumption in relation to other energy inputs.
A model which represents the embodied energy of residential areas in an urban
environment would contribute to knowledge in the following ways.
(a) Provide a comparison of modern dwelling configurations.
This applies to newer residential developments and uses the existing urban environment to
inform future development decisions. Different examples of dwelling configurations eg
apartments, town houses, detached house would be compared on the basis of life cycle
energy consumption. The embodied energy of the developments would form a baseline
upon which other components of life cycle energy consumption would be added. This
would enable a comparison of the energy consumption of existing residential areas with
different built configurations to be made. Such comparisons would offer useful guidance
for the design of future residential developments with lower energy consumption.
(b) Offer guidance on sunk costs
The evaluation of sunk costs in terms of embodied energy would provide a means of
valuing the existing built form. This information would be of use when comparing
alternative new urban developments where at least one proposal was based on existing
infrastructure. Urban development proposals utilising existing built infrastructure would
attract an energy credit based on the proposed embodied energy model.
(c) Assist the estimation of additional embodied energy
The embodied energy used to initially construct the built form can be used as a guide to
estimate future additional embodied energy for maintenance, repair and refurbishment.
33
This would be of interest in comparing future scenarios for the redevelopment of existing
suburbs. In the ‘business as usual’ scenario (ie no redevelopment), future energy
consumption would include additional embodied energy which would need to be included
in any comparison with alternative development scenarios.
(d) Quantify the potential for re-use and recycling of materials
A knowledge of the embodied energy of the existing urban environment would provide
information on the quantity of energy to be credited for the re-use and recycling of
materials in new development projects.
(e) Provide a means of compiling a materials inventory
It is necessary to quantify materials in buildings and infrastructure in order to evaluate their
embodied energy. Hence, a model of the embodied energy of the existing urban
environment would provide a means of compiling an inventory of materials stocks for
future re-use.
2.6 Representation of energy consumption in the urban environment
The method to comprehensively represent energy consumption in the urban environment
must have the ability to store and manipulate datasets corresponding to large numbers of
buildings. It should also be able to combine the different energy components consisting of
the embodied energy of buildings and infrastructure, operational energy and transport
energy.
The derivation of energy consumption data for these three main components is problematic
and a basic unit of scale must be selected on which the datasets can be assembled.
Research on embodied and operational energy suggests that some data, albeit very limited,
may be available at the scale of individual buildings. Hence, this research is based on the
assembly of data at that scale. Furthermore, as embodied energy is the first energy
component to be expended in the life cycle of the built form, the research is aimed at
providing a baseline of this energy expenditure upon which other components can be
added.
By themselves, datasets are inconvenient in presenting information in a format required for
a comprehensive analysis of energy consumption in the urban environment. They do not
easily allow for the assembly of the different components of energy consumption data
within urban areas such as neighbourhoods or suburbs. For instance, an estimate of
transport energy used by suburban residents may be achieved by analysing ‘journey to
work’ data from surveys by the Australian Bureau of Statistics. This information is
34
available at the scale of census collectors districts not individual dwellings. Furthermore,
the inclusion of the embodied energy of urban infrastructure such as roads requires an
urban area focus. In addition, datasets do not conveniently allow for the delineation,
comparison and observation of variability of energy consumption in different urban areas.
The provision of a spatial dimension to datasets overcomes these disadvantages and
provides additional benefits. Data maps are a powerful method for displaying statistical
information which direct attention to the substantive content of the data and allow for the
consideration of overall patterns as well as the detection of fine detail (Tufte, 2001). They
enable a layering and separation of information using colour and other devices which
conveys a large volume of data in a small space (Tufte, 2003). The production of data
maps can be accomplished using geographical information systems (GIS) that link
information to geography and allows the visualization and analysis of data. The potential
uses of spatially based data for recording and analyzing issues in urban planning was
described in the early 1990s (Huxhold, 1991) and it has taken some time for its more
widespread use. More recently, Okpala (2001) was advocating the greater use of GIS for
the effective planning of development of cities to achieve sustainability. He confirmed that
GIS was a powerful tool in decision making processes which integrated environmental,
social and economic factors in urban management.
The advantages of using GIS to represent energy consumption has been recognized by
Jones et al (2000; 2001) who developed a tool which assists in predicting the effect of
planning decisions on energy usage at the level of a whole city. The tool was developed
for the city of Cardiff in Wales and is known as the Energy and Environmental Prediction
model. It depicts building energy consumption down to the level of individual post-codes
which, typically, contain 10-15 houses. Domestic energy use was estimated using a
method whereby each house was categorised according to one of 20 types based on
dimensions, age and built form. The method assumes that the energy consumption of
houses in any given category is broadly the same. Hence, having ascribed each dwelling to
a category, the total domestic energy use and greenhouse gas emissions could be estimated.
One use of this model is to estimate the effects on energy consumption of retrofitting the
existing housing stock. Alternative retrofitting measures include double glazing, draught
proofing and improvements in insulation.
35
A more recent development is the use of GIS for depicting greenhouse gas emissions for
residences in Oxford, UK (Gupta, 2005). This research is able to display hot spots of
energy use and CO2 emissions on an individual dwelling level and then aggregate these to
an urban scale – street, district or city level. The next step in the use of this technology is
the analysis of the various options for greenhouse gas reductions and the development of
strategies to bring about these improvements. In terms of a comprehensive energy analysis
which considers the whole life cycle of residential buildings, a shortcoming of both the
Cardiff and Oxford research is that the embodied energy of the houses is not considered
and not depicted in the form of GIS maps.
Research aimed at mapping energy consumption including embodied energy has been
attempted in the city of Toyohashi in Japan (Matsumoto et al, 2006). The authors initially
report on embodied energy intensities (per square metre of floor area) of certain non-
residential buildings including retail stores, education institutional buildings and a hospital.
In 1995, department stores showed the highest embodied energy at 3.91 GJ/m2 followed by
convenience stores. However, the high ranking of the convenience stores was linked with
24 hour opening and illumination which is confusing as this would relate to operational
energy not embodied energy. The embodied energy intensities are also shown to increase
significantly between 1995 and 2000 which is not explained. A model of energy
consumption for the city is described comparing 2003 with 2053 and based on population
changes, variations to the energy supply system and new house construction in both the
central and suburban areas. The city was divided into cells and the energy consumption in
these cells was estimated using the model. Examples of the mapping of total energy
consumption suggest a reduction in the central area with little change in the suburb over
the 50 year period. There was no differentiation in the mapping between operational
energy and embodied energy or the significance of the latter. A possible difference
between future projections for Japanese cities compared with Australian cities is the effect
of substantial migration to the latter which has been a historical feature in urban
development in Australia.
2.7 Summary
The review of literature suggests that embodied energy is a significant part of total energy
consumption in the built environment. Although research has been carried out on the
embodied energy of individual buildings, this has not generally been applied to urban areas
36
such as residential suburbs, nor has it been integrated with other components of urban
energy consumption. An understanding of the embodied energy of the existing built form
can inform decision making on the planning and development of residential areas in the
future. A comprehensive analysis of energy consumption would be greatly assisted by a
model which spatially depicts embodied energy and forms a baseline upon which other
urban energy components can be added.
37
Chapter 3 Embodied Energy 3.1 Introduction
It has been argued in Chapter 2 that an understanding of the embodied energy of the
existing built form can inform decision making on the planning and development of
residential areas. This chapter focuses on the methods used to derive the embodied energy
of construction materials used in dwellings based on input-output analysis and other
datasets (Objective 2 stated in the outline of this thesis). Values for the embodied energy
of materials can then be used to estimate the embodied energy of dwellings and related
infrastructure. This is the first step in the development of a model for the spatial
representation of the embodied energy of residential areas in the urban environment.
3.2 Background
Post Second World War research into the thermal performance of buildings in Australia
was driven initially by thermal comfort considerations (Williamson, 1997). The
subsequent development of thermal simulation models enabled quantitative links with
energy consumption to be made. In the 1970s, attention became focused on energy
consumption as supplies of crude oil became disrupted arising from the Arab oil embargo
and Middle East war. Although factors affecting the operational energy of buildings
became an obvious area of research, the embodied energy of the built environment did not
go unnoticed. Researchers at the Division of Building Research at the CSIRO undertook
some calculations on the embodied energy of a typical house (at the time the term ‘capital
energy’ was more usual to describe embodied energy). Hill (1978) estimated that over the
lifetime of a house, the operational energy consumed in the house would exceed the
embodied energy by over 8 times. They concluded that efforts to minimize energy
consumption in the built environment would be better spent reducing operational energy
rather than embodied energy.
Research into embodied energy in construction was also being carried out in other parts of
the world with a significant contribution from Stein et al (1976) who surveyed energy used
in the U.S. construction industry using input-output analysis. Input-output analysis is
described in the next section. They found that 11.1% of national energy consumption
could be attributed to the construction industry. In the United Kingdom, Boustead and
Hancock (1979) published the Handbook of Industrial Energy Analysis which provided
information on the amount of energy required to manufacture a wide range of industrial
38
materials. In New Zealand, Baird and Chan (1983) found that the construction industry
consumed about 10% of national energy.
In Australia, research in the embodied energy of construction and buildings resumed in the
early 1990s as a result of the possible effect of greenhouse gas emissions from the
manufacture of construction materials on the environment (Tucker et al, 1993). Using
input-output analysis, Tucker and Treloar (1994) estimated that the proportion of national
energy consumed by the construction industry was considerably higher at about 20%. One
of the difficulties for researchers working in this area was the lack of reliable data on
embodied energy. To assist with this problem, Lawson published embodied energy data
for certain types of floor, wall and roof construction (Lawson, 1996) based on typical
values for the embodied energy of the component materials.
3.3 Methods for deriving embodied energy coefficients
The embodied energy of a material or component is the energy consumed in its production
including upstream activities such as raw material extraction, transport, manufacturing and
assembly. The embodied energy coefficient of a construction material is the quantity of
energy used in its production per unit and can be expressed in a variety of ways such as
MJ/kg, GJ/tonne, GJ/m3, etc. When expressed in the same way, comparisons can be made
between materials to select lower energy alternatives.
A classification of energy analysis was defined by the International Federation of Institutes
of Advanced Study (IFIAS, 1974) consisting of 4 levels which depended upon the energy
boundary stipulated and this can be used as a basis for considering embodied energy. A
Level 1 analysis considers only the direct energy consumed in a manufacturing process.
Level 2 encompasses the energy consumed in the preparation and transportation of raw
materials and other pre-manufactured inputs to the process. Higher levels (up to the
maximum of Level 4) consider other indirect energy inputs to a process such as the
primary energy required to supply delivered energy inputs and energy required to fabricate
plant and other process infrastructure.
3.3.1 Process analysis
The direct energy consumed in a manufacturing process (Level 1) can be determined by
process analysis which consists of an energy audit at the manufacturing location of a
product. Hence, for a Level 1 process analysis, the energy boundary will be the factory
39
fence and it is the energy consumed within this boundary which is calculated. For some
manufacturing processes such as fired clay products eg hard burnt bricks and terra cotta
roof tiles, the direct energy (Level 1) can amount to more than 50% of the total energy
consumed (sum of Levels 1, 2, 3 and 4). With other products such as window frames or
ready-mixed concrete, the direct energy is a much lower proportion due to large energy
inputs in upstream raw material manufacture. Hence, a Level 1 process analysis for a
product is substantially incomplete and requires further process analysis of upstream inputs
which becomes progressively more complex beyond Level 2. This results in a significant
disadvantage to the use of embodied energy coefficients derived from process analysis as
data from different sources may have different and uncertain energy boundaries.
Conversely, the advantage of process analysis is that it can be reasonably accurate and
specific to a particular product studied.
3.3.2 Input-output analysis
Input-output analysis is an alternative method for estimating the embodied energy of a
product and depends upon the use of national economic data representing all of the
economic transactions between the industrial sectors and includes imports. This approach
to the analysis of economic interrelationships is based on the original work by Leontief
(1941). Some time later, Bullard and Herendeen (1975) used input-output energy analysis
as a tool for determining the potential for energy savings in the US economy by means of
the substitution of products and services. In Australia, James (1980) developed a model of
the economy using input-output data to study the effects of energy supply variations on
particular industries and related environmental issues. Lenzen (1998) has used input-
output analysis to investigate energy and greenhouse gas flows in the Australian economy
and has developed this further (Lenzen, 2003) to assist in environmental policy design and
analysis. More recently, the application of input-output data has been extended to include
the capital input to products (Lenzen and Treloar, 2004) to provide a more complete
analysis.
Input-output tables for the Australian economy are produced every few years by the
Australian Bureau of Statistics in the form of a matrix of industrial sectors in rows and
columns normally just exceeding 100 x 100 sectors. The input-output tables based on
1996/97 data consisted of a 106 x 106 matrix and were available in the form of Direct
Requirement Coefficients (ABS, 2001a). The direct fiscal inputs of each of the sectors
(rows) are given for a unit output of each of the 106 sectors (columns). Four of the sectors
40
represent the energy supply to the economy, namely sectors 1100 Coal, oil and gas; 2501
Petroleum and coal products; 3601 Electricity supply and 3602 Gas supply. The fiscal
inputs of the four energy sectors to the 106 industrial sectors can be converted to energy
inputs by means of average energy costs making it possible to convert the direct input-
output matrix to a direct energy matrix. The mathematical inversion of direct input-output
matrices to provide total input-output matrices is well established (Leontif, 1966). This
technique enables all of the indirect inputs as well as the direct inputs to be included.
Hence, this approach can provide total energy requirements for particular industrial sectors
which is the equivalent to the sum of Levels 1, 2, 3 and 4 as classified by the International
Federation of Institutes of Advanced Study. These energy requirements are known as
energy intensities (MJ/$) and indicate the amount of energy used for every dollar output of
each industrial sector. Since, certain industrial sectors manufacture particular types of
construction materials (for example sector 2602 Ceramic products is dominated by the clay
brick manufacturing industry), the energy intensity for that sector can be used to evaluate
an embodied energy coefficient for hard burnt clay bricks using an average price for the
product.
An alternative technique for transforming a direct energy matrix to a total energy matrix is
to use the expansion of powers method (O’Connor and Henry, 1975). This is an iterative
process that calculates successive orders of indirect energy inputs. The sum of these
indirect energy inputs in combination with the direct energy inputs is similar to the results
of the matrix inversion providing sufficient rounds of calculations are undertaken,
normally about 8 stages (Miller and Blair, 1985). An advantage of the expansion of
powers method is that data can be scrutinized at each iteration. The relationship of the
expansion of powers and matrix inversion techniques is shown in Appendix 1a.
A range of embodied energy coefficients can be derived using input-output data which has
the advantage of including all levels of energy input resulting in a more ‘complete’ system
compared with process analysis. Furthermore, the coefficients represent national averages
for embodied energy for any given material which avoids the possibility of using
unrepresentative process analysis data from particular manufacturing plants or spurious
process analysis data where the definition of the energy boundary is uncertain.
41
3.3.3 Disadvantages of input-output analysis
There are a number of potential sources of error arising from input-output analysis
originating from the nature and manipulation of the source input-output data, energy costs
and materials prices. A previous study of these errors has estimated that embodied energy
coefficients derived from input-output analysis could be erroneous by as much as ± 20%
(Pullen, 1996).
Energy costs
The method described for deriving embodied energy coefficients uses national average
costs or tariffs for energy. In reality, building materials producers may pay different, and
possibly, lower energy prices which would have the effect of raising embodied energy
coefficients.
Materials prices
The most likely sources of error in building materials prices are associated with those
which were unavoidably derived from current prices (as opposed to those obtained from
1996/97 publications) and converted to 1996/97 values (which is the period when the
input-output data was compared) using building materials price indices.
Input-output data
Common potential sources of error when using the input-output data arise from the
phenomenon of double counting, homogeneity and proportionality (ABS, 2000). Double
counting refers to the inclusion of inter-energy transactions when energy sector
contributions to the relevant building materials sectors are calculated. This particular
source of error is dealt with further in the next section of this chapter and its minimisation
described.
Homogeneity in input-output analysis refers to the extent that the industrial sectors of the
national economy produce a single output and have a single input structure. In the
calculation of embodied energy coefficients, potential errors can arise from the assumption
of homogeneity ie that the energy intensity derived for any industrial sector is
representative of all of the sub-groups within that sector. Proportionality relates to whether
the change in output of an industry causes proportional changes in the amounts of
intermediate and primary inputs.
42
An assessment of potential error in estimating embodied energy in this research is provided
in the context of three case studies presented later in Chapters 5, 6 and 7.
3.3.4. Variations in input-output data
Variations in input-output data can arise from the methods used to harvest data from the
various industrial sectors. Lenzen (2001) reports on a personal communication with the
Australian Bureau of Statistics in 1999 where the standard error of source data items was
reported to be mostly in the 15% to 30% range, with the lowest at just over 1% and the
highest at 58%. The Australian Bureau of Statistics publishes input-output tables at
various cycles ranging from annual to triennial where the data collected refers to an annual
period of economic activity about 3½ years prior to the publication date. A longitudinal
comparison of energy intensities for particular industrial sectors derived using input-output
analysis over 4 rounds of publication serves to highlight possible variations in the data.
Table 3.1 Particular industrial sectors and their energy intensities
Table 3.1 lists 15 industrial sectors which are associated with the manufacture and supply
of construction materials and products. The energy intensities have been evaluated using
input-output analysis based on Table 9. Direct Requirement Coefficients for 1986/87,
1992/93, 1993/4 and 1996/97 (ABS, 1990, 1996, 1997 and 2001a) using the same method
and sources of supplementary data. Energy prices were derived by combining energy
supply data from the Australian Bureau of Agricultural and Resource Economics (Bush et
al, 1999) with the fiscal value of the various energy sectors given by the relevant Product
Code Industrial sector Energy Intensity (MJ/$) 86/87 92/93 93/94 96/97
1400 Other mining 24.8 10.4 14.7 11.6 2202 Textile products 36.0 27.4 24.9 19.0 2301 Sawmill products 35.0 23.7 19.8 21.6 2302 Other wood products 54.0 35.7 40.9 41.1 2503 Paints 67.6 46.3 62.0 37.0 2509 Plastic products 52.4 38.6 70.4 49.0 2601 Glass and glass products 38.6 25.6 17.7 15.6 2602 Ceramic products 45.1 29.6 25.6 23.2 2603 Cement, lime and concrete slurry 65.7 33.0 32.8 41.0 2604 Plaster and other concrete products 115.0 52.1 58.6 39.8 2605 Other non-metallic mineral products 40.3 21.5 15.2 18.7 2701 Iron and steel 24.7 12.8 12.4 18.2 2702 Basic non-ferrous metal and products 24.8 10.4 14.7 11.6 2704 Sheet metal products 36.0 27.4 24.9 19.0 2807 Household appliances 35.0 23.7 19.8 21.6
43
Details data from input-output tables (ABS, 2001b). The method used for combining the
data was the same for each period of input-output tables. The comparison of energy
intensities of relevant industrial sectors rather than embodied energy coefficients for
particular construction materials eliminates the use of materials prices and any associated
errors. The detailed explanation of the derivation of energy intensities and embodied
energy coefficients is given in the next section of this chapter.
Figure 3.1 shows the energy intensities of the selected industrial groups in a graphical form
indicating that some materials are susceptible to change with different datasets. There is
no increasing or decreasing trend of energy intensity with the passage of time over the ten
year period from 1986/87 to 1996/97. It may be inferred that the energy intensities based
on 1986/87 input-output data were exceptional with subsequent sets of data showing much
less variation. Although these differences may be significant for the comparison of
embodied energy coefficients of particular materials over time, their effect is minimised
when considering the embodied energy of whole buildings where there is a substantial mix
of different materials.
0.0
20.0
40.0
60.0
80.0
100.0
120.0
140.0
1400
2202
2301
2302
2503
2509
2601
2602
2603
2604
2605
2701
2702
2704
2807
Industrial Sector Code
Ener
gy In
tens
ity (M
J/$)
86/8792/9393/9496/97
Figure 3.1. Energy intensities for selected construction materials
This raises a further issue that is relevant to the estimation of the embodied energy in the
urban environment where existing buildings are composed of construction materials and
44
components which were supplied and manufactured over a period of time. In the case of
Australian cities this period spans up to over 200 years. This research derives embodied
energy coefficients based on input-output data collected in 1996/97 and applies them to the
existing built form. No attempt has been made to source data relevant to the manufacture
of materials over this period simply because such information required to develop historic
embodied energy coefficients is largely unobtainable. It has been necessary to assume that
the energy required to manufacture a given product in the past is similar to that required in
the present.
There are contradictory historical trends in the manufacture of products including building
materials. On the one hand, manufacturing processes have become more automated and
energy intensive whilst on the other hand, those processes have been subject to continuous
improvement with gains in energy efficiency. The energy intensity (amount of energy
required to produce one unit of output eg MJ/$) of goods and services in the Australian
economy has been studied over a 25 year period until 2000-01 (Tedesco and Thorpe,
2003). After eliminating the effects of structural change (the trend from manufacturing to
services industries) and fuel mix (the change from solid and liquid fuels to electricity and
gas) in the national economy, it was found that any technical improvements in energy
intensity due to technological advances or operational changes had been small in many
industrial sectors over the period.
The approach taken in this research is to estimate the energy embodied in the built
environment at contemporary levels. Put simply, the embodied energy estimates carried
out are the equivalent of the energy cost of replacement. This avoids the difficulty of
estimating historic embodied energy coefficients for materials manufactured in the past.
Replacement energy cost allows for the comparison of different existing built forms to
inform decisions about new developments. It also provides an accounting basis for the use
of existing built infrastructure and recycled construction materials where the embodied
energy of new materials would be avoided.
The use of replacement energy cost has direct parallels with the financial valuing of
buildings and infrastructure where the latter is often owned by governments entities.
Under the ‘deprival value’ approach, assets are valued according to the loss that might be
expected in the case of a government entity being deprived of the service potential of the
assets.
45
Thus the value to the entity in most cases will be measured by the replacement
cost of the services or benefits currently embodied in the asset, given that the
deprival value will normally represent the cost avoided as a result of
controlling the asset and that the replacement cost represents the amount of
cash necessary to obtain an equivalent or identical asset.
(Commonwealth of Australia, 1994).
3.3.5 Hybrid analysis
The combination of process analysis data with that derived from input-output analysis
overcomes many of the disadvantages of both types of analysis and was proposed by
Bullard et al (1978) and is the subject of more recent research in Australia (Treloar, 1997;
Lenzen & Dey, 2000). It offers a greater accuracy in evaluating embodied energy
coefficients as a result of the process energy component as well as the ‘completeness’ of
the analysis afforded by the input-output component.
Two types of hybrid analysis have been defined by Treloar et al (2001a) and later by
Crawford (2005) which are process based and input-output based. Process based hybrid
analysis uses a knowledge of the material inputs to a product as the framework for the
estimating its total embodied energy. The embodied energy of the individual material
inputs are calculated from process analysis for the direct energy (where known) and from
input-output tables for the indirect energy. These are summed in proportion to the material
quantities in the product. The quality of this process data is crucial and should include all
Level 1 energy inputs but no other higher level inputs. The energy of manufacture (direct
energy into the product) can also be estimated from input-output analysis and added to the
total. Clearly, this type of analysis is likely to be more complete for less complex
products.
Input-output hybrid analysis uses the energy intensity (from input-output analysis) of the
industrial sector relevant to the product subject to analysis as the framework for estimating
the total embodied energy. For instance, the energy intensity for the Residential
Construction Sector (ANZSIC group 4101) is used when estimating the embodied energy
of a house. The inputs to this sector of the economy are extracted from the input-output
model using a technique developed by Treloar (1998) and, once identified, the indirect
energy paths are replaced with data derived from process based input-output analysis. This
46
results in a more complete analysis where the total embodied energy can be as much as
twice that derived from process analysis or input-output analysis (Crawford, 2005).
3.3.6 Other approaches
A criticism of the methods described so far is that they are deterministic and present mean
values for the embodied energy of building materials. It is known that different plants
producing the same products can consume different amounts of energy for the same output
depending on a number of factors including the efficiency and age of the manufacturing
plant used and the extent of waste production and its re-use. For example, in a study of
two softwood timber mills at the same South Australian location, one plant consumed
twice as much non-renewable energy for the same output as the other (Pullen, 2000). This
was due to different steam raising and drying practices, materials handling techniques and
amounts of waste timber that are recycled. With construction products which are
manufactured in quite a number of locations, the deterministic approach does not give any
information about the uncertainty or distribution of individual embodied energy
coefficients.
Shipworth (2002) has proposed the use of a stochastic modelling framework using
statistical techniques based on probability distributions. The advantage of this approach is
that it is more suitable to the monitoring of improvements in fossil fuel consumption and
greenhouse gas emissions which is important for those countries who have agreed to
comply with the requirements of the Kyoto protocol. The establishment of baselines for
greenhouse gas emissions and the monitoring of reductions in emissions are an important
part of the protocol compliance. Stochastic data on greenhouse gas emissions resulting
from product manufacture (ie embodied energy) can indicate which industries have a larger
variability in emissions and would benefit from government policies encouraging better
production technologies.
Other researchers have used various statistical methods to estimate embodied energy
coefficients for particular construction materials. In some cases this has amounted to
collecting coefficients from a wide variety of sources and using average values for
particular materials (Lawson, 1996). The importance of ensuring that data is comparable is
emphasised by Hammond and Jones (2006) who have assembled a large database of 1400
records of embodied energy and carbon emissions values for materials in the UK. They
argue the preference for ‘cradle to site’ rather than ‘cradle to factory’ data and provide
47
minimum and maximum values for the embodied energy coefficient of each material as
well as an average value.
3.4 Derivation of embodied energy coefficients
The embodied energy coefficients for construction materials used in this research were
derived using a relatively simple form of hybrid analysis. This commenced with the
evaluation of the indirect embodied energy used to manufacture the product which was
estimated from the difference between embodied energy intensities from the total and
direct input-output data. The second step was to determine reliable direct embodied energy
from process analysis. The quality of this data is crucial and should include all Level 1
energy inputs but no other higher level inputs. The hybrid embodied energy coefficient
can then be calculated for a particular material by combining both the direct and indirect
components.
The input-output analysis was based on the most recently published data at the
commencement of the research, namely input-output tables for the year 1996/1997 (ABS,
2001a). This provided the indirect energy component of the embodied energy coefficients.
The direct components of the embodied energy coefficients of the most common
construction materials were obtained from process analysis. The following subsections
describe the development of the input-output and process analysis methods and the final
derivation of embodied energy coefficients used in this research.
3.4.1 Input-output analysis for indirect energy component
The starting point for deriving the indirect energy components of embodied energy
coefficients based on input-output analysis in this research is the Australian National
Accounts. Input-Output Tables (ABS, 2001a), specifically Table 9. Direct Requirements
Coefficients 1996-97 (see Appendix 1b for further information). The principles of this
method using previous input-output tables have been described earlier by Treloar (1994)
and Pullen (1995). Firstly, the direct energy intensities for the four energy sectors, 1100
Coal, oil and gas; 2501 Petroleum and coal products; 3601 Electricity supply and 3602 Gas
supply are calculated across the 106 output columns which represent the various sectors of
the Australian economy. Each direct requirement coefficient representing an energy input
is divided by the average energy cost and converted to a primary direct energy intensity by
multiplying by a primary energy factor. These are partial primary direct energy intensities
48
arising from each energy sector. The total primary direct energy intensity for each sector
output is then found by summing the four partial energy intensities.
DEI = (DRC x PEF1100) + (DRC x PEF2501) + (DRC x PEF3601) + (DRC x PEF3602) P1100 P2501 P3601 P3602
Where DEI is the total primary direct energy intensity for a
particular sector output (in MJ/$).
DRC is the direct requirements coefficient for that particular
sector output (in $/$100).
PEF is the relevant primary energy factor.
P is the relevant average energy cost (in c/MJ).
Average energy costs for the period 1996/1997 were derived by combining energy supply
data from Table D1 of Australian Energy: Market Development & Projections to 2014-15
produced by the Australian Bureau of Agricultural and Resource Economics (Bush et al,
1999) with the fiscal value data of the various energy sectors given by the relevant Product
Details data from 1996/97 input-output tables (ABS, 2001b). In the case of the energy
sector 1100 Coal, oil and gas, there is a slight mismatch between the energy products
described in the two data sources and selected figures must be used to ensure consistency.
The average energy costs are shown in Table 3.2. The primary energy factors indicate the
amount of primary energy used to manufacture, distribute and supply a unit of delivered
energy to the consumer. To be compatible with the input-output data, they must be
representative of the national economy as a whole and Table 3.2 summarises the primary
energy factors used which were developed by Treloar (1998). These factors are used in
combination with the elimination of inter-energy sector transactions to avoid the ‘double
counting’ of energy inputs. Due to the classification of energy supply sectors in the input-
output tables, there is potential for the double counting of energy inputs. This is avoided
by setting all coefficients in the direct requirements matrix which relate to purchases by the
energy supply sectors to zero.
49
Table 3.2 Average energy costs and primary energy factors
Energy sector Energy value
($million)
Energy supply
(PJ)
Average energy price
(c/MJ)
Primary energy factor
1100 Coal, oil and gas 12478.3 4316.6 0.29 1.2 2501 Petroleum and coal products 10650.6 1686.8 0.63 1.4 3601 Electricity supply 14808.9 718.1 2.06 3.4 3602 Gas supply 1552.9 823.3 0.19 1.4
A 106 by 106 energy matrix is created by multiplying each row of the direct requirements
coefficients by the relevant total primary direct energy intensity (DEI). The sum of the
columns of the energy matrix gives the first indirect energy intensity for each industrial
sector. A second energy matrix is then created by multiplying each row of the indirect
requirements coefficients by the relevant first indirect energy intensity. The sum of the
columns of the indirect energy matrix gives the second indirect energy intensity for each
industrial sector. This iterative process calculates successive orders of indirect energy
inputs and is known as the expansion of powers method. To ensure that the method
converged with that of matrix inversion, 12 rounds of calculations were undertaken. The
total energy intensity (TEI) for each economic sector was determined from the summation
of the direct energy intensity and the 12 indirect energy intensities. Due to the size of the
matrices involved, it is necessary to manipulate data using a computer spreadsheet
software such as Microsoft Excel. A summary of this process is shown in Appendix 1c.
With input-output analysis, the next stage in deriving embodied energy coefficients (in
MJ/kg or GJ/tonne) is to multiply the total energy intensity (in MJ/$) of a particular sector
with a cost (in $/tonne) of the construction material. For instance, the total energy
intensity for the sector ‘2602 Ceramic products’ is multiplied by a national average price
for bricks in 1996/97 corresponding to the year of formation of the input-output tables.
The prices for materials were taken from either Rawlinson’s Australian Construction
Handbook (Rawlinson, 1997) or Cordell’s Building Cost Guide (Cordell, 1997). For a
minority of materials which were not included by either of these two price guides, a local
merchants price was taken and converted to a 1996/97 price by means of Rawlinson’s
historical cost factors. Hence, a series of embodied energy coefficients can be derived
based solely on input-output analysis. However, the method used in this research uses
hybrid embodied energy coefficients where the direct energy is derived from process
analysis and the indirect energy from input-output analysis. For this purpose the difference
between the total energy intensity and the direct energy intensity (ie the indirect energy
50
intensity) is multiplied by the material cost. This indirect embodied energy is then
combined with the direct embodied energy (from process analysis) to form a total
embodied energy coefficient.
EEM = EED + [(TEI – DEI) x CM]
where EEM is the embodied energy coefficient of a material.
EED is the direct embodied energy of a material from process
analysis.
TEI is the total energy intensity of the material from input-
output analysis.
DEI is the direct energy intensity of the material from input-
output analysis.
CM is the cost of the material.
3.4.2 Process analysis for direct energy component
Since the research model is based on the Adelaide built environment, information on the
manufacture of construction materials in South Australia has been sought. This is possible
for materials where there is a significant local manufacturing base such as with clay brick
manufacture and ready mix concrete production. Other common materials such as ceramic
tiles are imported to the state and information has been obtained from likely suppliers.
Details of the direct energy used in the manufacture of certain construction materials from
process analysis is shown in Appendix 2 along with the source of this information and the
final derivation of the input-output hybrid embodied energy coefficient. In most cases,
information on process analysis was used which was relevant to the period of the input-
output data. For instance, data for timber was based on a process analysis carried out in
South Australia in 1997 (Pullen and Varley, 1997) and data for some non-locally
manufactured materials was drawn from the Life Cycle Inventory developed by the Centre
for Design at the Royal Melbourne Institute of Technology (RMIT, 1998).
The data in Appendix 2 is summarized for convenience in Table 3.3 which shows the
embodied energy coefficients of common construction materials derived from the hybrid
analysis. The direct component of the embodied energy coefficient is taken from process
analyses and the indirect component from the product of the price and indirect embodied
energy intensity. A number of embodied energy coefficients were evaluated for concrete
51
corresponding to 20, 30, 40 and 50MPa compressive strength and details of these are given
in Appendix 2.
Table 3.3 Summary of hybrid embodied energy coefficients for common materials
The derivation of the method used to determine embodied energy employed a sample of
records from the State Property Valuation Register with a geographical spread representing
northern, central and southern areas in the linear configuration of metropolitan Adelaide.
This selection comprised approximately half of the records in the ASD and contained a
range of building age from the older central areas to the newer post second world war
developments in the outer suburbs. The sample of records represented 254,165 land
parcels which formed the local government authorities of City of Salisbury (northern
suburb), City of Charles Sturt, Adelaide City Council, City of Norwood, Payneham &
St.Peters (city and inner central suburbs), City of Mitcham and the City of Onkaparinga
(southern suburbs). All of these council areas are part of the Adelaide Statistical Division
and make up approximately one third of the 19 local government authorities in the
Adelaide metropolitan area. The 254,165 records analysed were the current assessment of
property valuation as of 22 December 2003.
The records were supplied by the Land Services Group of the South Australian Department
of Administrative and Information Services in the form of a 42Mb tab delimited text file.
They were intended for research purposes only as described in Appendix 4.
Information on each record was provided in 29 fields as shown in Table 4.1. These fields
include identification numbers, land use code, address, year built, built area, construction
materials and building style. Particular fields or descriptors relating to built area, parcel
area, land use code, year built and construction materials represent important information
for estimating the embodied energy of the buildings on each parcel. The records are
updated on a regular basis as information is received about building approvals from local
government authorities (Shalders, 2007). This includes additions and extension to
buildings which affects the ‘built area’ field.
4.3 Derivation of embodied energy
Since land parcels with dwellings form the majority of the Property Valuation Register, the
embodied energy of houses on each parcel of land was estimated in conjunction with the
data contained in certain fields in Table 4.1. The origins of this method using typical
houses from particular historical periods were devised by the author of this thesis during a
pilot project for evaluating both operational energy and embodied energy consumption of
six selected metropolitan areas in Adelaide (Troy et al, 2003 and Pullen et al, 2002) and
referred to in Chapter 1. The following sections describe the development of this method
66
and its integration with the Property Valuation Register to estimate embodied energy and
depict this information spatially.
Table 4.1. List of fields used to describe land parcels from the Property Valuation Register
Descriptor Details Valnno Valuation number – the number allocated by the Valuer-General
to the ‘assessment” parcel Rectype Valuation type – 1 indicates a combined water and land
assessment, 2 indicates a land assessment only and 3 indicates a water assessment
Luc_code Land use code – a 4 digit code from a comprehensive list of 1017 uses (see Appendix 5a)
Lot_number Lot number Unit_no Unit number House_no House number Street_nam Street name Street_typ Street type Suburb_nam Suburb name Norooms Number of living rooms excluding bathrooms, toilets, etc Year_built Year building constructed Btype Text description for non-residential buildings Barea Built area (see Appendix 5b for definition) Wall_mater Wall material – major material using in external wall construction
up to 34 categories (see Appendix 5c) Roof_mater Roof material – major material used in roof cladding up to 14
categories (see Appendix 5d) Bldstyle Building style (see Appendix 5e for codes) Ensuite Presence of ensuite bathroom in houses Mlystys Number of storeys Buildcond Building condition (Good 7 – 9, Average 4 – 6, Poor 1 – 3) Bedrooms Number of bedrooms Plidtype Plan type (see Appendix 5f) Plnumber Plan number – the reference number of the plan generating the
associated parcel Parctype Parcel type (see Appendix 5f) Parcelno Parcel number – the identifier shown on plan or map which forms
part of the legal land description of title Titlpref Land title prefix Titlvol Land title volume Titlfol Land title folio Area Area of land parcel (in hectares) Postcode Postcode
4.4. Rules for estimating embodied energy
For houses and other residential buildings, a series of rules were devised so that the
descriptors of the building on each land parcel could be used to estimate embodied energy.
The age of the buildings was used as a surrogate identification of certain construction
67
features. Although the Property Valuation Register provided information on wall and roof
materials of existing residential buildings, no information was available from this source
on ceiling and wall heights, type of flooring construction or material used for window
frames. These three factors are important in determining the embodied energy of a house.
The ceiling height can indicate wall height and this parameter affects the area of external
and internal walls. The size of window openings can also affect these areas. The type of
flooring construction determines whether concrete-slab-on-ground/raft footing or
suspended timber floor with strip footings have been used. Window frames are commonly
manufactured in timber or aluminium which have widely different embodied energy.
Furthermore, the information on wall and roof materials did not distinguish entirely
between particular materials with some similarity. Hence, certain assumptions were made
about the construction features and materials depending on the age of the houses. The
following section outlines the rationale for these assumptions.
(a) Floor and footing types
The Property Valuation Register does not provide any information regarding the floor
and footing type used in the construction of houses. The predominant type used in
modern houses is the concrete-slab-on-ground/raft footing. The introduction of this
type of footing/floor system occurred over a period of time as designs and techniques
changed. In discussing the design of footings prior to the introduction of modern
footings, Persse and Rose (1994, p148) state:
There was no significant progress until the early 1960’s when Consulting Engineers
began recommending footings for different problems, e.g. grillage raft and pier and
beam. Raft footings/floor systems prevail today.
In a history of the development of footing types in Adelaide, Fargher (1979) reported
that grillage raft footings were first adopted in 1965. The common shallow stiffened
slab was becoming common by the early 1970s assisted by the development and
improvement of design methods (Walsh, 1975). By the early 1980s, the transition to
the reinforced concrete raft footing in South Australia was complete with more
rigorous design approaches (Mitchell, 1984).
Prior to the concrete-slab-on-ground, the normal footing and floor system consisted of
steel reinforced concrete strip footings with a suspended timber floor. In a review of
68
cracking in houses, Khor (2003) describes the introduction of pier-and-beam footing
for reactive soils in the 1950s by the Department of Mines and Energy. Strip footings
in various forms with timber floors have been in use since the early 1900s. At the turn
of the century, the use of shale quarry overburden with lime mortar was common as a
footing although concrete footings were employed after 1905. Initially, little
reinforcement was used other than a light expanded metal in the top and bottom of the
strip footing (Fargher, 2006). From 1923, when the South Australian Building Act and
Regulation came into effect, mild steel rods were used to reinforce the concrete except
for footings less than 9 inches (225mm) deep. Although not specified, the depth of
reinforced footings was up to 24 inches (600mm) whereas the width was prescribed at
1½ times the aggregate width of the wall which for footings for external walls
amounted to approximately 15 inches (400mm). In the 19th century, limestone rubble
and mortar was the common footing material with sizes varying according to
conditions. These were laid in shallow trenches with common dimensions of 12 – 16
inches (300 – 400mm) deep and 24 inches (600mm) wide. Some further information
on strip footings are provided in Appendix 6 and this has been used in the estimation of
the embodied energy of older houses.
In modern times, by far the most significant change in floor and footing design was the
transition referred to earlier from timber suspended floors to reinforced concrete slab-
on-ground. A further indication of when this occurred can be found in a study of the
thermal performance of houses in three locations in Australia which included the inner
suburb of Norwood in Adelaide (Williamson et al, 1989). As part of this research, the
materials used in the construction of the sample of single storey dwellings in Norwood
were determined by means of a survey of over 800 households. An analysis of these
results shows that the transition from timber suspended floors (with reinforced concrete
strip footings) to reinforced concrete slab-on-ground commenced in the mid 1960s with
the latter dominating house construction by the early 1980s. It should be highlighted
that this study focused on an established inner suburban area which commenced
development over 60 years ago and for this reason the proportion of data relating to the
period in question was relatively small (approximately 60 houses out of the total). In
addition, it is possible that the methods used to construct new houses in older suburbs
may have been conservative compared with housing developments on green-field sites.
The data are presented in Figure 4.2 by taking a chronological midpoint for each of the
periods chosen in the study which were 1983 – 1988, 1968 – 1982, 1938 – 1967 and
older than 1938. This shows that concrete slab-on ground became the dominant
flooring material by the late 1970s.
NOTE: This figure is included on page 69 of the print copy of the thesis held in the University of Adelaide Library.
Figure 4.2 Proportion of houses with either timber of concrete floors (based on data derived by Williamson et al, 1989)
The period of this transition is broadly confirmed by the Timber Development
Association (Lewellyn, 2006) who observed a rapid decline in the use of timber for
floors between 1970 and early 1980s whilst seeing a coincident increase in the use of
timber for brick veneer walls in residential construction.
Based on the survey data and anecdotal information, the end of 1978 was nominated
as the mid point in time for the transition period. For the purposes of evaluating
embodied energy, houses built before this date (up to and including 1978) as shown in
the property register would be treated as having suspended timber floors. Similarly,
houses built after this date would be treated as having concrete slab-on-ground floors.
The potential for error arising for houses within the transition period being assigned
floor and footing types which were different from reality is discussed at the end of the
next chapter.
(b) Ceiling and wall heights
No information on ceiling or wall height is included in the Property Valuation
Register and assumptions must be made for estimating the embodied energy of
residential buildings.
69
In the publication ‘House Styles in Adelaide – A Pictorial History’, Persse and Rose
(1994, p148) summarise walling heights as follows:
Walling heights of 4.27metres (14 feet) or more were not uncommon in the better
class of construction. These were reduced to about 3.66 metres (12 feet) at the turn of
the century and to 3.05 metres (10 feet) after World War 1. After the Second World
War, a further reduction to 2.74 metres (9 feet), due to a shortage of materials,
occurred and in the early 1970’s to 2.44 metres (8 feet).
As suggested, the ‘class of construction’ would have an effect on walling height
particularly prior to the Second World War when building practices were more varied.
Pikusa (1986) provides more detailed information on ceiling heights of particular
nineteenth century houses in Adelaide and these have been summarised in Table 4.2.
Table 4.2 Ceiling heights of particular houses (extracted from Pikusa, 1986).
NOTE: This table is included on page 70 of the print copy of the thesis held in the University of Adelaide Library.
Of course, ceiling and wall heights of a building are not necessarily the same but there
is normally a correlation between in dwellings constructed prior to the Second World
War. After this time, it became common to set the eaves height lower than the ceiling
height such that the height of the outer leaf of a cavity external wall was lower than
the inner leaf.
70
71
Based on this information, the ceiling heights assumed for the evaluation of
embodied energy are 3.6m up until 1900, 3.3m between 1901 and 1945, 2.7m
between 1946 and 1978, and 2.4 m from 1979 to 2003.
(c) Window frame materials
Aluminium window frames were introduced into the South Australian market in the
mid 1960s in a basic mill finish but it was not until the early 1970s that coloured
anodized finishes became available and these frames became a major competitor to
timber (Judd, 2004). Based on this information, it has been assumed that aluminium
was the main window framing material from 1979 onwards.
(d) Wall types
The wall material codes used in the Property Valuation Register include Brick, Iron,
Rendered, Asbestos, Weatherboard, Log, various types of natural building Stones and
Block and the full list can be seen in Appendix 5c. Other composite types are also
specified which include concrete and walls including framing but these codes are only
used to describe the walls of non-residential buildings. For houses, the codes used in
the Property Valuation Register do not distinguish between double brick (either solid or
cavity) and brick veneer construction, the latter possessing considerably lower
embodied energy.
Brick veneer construction was gradually adopted in South Australia during the mid
1960s and became commonplace by the late 1970s. The transition from double brick
to brick veneer coincided with the introduction of articulation in masonry walls to
minimize cracking arising from reactive soils.
The South Australian Housing Trust was an early innovator with brick veneer
construction (SAHT, 1972) for State owned housing. The 1972 Annual Report states:
There has, over the last few years, been a gradual change in the method of house
construction in South Australia. The traditional method of construction is to build a
house of double brick walls supported on a concrete foundation which is an excellent
method for building on good quality, low clay content soils and provides a house of
strength and endurance. However, on poorer soils such a house needs to have special
and costly foundations to prevent walls cracking and often, even then, requires
expensive maintenance. In some localities even the most costly foundations appear to
give no assurance that, with the inevitable soil movement, cracking will not occur.
As Adelaide has grown larger, almost all the stable good building quality soils have
been used. Much of the land now available for development is of doubtful building
character and the Trust is satisfied that in many parts of the State the only sensible
way to build is to use brick veneer construction.
In the private sector, the transition to brick veneer appears to have been slower. The
study by Williamson et al (1989) on the thermal performance of houses in established
suburbs provides some data on this as shown in Figure 4.3 indicating that the
transition occurred by the early 1980s. The Timber Development Association
(Lewellyn, 2006) observed that the majority of new houses on green-field sites were
constructed with brick veneer external walls by the late 1970s.
NOTE: This figure is included on page 72 of the print copy of the thesis held in the University of Adelaide Library.
Figure 4.3 Proportion of houses with either double brick or brick veneer
external walls (based on data derived by Williamson et al, 1989)
Based on this background, brick veneer rather than double brick has been assumed for
houses built in and since 1979 which have been ascribed wall type ‘code 1 – Brick’
for the purposes of estimating embodied energy. Prior to this date, houses with the
wall type ‘code 1 – Brick’ have assumed to be constructed from double brick.
72
73
(e) Roof tile types
The Property Valuation Register includes a field which describes roof type and this
includes 14 variations including Tiled, Galvanised iron, Corrugated asbestos, Steel
decking, Slate and asbestos Shingles. The complete list is shown in Appendix 5d. The
roof type ‘code 1 – Tiled’ includes both concrete and terra cotta (clay) type of roof tiles
and these construction materials have quite different embodied energy. Persse and
Rose (1994) indicate that corrugated steel ie galvanised iron roof cladding was a
common roofing material from early settlement of the South Australian colony through
to the post Second World War period and this material is taken as the default for this
period (1836 – 1945). From that time, concrete roof tiles became the common roofing
material. Terra cotta tiles were used from 1911 with a period of popularity between the
World Wars coinciding with the availability of Wunderlich clay tiles from 1921.
Hence, the specification of roof tile type ‘code 1 – Tiled’ in the property Valuation
Register was assumed to mean terra cotta tiles between 1900 and 1945 and concrete
tiles in the post Second World War period.
(f) Floor to window area ratio
A further change in the design and construction of houses that has occurred in South
Australia since the proclamation of the colony in 1836 and which affects the estimation
of materials used is the size of the external openings, particularly the fenestration. For
the purposes of estimating the quantities of materials in houses from particular periods,
the approximate floor to window area ratios have been examined as summarized in
Appendix 7. This indicates that for existing houses constructed prior to the Second
World War, window sizes were proportionally smaller with the floor to window ratio at
approximately 9 to 10. During the last 60 years, houses were constructed with larger
expanses of glazed area to enhance interior daylighting of homes and the ratio reduced
to approximately 5½ to 6.
4.5 Default combinations of construction features
The most representative combinations of construction features for houses at different
periods of time are summarised in Table 4.3. These combinations were used as a default
when estimating the embodied energy of houses. Essentially, a number of embodied
energy intensities (embodied energy per square metre of floor area) were calculated for
typical (default) houses constructed in the historical periods shown in Table 4.3.
74
Table 4.3 Summary of default combinations of materials
Period Ceiling height
(m)
Footing/floor type Wall construction when ‘Type 1 – Brick’ specified
Window frames
Roof material
1836 – 1900 3.6 Rubble and mortar
strip/timber
Double brick Timber Galvanised iron
1901 – 1945 3.3 Lightly reinforced concrete
strip/timber
Double brick Timber Galvanised iron (terra cotta tiles when ‘Type 1 – Tiled’ specified
In addition to these combinations of main materials and dimensions, further refinements of
the features were used relating to wall and ceiling insulation, window/floor area ratio and
roof pitch as follows:
• Insulation. For the 1979 to 2003 period, R2.5 ceiling with R1.5 wall insulation was
included. For the 1946 to 1978 period, only R1.5 ceiling insulation was included.
No insulation was specified for houses from the earlier two periods.
• Window/floor area ratio. For the two periods since the Second World War, a ratio
of between 9 and 10 was used whereas a ratio of between 5½ and 6 was used for
earlier two periods.
• Roof pitch. For the two periods since the Second World War, a pitch of 20º was
used whereas a pitch of 30º was used for earlier two periods.
The embodied energy intensities were derived using a spreadsheet technique (Pullen,
2000a) which estimates embodied energy from basic dimensions and material
specifications of houses. The spreadsheet can accept embodied energy coefficients for
building materials from a variety of sources to accommodate advances in the knowledge of
energy consumption for the manufacture of building materials. For this exercise, the
hybrid embodied energy coefficients were used as described in Chapter 3 based on process
analysis and input-output tables based on statistics from the period 1996/97 (ABS, 2001a).
Figure 4.4 shows a general representation of the process for estimating embodied energy of
houses.
75
Figure 4.4 General representation of process for estimating embodied energy of houses
(source: Pullen et al, 2002).
Basic data about the dimensions of each typical house is entered into the spreadsheet
including total floor area, wet floor area, external wall area, internal wall area, window
area, insulated roof area, external paved or concreted areas, shed area, pergola or covered
area and fence length. For each building element, the default combination of materials
for the period of the typical house were selected. The materials intensities are factors that
convert areas of building elements to weights of the materials comprising those elements.
For instance, 10m2 of floor area for a post 1979 house will convert to approximately 85 kg
of steel reinforcement, 5 tonnes of concrete, 800 kg of dolomite blinding and 3kg of
polyethylene membrane and 11kg of PVC waste pipe. The material intensities are
embedded in the spreadsheet. Similarly, the embodied energy coefficients which convert
the weights of materials to embodied energy are embedded in the spreadsheet. The
embodied energy of each typical house was expressed as an embodied energy intensity
(MJ/m2) ie per square metre of floor area.
Four spreadsheets corresponding to houses typical of the four periods are shown in
Appendix 8. Initially, the spreadsheet for the period 1979 to 2003 (Appendix 8a) was
derived corresponding to the default combination of dimensions and materials as shown in
Table 4.3 and with the feature refinements. This spreadsheet was modified to form the
default spreadsheet for the period 1946 to 1978 (Appendix 8b) by changing material
Databases and Spreadsheets Property Register calculations
Elem e nt Su b - De ta il A re a o r M a te ria l M a te ria l Ene rg y Em b o d ied Ele m e nt n um b e r Inte nsity C o e ff. Ene rg y
m 2 or no . (kg / m 2) (M J/ kg ) (M J) Pro p . o f(e x ce p t it e m s) Tot a l(%)
01 Fo o tin g s/ Flo o r C o nc ret e sla b o n g ro un d 170 Ste e l 8.6C o nc ret e 528.0Blin d ing 80.0M e m b ra ne 0.3
Susp e n d e d t im b e r Ste e l 5.2C o nc ret e 348.0Bric kw o rk 29.5Tim b e r 18.0
Susp e n d e d t im b e r Ste e l 2.2(AS2870.1 d e sig n - V ic t o ria ) C o nc ret e 165.8
Bric kw o rk 56.0Tim b e r 26.9Dra in s 0.3
05 Ro o f Fra m in g Tim b er 170 Tim b e r 16.8St ee l Ste e l 16.3
C la d d in g C o n c re te Til e 170 C o nc ret e Tile 52.6C la y Tile C la y Ti le 48.1St ee l She e t Ste e l Sh e e t 4.3
Ea v e s so ff it 1.6In su la t io n(R2) 144 In su la t io n 1.0Re fle c . In su l. A lum in ium Fo il 0.4C e ilin g Pla st e rb o a rd 7.6G ut te rin g Ste e l 0.5
06 Exte rn a l W a l ls Do u b le Bric k Bric k(St a n d a rd ) 352.0DPC 0.1M o rt a r 48.6Pla st e r 14.0
Bric k V en e e r St a nd a rd b r ic k 80 Bric k(St a n d a rd ) 176.0M o rt a r 23.4
M o d ula r b ric k Bric k(M o d ula r) 143.0M o rt a r 16.2
Tim b er f ra m in g 124 Tim b e r Fra m ing 7.1St ee l f ra m in g Ste e l Fra m ing 6.2In su la t io n (R1.5) 124 In su la t io n (R1.5) 1.0
DPC 0.1Pla st e r Bo a rd 7.6
AA C Blo c k 200m m t h ic k AA C Blo c k 100.04m m Ren d e r 8.0C o a t ing 0.1Pla st e r Bo a rd 7.6
Tim b e r c la d C la d d in g C la d d in g 10.0Pa in t Pa in t 0.2Tim b er f ra m in g Tim b e r f ra m in g 7.1In su la t io n (R1.5) In su la t io n(R1.5) 1.0
DPC 0.1Pla st e r Bo a rd 7.6
07 W in d o w s Fra m e s Tim b er Tim b e r 16.3A lum in ium 28 Alum in ium 6.0
G la ss 7.5
EM BO DIED ENERG Y SPREA DSHEET
Elem e nt Su b - D eta il Are a or M a te ria l M a teria l En erg y Emb o d ied Ele m en t num b e r Inte n sity C oe ff. Ene rg y
m 2 o r n o. (k g / m 2) (M J / kg ) (M J) Pro p . o f(exc e p t ite m s) Tot a l(%)
01 Fo o tin g s/ Flo o r C o nc re t e sla b o n g rou nd 170 St e e l 8 .6C on c re te 528.0Blind in g 80.0M e m b ra n e 0.3
Susp e n d e d tim b e r St e e l 5 .2C on c re te 348.0Bric k w o rk 29.5Tim b e r 18.0
Susp e n d e d tim b e r St e e l 2 .2(A S2870.1 d esig n - V ic t or ia ) C on c re te 165.8
Bric k w o rk 56.0Tim b e r 26.9D ra ins 0.3
05 Ro o f Fra m ing Tim b e r 170 Tim b e r 16.8Ste e l St e e l 16.3
C la d d ing C o nc re t e Tile 170 C on c re te Tile 52.6C la y Til e C la y Tile 48.1Ste e l Sh e et St e e l She e t 4 .3
Ea v e s so ffit 1 .6Insu la t io n(R2) 144 Insu la tio n 1.0Re fle c . In su l. A lum in ium Fo il 0 .4C e ilin g Pla ste rb oa rd 7.6G utt e rin g St e e l 0 .5
06 Ex te rn a l W a lls D o u b le Bric k Bric k (Sta nd a rd ) 352.0D PC 0.1M o rta r 48.6Pla ste r 14.0
Bric k Ve n e er Sta n d a rd b ric k 80 Bric k (Sta nd a rd ) 176.0M o rta r 23.4
M o d u la r b ric k Bric k (M o d u la r) 143.0M o rta r 16.2
Tim b e r fra m ing 124 Tim b e r Fra m in g 7.1Ste e l fra m ing St e e l Fra m in g 6.2Insu la t io n(R1.5) 124 Insu la tio n (R1.5) 1 .0
D PC 0.1Pla ste r Bo a rd 7.6
A A C Bloc k 200m m th ic k A A C Blo c k 100.04m m Re nd er 8.0C oa tin g 0.1Pla ste r Bo a rd 7.6
Tim b e r c la d C la d d ing C la d d ing 10.0Pa in t Pa in t 0 .2Tim b e r fra m ing Tim b e r fra m ing 7.1Insu la t io n(R1.5) Insu la tio n (R1.5) 1 .0
D PC 0.1Pla ste r Bo a rd 7.6
07 W in d o w s Fra m e s Tim b e r Tim b e r 16.3A lu m in iu m 28 A lum in ium 6.0
G la ss 7.5
EM BO DIED ENERG Y SPREA DSHEET
Elem e nt Su b - D eta il Are a or M a te ria l M a teria l En erg y Emb o d ied Ele m en t num b e r Inte n sity C oe ff. Ene rg y
m 2 o r n o. (kg / m 2) (M J / kg ) (M J) Pro p . o f( ex c e p t i te m s) Tot a l(%)
01 Fo o tin g s/ Flo o r C o nc re t e sla b o n g rou nd 170 St e e l 8 .6C on c re te 528.0Bli nd in g 80.0M e m b ra n e 0.3
Susp e n d e d tim b e r St e e l 5 .2C on c re te 348.0Bri c k w o rk 29.5Tim b e r 18.0
Susp e n d e d tim b e r St e e l 2 .2(A S2870.1 d esig n - V ic t oria ) C on c re te 165.8
Bri c k w o rk 56.0Tim b e r 26.9D ra ins 0.3
05 Ro o f Fra m ing Tim b e r 170 Tim b e r 16.8Ste e l St e e l 16.3
C la d d ing C o nc re t e Tile 170 C on c re te Tile 52.6C la y Tile C la y Tile 48.1Ste e l Sh e et St e e l She e t 4 .3
Ea v e s so ffit 1 .6Insu la t io n(R2) 144 Insu la tio n 1.0Re fle c . In su l. A lum in ium Fo il 0 .4C e ilin g Pla ste rb oa rd 7.6G utt e rin g St e e l 0 .5
06 Ex te rn a l W a ll s D o u b le Bric k Bri c k (Sta nd a rd ) 352.0D PC 0.1M o rta r 48.6Pla ste r 14.0
Bric k Ve n e er Sta n d a rd b ric k 80 Bri c k (Sta nd a rd ) 176.0M o rta r 23.4
M o d u la r b ri ck Bri c k (M o d u la r) 143.0M o rta r 16.2
Tim b e r fra m ing 124 Tim b e r Fra m in g 7.1Ste e l fra m ing St e e l Fra m in g 6.2Insu la t io n(R1.5) 124 Insu la tio n (R1.5) 1 .0
D PC 0.1Pla ste r Bo a rd 7.6
A A C Bloc k 200m m th ic k A A C Blo c k 100.04m m Re nd er 8.0C oa tin g 0.1Pla ste r Bo a rd 7.6
Tim b e r c la d C la d d ing C la d d ing 10.0Pa in t Pa in t 0 .2Tim b e r fra m ing Tim b e r fra m ing 7.1Insu la t io n(R1.5) Insu la tio n (R1.5) 1 .0
D PC 0.1Pla ste r Bo a rd 7.6
07 W in d o w s Fra m e s Tim b e r Tim b e r 16.3A lu m in iu m 28 A lum in ium 6.0
G la ss 7.5
EM BO DIED ENERG Y SPREA DSHEET
Elem e nt Su b - D eta il Are a or M a te ria l M a teria l En erg y Emb o d ied Ele m en t num b e r Inte n sity C oe ff. Ene rg y
m 2 o r n o. (k g / m 2) (M J / kg ) (M J) Pro p . o f( ex c e p t ite m s) Tot a l(%)
01 Fo o tin g s/ Flo o r C o nc re t e sla b o n g rou nd 170 St e e l 8 .6C on c re te 528.0Bli nd in g 80.0M e m b ra n e 0.3
Susp e n d e d tim b e r St e e l 5 .2C on c re te 348.0Bric k w o rk 29.5Tim b e r 18.0
Susp e n d e d tim b e r St e e l 2 .2(A S2870.1 d esig n - V ic t or ia ) C on c re te 165.8
Bric k w o rk 56.0Tim b e r 26.9D ra ins 0.3
05 Ro o f Fra m ing Tim b e r 170 Tim b e r 16.8Ste e l St e e l 16.3
C la d d ing C o nc re t e Tile 170 C on c re te Tile 52.6C la y Tile C la y Tile 48.1Ste e l Sh e et St e e l She e t 4 .3
Ea v e s so ffit 1 .6Insu la t io n(R2) 144 Insu la tio n 1.0Re fle c . In su l. A lum in ium Fo il 0 .4C e ilin g Pla ste rb oa rd 7.6G utt e rin g St e e l 0 .5
06 Ex te rn a l W a ll s D o u b le Bric k Bric k (Sta nd a rd ) 352.0D PC 0.1M o rta r 48.6Pla ste r 14.0
Bric k Ve n e er Sta n d a rd b ric k 80 Bric k (Sta nd a rd ) 176.0M o rta r 23.4
M o d u la r b ri c k Bric k (M o d u la r) 143.0M o rta r 16.2
Tim b e r fra m ing 124 Tim b e r Fra m in g 7.1Ste e l fra m ing St e e l Fra m in g 6.2Insu la t io n(R1.5) 124 Insu la tio n (R1.5) 1 .0
D PC 0.1Pla ste r Bo a rd 7.6
A A C Bloc k 200m m th ic k A A C Blo c k 100.04m m Re nd er 8.0C oa tin g 0.1Pla ste r Bo a rd 7.6
Tim b e r c la d C la d d ing C la d d ing 10.0Pa in t Pa int 0 .2Tim b e r fra m ing Tim b e r fra m ing 7.1Insu la t io n(R1.5) Insu la tio n (R1.5) 1 .0
D PC 0.1Pla ste r Bo a rd 7.6
07 W in d o w s Fra m e s Tim b e r Tim b e r 16.3A lu m in iu m 28 A lum in ium 6.0
G la ss 7.5
EM BO DIED ENERG Y SPREA DSHEETAddress Age Built Area W all type Roof type
Embodied energy of each building
Principal Dimensions
Materials Intensities
Materials & Energy Prices
Input-output Tables
Embodied Energy
Coefficients
76
selections and dimensions. Both of these spreadsheets incorporated a 300mm difference in
inner and outer leaf height of external walls reflecting the construction feature of most post
Word War Two houses where the roof eaves height is lower than the ceiling height. The
1901 – 1945 spreadsheet (Appendix 8c) and the 1836 - 1900 (Appendix 8d) spreadsheet
were then derived with appropriate default material selections and dimensions. These two
spreadsheets were adjusted corresponding to the higher floor to window area ratio, similar
external wall inner and outer leaf heights and higher roof pitch as well as their individual
differences in materials and dimensions.
Each of the four spreadsheets was then adjusted for alternative external walling materials
as defined in the Property Register and the variations in embodied energy intensity were
noted in each case. Depending on the external wall material, appropriate assumptions were
made with regard to internal wall construction. For example, where asbestos cement or
weatherboard external walls were listed (as opposed to the default of brick), it was
assumed that the internal walls consisted of plasterboard lining on a timber frame rather
than brick. After returning to the default combinations of materials and dimensions, a
similar exercise was carried out for the alternative roofing materials and the variations in
embodied energy coefficient was noted. The footings/floor section of the original SEED
spreadsheet (Appendix 8a) was also modified to incorporate materials intensities for strip
footings and timber floors of older houses as detailed in Appendix 6. These are shown in
parts b, c and d of Appendix 8.
Table 4.4 summarises the embodied energy intensities (in GJ/m2) for the default
combinations for houses in each of the four historical periods. In addition, variations from
these intensities are shown corresponding to possible alternative materials which might be
recorded for houses on particular titles in the Property Valuation Register.
77
Table 4.4 Embodied energy intensities and variations for alternative materials (GJ/m2)