Engineering Fundamentals
of
Energy Efficiency
Jonathan M. Cullen
Clare Hall
September 2009
This dissertation is submitted for the degree of
Doctor of Philosophy
Engineering fundamentals of Energy Efficiency
Jonathan M. Cullen
Using energy more efficiently is essential if carbon emissions are to be
reduced. According to the International Energy Agency (IEA), energy
efficiency improvements represent the largest and least costly savings in
carbon emissions, even when compared with renewables, nuclear power
and carbon capture and storage. Yet, how should future priorities be
directed? Should efforts be focused on light bulbs or diesel engines,
insulating houses or improving coal-fired power stations?
Previous attempts to assess energy efficiency options provide a use-
ful snapshot for directing short-term responses, but are limited to only
known technologies developed under current economic conditions. To-
morrow’s economic drivers are not easy to forecast, and new technical
solutions often present in a disruptive manner. Fortunately, the theo-
retical and practical efficiency limits do not vary with time, allowing
the uncertainty of economic forecasts to be avoided and the potential
of yet to be discovered efficient designs to be captured.
This research aims to provide a rational basis for assessing all future
developments in energy efficiency. The global flow of energy through
technical devices is traced from fuels to final services, and presented as
an energy map to convey visually the scale of energy use. An important
distinction is made between conversion devices, which upgrade energy
into more useable forms, and passive systems, from which energy is lost
as low temperature heat, in exchange for final services. Theoretical effi-
ciency limits are calculated for conversion devices using exergy analysis,
and show a 89% potential reduction in energy use. Efforts should be
focused on improving the efficiency of, in relative order: biomass burn-
ers, refrigeration systems, gas burners and petrol engines. For passive
systems, practical utilisation limits are calculated based on engineering
models, and demonstrate energy savings of 73% are achievable. Signif-
icant gains are found in technical solutions that increase the thermal
insulation of building fabrics and reduce the mass of vehicles.
The result of this work is a consistent basis for comparing efficiency
options, that can enable future technical research and energy policy to
be directed towards the actions that will make the most difference.
i
Acknowledgements
The work was supported by the Overseas Research Scheme, the
Cambridge Commonwealth Trust, the C.T. Taylor Fund and the
Clare Hall bursary scheme.
I am grateful for the advice, support and friendship of my super-
visor, Dr Julian M. Allwood, without whom there would be no
thesis.
Thanks must also go to Professor Michael F. Ashby (advisor), Ed-
ward H. Borgstein, Jonas A.F. Roose, Rachel L. Milford (for their
detailed number work and contributions to the journal papers),
and Michael J. Woods, Janne M. Cullen, and Dr Conrad Guettler
(for their proof-reading).
Finally, I thank my family—Maria, Leni and Mabel—and my par-
ents for their faith in me.
This thesis contains fifty six thousand words, eighteen figures
and forty five tables. It is the result of my own work.
This thesis is printed on 100% recycled paper.
ii
Previous publications
Some of the material in this thesis has been published in
Journals and presented at conferences.
Journal papers (published)
1. The role of washing machines in life cycle assessment studies.
Journal of Industrial Ecology, 13(1):27–37, 2009.
doi:10.1111/j.1530-9290.2009.00107.x
J.M. Cullen and J.M. Allwood
2. The efficient use of energy: tracing the global flow of energy from
fuel to service. Energy Policy, 38(1):75–81, 2010.
doi:10.1016/j.enpol.2009.08.054
J.M. Cullen and J.M. Allwood
Journal papers (pending)
1. Theoretical efficiency limits for energy conversion devices.
Accepted for publication in Energy, 2010.
J.M. Cullen and J.M. Allwood
2. Options for achieving a 50% cut in industrial carbon emissions
by 2050. Accepted for publication in Environmental Science &
Technology, 2010.
J.M. Allwood, J.M. Cullen, and R.L. Milford
3. Practical efficiency limits for passive energy systems.
Energy Conversion and Management, 2009.
J.M. Cullen, J.M. Allwood, and E.H. Borgstein
iii
Conferences papers and posters
1. Limits to global energy efficiency. In R’09 Twin World Congress:
Resource Management and Technology for Material and Energy
Efficiency, Davos, Switzerland, September 2009.
J.M. Cullen and J.M. Allwood
2. Steel, aluminium and carbon: alternative strategies for meeting
the 2050 carbon emission targets. In R’09 Twin World Congress:
Resource Management and Technology for Material and Energy
Efficiency, Davos, Switzerland, September 2009.
J.M. Allwood and J.M. Cullen
3. Determining the best options for improving global energy
efficiency. In Transitions Toward Sustainability, 5th International
Conference on Industrial Ecology, Lisbon, Portugal, June 2009.
J.M. Cullen and J.M. Allwood
4. Prioritising energy efficiency opportunities for practical change.
Awarded best poster, in Sustainable Energy: New Solutions from
Physics and Engineering, Institute of Physics, London, United
Kingdom, October 2008.
J.M. Cullen and J.M. Allwood
5. Engineering fundamentals of energy efficiency. Keynote paper, in
2008 Global Symposium on Recycling, Waste Treatment and
Clean Technology (REWAS), Cancun, Mexico, October 2008.
J.M. Cullen and J.M. Allwood
6. Fundamental Analysis of Carbon Emissions for Manufacturing
with Aluminium and Steel. In 4th International Conference of
the International Society for Industrial Ecology, Toronto,
Canada, June 2007.
J.M. Cullen and J.M. Allwood
7. Manufacturing with aluminium and steel: a fundamental analysis
of energy use and carbon emissions. Awarded best paper, in
Institute for Manufacturing First Year Conference, Cambridge,
United Kingdom, May 2007.
J.M. Cullen and J.M. Allwood
iv
Contents
1 Introduction: the efficient use of energy 1
1.1 Climate change and energy related emissions . . . . . . 1
1.2 Technical options for reducing carbon emissions . . . . 3
1.3 Engineering: key to a low carbon future . . . . . . . . . 6
1.4 Organisation of this thesis . . . . . . . . . . . . . . . . 7
2 Review: prioritising energy efficiency options 9
2.1 Units for measuring energy . . . . . . . . . . . . . . . . 10
2.2 Allocating energy into suitable groupings . . . . . . . . 19
2.3 Determining energy efficiency targets . . . . . . . . . . 30
2.4 Proposed framework for assessing efficiency gains . . . . 43
3 Tracing the global flow of energy from fuel to service 46
3.1 Potential gains from energy efficiency . . . . . . . . . . 46
3.2 Drawing a map of global energy flow . . . . . . . . . . . 47
3.3 Results and discussion: what do we now know? . . . . . 56
3.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . 64
4 Theoretical efficiency limits in conversion devices 65
4.1 Constructing a map of global energy efficiency . . . . . 65
4.2 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . 80
4.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . 86
5 Practical efficiency limits in passive systems 88
5.1 Methodology: assessing the practical energy savings . . 89
5.2 Practical energy savings in buildings . . . . . . . . . . . 93
5.3 Practical energy saving in factories . . . . . . . . . . . . 127
5.4 Practical energy savings in vehicles . . . . . . . . . . . 139
5.5 Results and discussion . . . . . . . . . . . . . . . . . . . 160
6 Discussion and conclusions 163
6.1 What new conclusions can now be made? . . . . . . . . 163
6.2 Outline of future research . . . . . . . . . . . . . . . . . 170
6.3 Two promising ideas . . . . . . . . . . . . . . . . . . . . 173
6.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . 178
v
References 181
Figures 195
Tables 196
Nomenclature 198
vi
1 Introduction: the
Efficient Use of Energy
Reducing energy demand by using energy more efficiently is the
most cost effective strategy available for reducing carbon dioxide
emissions. The International Energy Agency (IEA) asserts that
‘energy efficiency improvements in buildings, appliances, trans-
port, industry and power generation represent the largest and least
costly savings’ in emissions.1(p.40) Many nations agree, including
the United Kingdom (UK) which claims the starting point for
addressing climate change risks is ‘to reduce our overall energy
use through greater energy efficiency’.2(p.107) However, despite this
great potential, energy efficiency is often neglected amidst the po-
litical excitement surrounding alternative strategies such as renew-
able energy and the resurgence of nuclear power. It is important
that engineers are actively engaged in the climate change debate,
and give equal focus to both the development of low carbon energy
supplies and technologies which improve energy efficiency.
1.1 Climate change and energy related emissions
‘Climate change is real, and the causal link to increased greenhouse
emissions is now well established.’
David King3(p.176)
Chief Scientific Adviser (2000–2007), UK
Climate change is the most important environmental challenge
facing our world today. The release of greenhouse gases (GHGs)
into the atmosphere, at ever increasing rates, is pushing global
temperatures to elevated levels. Worldwide GHG emissions are
1
§1.1
dominated by carbon dioxide (CO2); the majority of CO2 is re-
leased when fossil fuels are burned for human energy use. There-
fore, any long-term strategy which promotes a low carbon future
must reduce the consumption of energy from carbon bearing fossil
fuels.
The latest report from the Intergovernmental Panel on Cli-
mate Change (IPCC)4 asserts that atmospheric concentrations
of the three primary GHGs—CO2, methane (CH4) and nitrous
oxide (N2O)—have increased significantly due to human activities
over the last 250 years. Ice cores show that current GHG concen-
trations far exceed recorded levels over a period of ten thousand
years. Furthermore, mid-range reference case forecasts suggest
GHG emissions will continue to rise another 50% by 2025. GHG
producing activities have historically been crucial for economic de-
velopment making the reversal of this trend a daunting task for
modern society.
The conclusion of the IPCC panel is that ‘[w]arming of the
climate system is unequivocal’,4(p.5) based on observations of in-
creased air and ocean temperatures, rising average sea levels, and
ice and snow melt. Average near-surface air temperature has
risen 0.74 ◦C between 1906 and 2005, and the rate of tempera-
ture increase is accelerating. Eleven of the twelve warmest years,
recorded since 1850, have occurred in the last twelve years (1995–
2006), and global surface temperatures are projected to rise be-
tween 1.1 ◦C and 6.4 ◦C by the end of this century. This corre-
sponds to an estimated sea level rise of 0.18 m to 0.59 m, without
accounting for future rapid non-linear changes in ice flow. If the
planet continues to warm at current rates, dramatic changes to
the human environment are likely to occur.
The IPCC state that ‘[m]ost of the observed increase in global
average temperatures since the mid-20th century is very likely
[90% likelihood] due to the observed increase in anthropogenic
greenhouse gas concentrations’.4(p.10) To stabilise global mean tem-
perature rise between 2.0 ◦C and 2.4 ◦C above pre-industral levels
2
§1.2
will require a minimum cut in total annual global emissions of 50%
to 85% from 2000 levels, by 2050. Governments have responded to
such forecasts by agreeing a global reduction of 50% before 2050
at the recent G8 Hokkaido Toyako summit.5 National targets are
entering policy as law, for example the UK Climate Change Act
2008 commits to reducing the net UK carbon account for the year
2050, by at least 80% below the 1990 baseline UK.6
The combustion of fossil fuels to provide energy releases large
quantities of CO2 emissions. The World Resources Institute (WRI)
report Navigating the numbers,7 using 100-year global warming po-
tentials, shows that 77% of all GHG emissions are in the form of
CO2—some 32 Gt CO2. The balance is composed of CH4 (14%),
N2O (8%), and fluorintated gases (1%). Approximately 75% of
all carbon emissions are derived from human energy consump-
tion. This has led the IPCC to conclude that ‘[e]missions of CO2
due to fossil fuel burning are virtually certain to be the dominant
influence on the trends in atmospheric CO2 concentration dur-
ing the 21st century’.8(p.12) Thus, reducing energy-related carbon
emissions has become a priority in the current debate surrounding
climate change.
1.2 Technical options for reducing carbon emissions
‘We need to actively reduce our dependence on fossil fuels, moving to
a low-cost, carbon-free energy system, focusing on renewables and on
energy-efficiency gains.’
David King9(p.781)
Chief Scientific Adviser (2000–2007), UK
Achieving climate change targets will require significant techni-
cal changes to the way that energy is supplied and used. The Kaya
identity10,11 expresses the generation of energy-based carbon emis-
sions as the product of four drivers: population, affluence, energy
3
§1.2
intensity and carbon intensity. It has been used widely in liter-
ature, for instance in papers by Schipper et al.,12 Ramanathan,13
and Raupach et al.,14 and forms the basis for the scenario models
used in the IPCC assessment models.4,8,15,16 It can be written in
equation form as:
Carbon = Population× GDP
Population
× Energy
GDP
× Carbon
Energy
(1.1)
The first two terms of the equation are socio-economic drivers.
Placing limitations on population growth or access to economic
wealth is unpopular, despite being influenced to some degree by
political choices. Energy and carbon intensity (the third and
fourth terms) are technical drivers influenced by trends in design
and innovation. Thus, the technical options for reducing carbon
emissions are to use energy more efficiently (which lowers energy
intensity) and to de-carbonise the energy supply (which reduces
carbon intensity).
A simple projection for 2050 is used to illustrate the large tech-
nical changes that will be required to balance modest forecasts of
population and affluence. According to UN report, World Popu-
lation Prospects,17 the global population is estimated to grow from
6.5 to 9 billion over the period 2005 to 2050. Using Equation 1.1,
this has the effect of multiplying CO2 emissions by approximately
one and half times. During the same period, the IPCC 8 expects
global per capita income to rise by around 2% per year, or two and
a half times by 2050. To maintain current CO2 emission levels with
this increase in the socio-economic drivers, will require a four-fold
improvement from the technical drivers. Achieving this target will
be demanding, and yet this makes no additional allowance for ac-
tually reducing annual carbon emissions to the atmosphere.
Emission reduction strategies have to date focused primarily
on carbon intensity, by substituting carbon intensive fossil fu-
els with low-carbon energy sources. This bias is reflected in the
4
§1.2
International Energy Agency (IEA) figures for worldwide research
and development expenditure, where less than 10% is allocated
to energy efficiency in comparison with 40% for nuclear fission
and fusion.1(p.173) Viable decarbonisation options include switch-
ing to less carbon intensive fuels (for instance, from coal to natu-
ral gas), developing more renewable energy sources (for example,
solar, wind, wave and geothermal energy), increasing the use of
bio-energy (from wood and plants) and nuclear energy, or by de-
veloping Carbon Capture and Storage (CCS). Numerous publi-
cations describe the potential of such technologies, including the
papers Decarbonization: doing more with less by Nakicenovic 18
and Renewable energy strategies for sustainable development by
Lund.19 Despite enthusiastic lobbies for nuclear and renewable en-
ergy and the apparent political preference for supply substitution,
there is little evidence that sufficient renewable energy supply will
be available to reach carbon emission targets.
MacKay 20 demonstrates in his book Sustainable energy—
without the hot air that even before considering economic and
social barriers, there is not sufficient renewable energy potential
to meet current UK demand for energy. Van der Veer, the ex-chief
executive of Royal Dutch Shell, explains that most Americans and
Europeans believe renewable energy will replace fossil energy sup-
ply by 2050, whereas even the most optimistic forecasts involving
significant technological breakthroughs limit the growth of renew-
able energy to around 30%.21 Such opinions are supported by the
IEA in their aggressive BLUE scenario which targets a 50% reduc-
tion in CO2 emissions from current levels by 2050. They estimate
that 21% of the emissions savings will come from renewable en-
ergy, 19% from CCS, 18% from fuel switching and 6% from nuclear
generation. In comparison energy efficiency accounts for 36% of
the estimated saving,1(p.65) and this figure is over and above the
generous 0.9% per year baseline efficiency improvements. Energy
efficiency gains can also be achieved at lower marginal costs than
the alternatives.
5
§1.3
If the forecasts above are correct, then a shift in engineering
focus is required towards the development of energy efficieincy
technologies.
1.3 Engineering: key to a low carbon future
‘The engineering profession is uniquely placed to understand what tech-
nologies can be deployed to reduce carbon dioxide emissions as effi-
ciently and effectively as possible because we are responsible for de-
signing and developing new products and technologies.’
Sue Ion22
Vice-President of the Royal Academy of Engineering, UK
Engineers have a responsibility to be actively engaged in the
climate change debate. Ulaby 23 states that it is important for
scientists and engineers to remain at the centre of the climate
change discourse. He laments, however, that the climate change
debate ‘has been co-opted by politicians whose agendas are more
economic than scientific’.23(p.1471) Engineers, according to Ion,22 are
ideally placed to understand technologies that deliver materials,
processes, products and services to society with significantly lower
carbon emissions.
Nevertheless, it is simplistic to assume that technical efficiency
solutions alone will lead to a corresponding reductions in carbon
emissions. Gutowski 24 explains that introducing a new technology
not only reduces environmental impact, but also acts to stimulate
the economy, thus driving up per capita income. This is known
as the rebound effect or Jevons’ paradox,25 after the economist’s
observation that producing and using a resource more efficiently
(in his case coal) often led to greater consumption of the resource,
rather than less (see Alcott 26 and Polimeni and Polimeni 27 for a
discussion of this effect).
Princen 28 suggests that the dominant logic of coupled effi-
ciency and expansion should be replaced with sufficiency. He
6
§1.4
argues that society should not only use energy more efficiently,
but it must at the same time reduce energy consumption. Jevon’s
paradox is real and unavoidable, and requires a politically created
constraint on total energy consumption to compliment technical
advances in energy efficiency. Efficiency is concerned with deliv-
ering the most possible goods and services within the constraints
that society places on energy use. Efficiency does not lead to
demand reduction, but reduces the ‘pain’ of reaching the chosen
reduction target. Engineers must be engaged in every aspect of
this process, from the development new technical solutions to im-
prove efficiency, to the debate concerning the wider economic and
political implications of energy policy.
1.4 Organisation of this thesis
‘All agree that something must be done urgently, but what?’
David MacKay20(p.2)
Author of Sustainable energy—without the hot air.
So, where should engineers focus their efforts? Are the great-
est efficiency gains to be found in light bulbs or diesel engines,
insulating houses or improving coal-fired power stations? What
are the limits to energy efficiency? How should future research
priorities be directed?
This research aims to provide a rational basis for assessing all
future developments in energy efficiency. The task is approached in
much the same way as MacKay tackles energy supply, in Sustain-
able energy—without the hot air.20 It intentionally avoids ethical
questions concerning how much energy humans should consume,
or how best to distribute energy fairly between nations and peo-
ple groups. Neither does it debate the sustainability of future
growth in population, wealth and resource consumption. Instead
it uses simple physical models to determine the fundamental lim-
7
§1.4
its to energy efficiency, limits which unlike economic or political
benchmarks, do not change.
The analysis is global, and therefore imprecise. Energy quanti-
ties are rounded to the nearest exajoule ( EJ = 1018 J), roughly the
total primary energy supply of Portugal. It deliberately focuses
on the technical pieces of equipment—‘conversion devices’ like en-
gines and furnaces, and ‘passive systems’ such as cars and houses—
where engineering solutions can be practically applied, rather than
compare economic sectors or historical efficiency trends.
The chapter outline for this thesis is as follows. Chapter 2
presents a summary of techniques for measuring energy efficiency
and reviews previous attempts to assess the potential of energy ef-
ficiency options. In Chapter 3 , a global map of energy flow from
fuels to final services is constructed, allowing the identification of
the technical components where large efficiency gains are likely
to be found. A novel distinction is made between conversion de-
vices, which upgrade energy into more useable forms, and passive
systems, from which energy is lost as low temperature heat, in
exchange for final services. The theoretical efficiency limits for en-
ergy conversion devices are explored in chapter 4 and the practical
efficiency limits for passive energy systems, in chapter 5. Finally,
in chapter 6 the implications of the research are discussed and a
list of future research projects is proposed.
Together, these chapters aim to provide a thorough exploration
into the engineering fundamentals of energy efficiency.
8
2 Review: prioritising
Energy Efficiency Options
In the 1975 conference Efficient use of energy, Ford et al. 29 state
that the primary objective of any technical energy study is to
define a target ‘standard of performance’ against which current
energy consumption can be compared. Targets may be chosen
from several different options, for example: current best practice,
the extrapolation of an historical trend, the projected gains from
a specific design innovation, or a fundamental physical limit. The
difference between today’s energy use and this target provides a
measure of the possible energy savings, in a device, system or
energy sector. This can be expressed as:
Potential for
saving energy
=
Scale of
energy flow
×
[
1− Target energy use
Current energy use
]
(2.1)
The scale of energy flow can be measured using a variety of
units—for example, joules, barrels of oil, cubic metres of natu-
ral gas, or economic cost—each with its own advantages. Fur-
thermore, global energy flow can be broken down according to
different groupings, such as regions, countries, economic sectors,
technical devices, or consumer products. The first two sections of
this chapter explore the diverse ways of measuring energy (§2.1)
and methods for allocating energy use to activities (§2.2).
The ratio of target energy use to current energy use, is a simple
proxy for energy efficiency. This measure and thus the potential
for saving energy can vary greatly depending on the specific target
performance that is chosen. For example, if a target constrained
by market forces is selected then the economic potential is found,
9
§2.1
whereas a technical potential sets a target based on practical design
and material limitations, and a theoretical potential reflects the
constraints of thermodynamic limits. Figure 2.1, which is adapted
from Dyer et al. 30(p.4437) summarises these approaches for calcu-
lating possible energy savings and demonstrates with indicative
values the wide range of performance targets.
Figure 2.1 Diagram of the potential gains from energy efficiency
Selecting a suitable target efficiency that is both objective and
technically defensible is essential if the full potential of efficiency
measures is to be gauged. Basing long-term targets on economic
potentials—by tracking historic efficiency indicators or surveying
known technologies—is risky because future economic drivers are
difficult to forecast over long time periods. This is in contrast to
technical and theoretical efficiency limits which do not vary with
time. The third section of this chapter (§2.3) presents a critical
review of the current methods used for predicting future efficiency
gains, divided into four groups: comparative methods, top-down
models, bottom-up models and theoretical models.
In the fourth section (§2.4), a new technical framework for
assessing future efficiency gains is proposed. The overall structure
of this chapter is summarised in figure 2.2.
2.1 Units for measuring energy
Smil 31 in his book Energy in nature and society, provides a system-
10
§2.1
2.1 UNIT MEASURES
Primary
Normalised
Quality
Value
Statistical analysis
Input-output analysis
Index decomposition
Process analysis
2.2 ALLOCATION
2.3 PRIORITISATION 
Comparative
Top-down
Bottom-up
Theoretical
2.4 PROPOSED FRAMEWORK 
Figure 2.2 Outline of chapter
atic study of the energy sources, storages, flows and conversions,
which form the complex energy network. The primary aim of this
field of study, known as energetics, is to select unifying energy met-
rics which allow energy flows and transformation to be compared.
Selecting an appropriate unit of measure is challenging.
A variety of measures have been proposed and used histor-
ically, but none has gained absolute universal acceptance. The
most established is the simple unit for energy, the joule ( J), which
is useful for measuring the conversion of energy from one form to
another. Alternative units have been developed: to make com-
parison simpler, to measure the quality of energy in addition to
quantity, and to allow integration with economic measures. Vari-
ous units for measuring energy use have been organised into four
groups, as shown in table 2.1, and are described in this section.
11
§2.1
Table 2.1 Units for measuring energy use
Group Measure Units
Primary Energy J, BTU, kcal
Fuel energy toe, boe, tce
Electricity kWh
Fuel mass kg
Carbon dioxide t CO2
Normalised Energy intensity J/£, J/kg
Carbon intensity t CO2 /J, kg CO2 /kWh
Quality Entropy J/K
Exergy J
Consumer Price £, $
2.1.1 Primary measures
Primary energy is the term used to describe the energy contained
in raw fuels. Bullard and Herendeen 32(p.268) state that ‘primary
energy is extracted from the earth, is processed by the economy,
and ultimately gravitates to final demand’. Energy is conserved
through this process, according to the first law of thermodynamics,
allowing final energy use to be calculated from fuel consumption
data using conventional energy balance methods. The energy con-
tent of fuels can be measured directly in joules, or with the use of
proxies such as mass, volume or carbon dioxide emissions. Most
official statistics are published using primary energy units to mea-
sure supply or demand.
Consumption of primary energy is commonly measured in joules
( J). Alternative energy units include British Thermal Unit (BTU)
which is equal to 1,055 J, and kilogram calorie ( kcal) which is equal
to 4,184 J. Hydrocarbon fuels contain chemical energy and when
combusted release energy which can be used to provide heat or
converted into useful work. The most widely used fuels are oil,
natural gas and coal. Consumption of electrical energy—which is
not considered a form of primary energy—is compared with pri-
mary energy consumption by taking account of the particular mix
12
§2.1
of fuels used to produce the electricity, the conversion efficiency
of power stations and the transmission and distribution losses.
Electricity consumption, including energy from hydroelectric, re-
newable and nuclear sources, is normally measured in watt-hours
(1 Wh equals 3,600 J).
Several proxy units of measure are used for convenience. Due
to the dominance of oil and coal in international trade, energy
statistics for fuels are normally quoted in homogenised units, such
as tonnes of oil equivalent ( toe), barrels of oil equivalent ( boe),
or tonnes of coal equivalent ( tce). Consumption of natural gas is
typically measured in normal cubic metres ( m3). Using mass or
volume proxies allows comparisons to be made between differing
fuels, however errors can arise in the conversion process if accurate
enthalpies of combustion are not available.
More recently, primary energy consumption has been equated
with environmental impact from the release of greenhouse gas
(GHG) emissions to the atmosphere. GHG emissions are typi-
cally measured in equivalent tonnes of carbon dioxide ( t CO2e)
using a 100-year weighting system from the IPCC.8(pp.388–389) Car-
bon dioxide emissions resulting from the combustion of fuels can
be calculated using tables based on the stoichiometric products of
combustion. Emissions derived from the production of electricity
are estimated from a country’s specific generation mix. For exam-
ple, the United Kingdom (UK) DEFRA 33 has published emission
factors (in kg CO2 /kWh) for various fuels and for electricity sup-
plied from the public network. The use of environmental pressure
indicators, such as carbon dioxide, is valuable for focusing atten-
tion on the impacts of energy consumption.
Measuring energy use with primary measures is popular and
well-understood. The use of simple units allows for uncomplicated
aggregation and avoids the introduction of errors through addi-
tional mathematical manipulations. Energy quality and embod-
ied energy are not implicitly measured, leading to some ambiguity
when system boundaries are not defined carefully. Giampietro 34
13
§2.1
also argues that the amount of useful work delivered to an eco-
nomic system is more relevant than the amount of primary energy
consumed. However, the use of primary measures affords compar-
ison of direct energy consumption and provides an overall basis
from which to prioritise energy reduction.
2.1.2 Normalised measures
The normalisation of energy data permits more meaningful com-
parisons between data groups (i.e. countries, sectors, processes or
materials). Normalisation refers to the statistical method of divid-
ing data series by a common variable. This permits the essential
features of the data to be compared in the absence of any influ-
ence from the isolated variable. The most commonly used ratio
or indicator is energy intensity, which broadly refers to the energy
consumed per unit of activity, for example, energy consumption
per unit of economic output ( J/£). According to Schipper et al.,12
the use of energy intensity allows for the comparison of wide ranges
of data, but at the expense of inaccuracies introduced by using an
economic normalisation variable.
Carbon intensity refers to the normalisation of carbon dioxide
(CO2) emissions on the basis of energy consumption ( t CO2 /J) or
electricity generation ( t CO2 /kWh). This measure is commonly
used to rate the environmental impact of different energy supply
options. The term carbon footprint has been used more recently
to define the CO2 emissions per year ( t CO2 /year) for a person,
household, organisation or country. Despite using the term ‘foot-
print’, it does not refer to a physical footprint of land and therefore
differs in concept from the unit ecological footprint.
A separate approach measures energy consumption or carbon
emissions, per unit output of final service. Patterson 35 describes
this as physical-thermodynamic indicator for efficiency, expressed
as:
14
§2.1
η =
physical output (useful)
energy input
(2.2)
or the inverse of this ratio, which Phylipsen et al. 36 have defined as
the Specific Energy Consumption (SEC). Lovins, for example lists
functions for the residential sector as: ‘space heat, water heat, air-
conditioning, refrigeration, cooking, lighting, clothes drying and
other electrical’.37(p.80) Access to detailed physical data enables
energy intensity to be defined in physical terms—for example, en-
ergy consumption per mass of clothes dried ( J/kg). More recently,
Schenk and Moll 38 have argued that the use of such physical in-
dicators leads to a better understanding of energy consumption,
whereas Farla and Blok 39 and Schipper et al. 12 settled on a hybrid
approach using a mix of physical and monetary indicators for in-
dustry, manufacturing and service sectors, but relying on physical
indicators alone for transport, freight and households.
These studies are useful for identifying structural changes in
energy use over several years. However, considerable debate exists
over the most appropriate intensity ratio for assessing changes in
energy use patterns and no one measure is appropriate for all data.
Smil proposes the use of fundamental unifying energy metrics,
such as power density ( W/m2) and energy intensity ( J/kg), for
comparing energy flows and transformations, but also makes the
qualifying statement that:
‘There is no single or best yardstick to assess the performance
of energy transformations; the most commonly used ratio is not
necessarily the most revealing one; the quest for the highest rate is
not always the most desirable goal; and inevitable preconversion
energy losses may be far greater than any conceivable conversion
improvement’.40(p.15)
Despite this, normalisation measures are particularly valuable for
15
§2.1
economic-statistical methods such as input-output analysis and in-
dex decomposition. Normalised energy data can also be compared
easily to defined benchmark values, average consumption figures
and Best Available Technologies (BAT).
2.1.3 Quality measures
Whereas primary and normalised measures are firmly based within
the thermodynamic principle of energy conservation, quality mea-
sures, in addition, attempt to incorporate the second law of ther-
modynamics which asserts that energy has quality as well as quan-
tity. Ahern 41 explains that 1 J of energy at 1000 K can perform
more work than 1 J of energy at 100 K. Therefore, energy at a
higher temperature is more valuable than energy at a low temper-
ature. Work is a higher quality form of energy than heat since
work can be completely converted to heat, whereas not all heat
can be converted to work. Ford et al. 29 states that work is con-
sequently the most valuable form of energy, equivalent to heat at
infinite temperature. The same high value is given to electricity
which for practical purposes is interchangeable with work.
In any real conversion process energy is degraded to a lower
quality, meaning less work is available for any subsequent process.
Irreversibilities in real processes are observed as an increase in
entropy (S), that is not matched by an equivalent production of
work. Thus minimising the generation of entropy is equivalent to
conserving the quality of energy. Entropy is useful for defining the
minimum theoretical energy requirement for a process, as demon-
strated in the iron and steel making study by de Beer et al. 42 It
is also a measure of the disorder or randomness of a system, and
unlike energy, is not conserved.
In thermodynamic literature entropy is described as an exten-
sive state variable (proportional to the size of the system) that is
definable for any material substance or any system, and measured
in joules per kelvin ( J/K). It is calculated using the differential
16
§2.1
quantity dS = δQ/T where δQ is the amount of heat absorbed
in a reversible process (for a system state change), and T is the
absolute temperature. However, C¸engel and Boles 43(p.331) also de-
scribe entropy as a ‘somewhat abstract property’. The problem
for engineers is that classical thermodynamic approaches to re-
versible (ideal) and irreversible (real) processes involve complex
physical and mathematical proofs, often bounded by specific the-
oretical conditions. These bear little resemblance to real processes
and make aggregation of energy data problematic. In response, an
engineering form of thermodynamic property entropy has been de-
veloped: exergy.
Exergy (B) is a measure of both resource quantity and qual-
ity, and is useful for aggregating heterogeneous energy sources and
materials. Exergy can be defined as ‘the potential work that can
be extracted from a system by reversible processes as the system
equilibrates with its surroundings’, from Ayres.44(p.192) Other de-
scriptions include: ‘available work’, ‘the useful part of energy’, ‘the
potential to do work’, ‘free energy’, ‘work capacity’, and ‘the use-
ful work obtainable from an energy source or material’ (see Ford
et al. 29 , Giampietro 34 , Ahern 41 , Cleveland et al. 45). Exergy uses
mechanical work rather than energy as the measurement basis—
mechanical work being the highest quality, lowest entropy form of
energy. Like energy, exergy is measured in joules ( J).
Exergy can be neatly divided into four components: kinetic,
potential (gravitational or electromagnetic), physical (pressure or
temperature) and chemical. For most energy conversion processes
only the chemical component of exergy is significant. Chemi-
cal exergy measures the available work, normally referenced rel-
ative to either the earth’s crust, ocean or atmosphere, as dis-
cussed by De Meester et al. 46 Therefore, the standard chemical
exergy per mole (B) is defined in reference to an equilibrium
state (temperature (T0), entropy (S0) and the component species
(µi0)) found in standard exergy tables, for example Ayres and
Ayres.47(Appendix B–D) Exergy is mathematically defined as:
17
§2.1
B = (H −H0)− T0(S − S0) +
∑
i
Ni(µi − µi0) (2.3)
where H denotes enthalpy and Ni are molar fractions of the chem-
ical elements.
Exergy values for many material resources have been previ-
ously calculated. For fossil fuels, the ratio of chemical exergy to
net calorific value is close to unity—exergy values are only 4–11%
higher for typical fuels according to Ertesvag and Mielnik.48(p.959)
The difference results from the inclusion of post-combustion wa-
ter vapour (lower heating value) and flue-gas components in the
exergy calculation. The conversion of heat, a lower quality form
of energy, to exergy (or mechanical work) is performed by multi-
plying the heat energy by the reversible Carnot engine equation:
|T−T0
T
|, where T is the temperature of the heat carrier and T0 is
the ambient temperature, both in Kelvin.
Despite the potential of exergy as an absolute measure of en-
ergy quality, it is seldom used in global energy analysis. De Meester
et al. 46 reasons that exergy data for many resources, in particular
mineral resources, is still incomplete and inconsistent. However,
perhaps a greater barrier is the conceptually challenging nature of
quality measures, and the economic preference for the more crude
unit of measure, primary energy.
2.1.4 Value measures
The price paid for energy is perhaps the most comprehensive mea-
sure of the utility of a fuel. In neoclassical economic theory the
price of a fuel (per energy equivalent) equals its marginal value
product, or economic usefulness. Cleveland et al. 45 argue that a
fuel’s price encompasses factors such as energy density, scarcity,
cleanliness, emission profile, flexibility and ease of storage. Price
is therefore a value-based unit of measure. Economic transactions
18
§2.2
are measured in monetary currencies such as the British pound
(£) or the United States of America (USA) dollar ($). For multi-
country studies, transaction data is adjusted according to cur-
rency exchange rate indexes or Purchasing Power Parities (PPP)
databases (for example the International Financial Statistics (IFS)
database which is maintained by the International Monetary Fund
(IMF)). Price data collated over longer time periods must be cor-
rected for inflation.
The most obvious benefit of using price to measure energy is
the availability of detailed data for analysis. Accounting practices
ensure that energy production and consumption is measured in
monetary terms at all levels of society. Price based measures are
also familiar to consumers. However, the use of value measures
is not immune to market imperfections—Cleveland et al. 45 report
that energy prices often fail to include many negative social and
environmental impacts (externalities) associated with energy con-
sumption.
There are numerous ways to measure energy, each with its own
advantages. Ideally, for a physical based study, a quality measure
of energy would be chosen. Yet, Giampietro 34(p.177) comments that
the ‘history of energy analysis is the history of the struggle with
the conundrum of how to deal with the problem of aggregation of
energy forms of different quality’. In practice, the unit of measure
is more likely to be selected based on the coverage and accuracy
of the available energy data.
2.2 Allocating energy into suitable groupings
In order to compare the energy efficiency of two conversion devices,
the energy flow through each device must be known. Several dif-
ferent approaches have been developed to allocate energy flow to
diverse grouping such as countries, sectors, conversion devices or
19
§2.2
Table 2.2 Methods for aggregating energy use
Method Advantages Limitations
Statistical
analysis
Simple, established and
commonly used; broad
coverage
Measures only first
order, direct inputs;
errors from survey
collection practices and
aggregation methods
Input-output
analysis
Aggregates data at the
sectoral level; accounts
for higher order inputs
(direct and indirect)
Data collection is both
time-consuming and
error prone; conversion
from monetary value to
energy use is difficult
Index
decomposition
Separates out the
impacts of influencing
variables; provides
understanding
Mathematical models
introduce errors;
neglects contributions
from non-energy
emissions
Process
analysis
Accurate and specific
within a defined system
boundary; highlights
possible energy savings
Conditional on the
chosen system
boundary; truncation
leads to errors
products. Four methods for collecting, ordering and allocating en-
ergy consumption data are described in this section; these are sum-
marised in table 2.2. Three of the analysis methods—statistical,
input-output and process—are derived from the original work of
Chapman 49 in 1974. Input-output analysis and process analysis
were also described in detail by Bullard et al. 50 four years later.
Index decomposition, a more recently developed method, has been
added to the list. A brief description of each method is given, fol-
lowed by literature examples and a discussion of the advantages
and limitations of each approach. The best choice of methodology
is determined case by case, taking into account the accuracy of the
energy data, the chosen unit of measure, the data coverage and
the system boundary of the study.
20
§2.2
2.2.1 Statistical analysis
General statistical energy data is collected and published by inter-
national organisations, governments, industry sector associations
and large companies. Energy data is normally extracted from sur-
veys completed by relevant stakeholders and published typically
on a yearly basis. Attention is focused on primary energy from
fossil fuels (coal, oil and gas) and electricity (nuclear, hydro and
renewable sources) because it is easier to collect data from large
centralised energy systems. Energy sources such as food, direct
sunlight and biomass are typically ignored, despite preliminary
estimates by Haberl 51 revealing that unaccounted biomass con-
tributes 235 EJ/year or 39% of global human ‘energetic’ needs.
The law of energy conservation allows primary energy data to be
divided up according to various groupings and tracked through
numerous energy transformations. Thus snapshots of energy use
in society can be taken from almost any angle and energy data
can be aggregated by simple addition.
Several international organisations collate primary energy data
at a global level and publish annual energy reports. Examples in-
clude the United Nations (UN) Energy Statistics Yearbook 2003,52
the International Energy Agency (IEA) World Energy Outlook
2004 53 and the World Energy Council (WEC) Survey of Energy
Resources 2004.54 In addition, some private companies report on
global energy data, for instance BP 55 and Enerdata.56 Interna-
tional publications report on trends in overall energy consumption
rates and also track changes in energy distribution over time and
between energy sources, countries and sectors. Entities such as
the European Union (EU) and the Organisation for Economic Co-
operation and Development (OECD) collate regional energy data
from member countries.
In most cases, global energy data is sourced from the gov-
ernmental agencies of individual countries. Government agencies
typically publish their own energy statistics, for example, the UK
21
§2.2
Department of Trade and Industry (DTI)57 and the USA Energy
Information Administration (EIA).58 Detailed energy data is col-
lected from numerous sources including trade records, private com-
panies, government department records and fuel tax accounts.
Trade sector associations represent commercial and industrial sec-
tors at international and national levels and collect energy data
from detailed surveys. Two such associations include the World
Coal Institute59 and World Steel.60 Large companies aggregate en-
ergy information across numerous sites in order to assess perfor-
mance against targets. Energy benchmarking of this type, espe-
cially in relation to climate change impacts, is an important aspect
of corporate sustainability reporting.
Recent attention to climate change impacts has prompted or-
ganisations to collect and publish GHG emission data. From a
global perspective, two reports are particularly valuable: Navi-
gating the Numbers from the World Resources Institute (WRI)7
and Key GHG Data produced by the United Nations Framework
Convention on Climate Change (UNFCCC).61 The UK Carbon
Trust 62 has made an ambitious attempt to attribute carbon emis-
sions from primary fuel consumption throughout multiple levels
of the economy. Six ‘carbon accounts’ were chosen, culminating
at the level of ‘high-level consumer needs’ including: recreation
and leisure, space heating, food and catering, household, health
and hygiene, clothing and footwear, commuting, education, other
government and communication. Such methods of redistributing
carbon emissions are important for linking consumer actions di-
rectly with environmental impacts.
Some reservations remain over the accuracy of statistical data
analysis. Energy data is derived from first order energy inputs (or
direct inputs) and therefore excludes energy inputs from higher
orders. For example, the energy used in the process of refining oil
or the energy required for construction of a steel mill, is typically
ignored. Farla and Blok 63 also point out that survey collection
practices are subject to errors from: incomplete surveys, limited
22
§2.2
sector coverage (requiring scale-up), errors in interpreting ques-
tionnaires and publication mistakes. Accounting for discrepancies
in energy definitions, energy types, system boundaries, non-energy
use and industry classifications, causes further errors according to
Karbuz.64 For these reasons, global energy statistics from different
organisations rarely agree, and the aggregation of energy data from
several different sources is difficult. However, statistical methods
are well established, accepted and readily available for making en-
ergy comparisons.
2.2.2 Input-output analysis
Input-output analysis is an economic-statistical approach used to
determine energy demand at lower sectoral levels. Monetary values
of transactions between various sectors of an economy are collated
into a square input-output matrix. Each sector is listed as a sup-
plier (in rows) and a consumer (in columns) in the input-output
matrix. For a matrix A, the element Aij represents the supply
of resources from a sector i in order to produce one unit of out-
put from the sector j. The matrix approach has advantages over
primary energy analysis because a total energy demand for each
sector can be calculated which sums all direct and indirect energy
inputs. This is demonstrated in the following input-output matrix
example, adapted from Boustead and Hancock.65
Table 2.3 shows that to produce 1 unit of steel requires 0.1 units
of steel and 0.2 units of electricity, and to produce 1 unit of elec-
tricity requires 0.3 units of steel and 0.4 units of electricity. These
values measure the first-order or primary resource demand, and
can be written mathematically as the matrix A.
The second order consumption for steel and electricity can be cal-
culated using the same matrix, as shown in table 2.4.
Therefore the matrix for the second order consumption is:∣∣∣∣∣ 0.07 0.150.10 0.22
∣∣∣∣∣ =
∣∣∣∣∣ 0.1 0.30.2 0.4
∣∣∣∣∣
2
= A2
23
§2.2
Table 2.3 Input-output matrix example
Consumer
Supplier Steel Electricity
Steel 0.1 0.3
Electricity 0.2 0.4
A =
∣∣∣∣∣ 0.1 0.30.2 0.4
∣∣∣∣∣
It can be shown that the third order consumption corresponds
to A3 and so forth. The total resource consumption B is given by:
B = A+ A2 + A3 + . . . = (I − A)−1 − I (2.4)
where I is the identity matrix.
Input-output analysis was originally developed by Leontief 66
to predict the economic effect caused by changes to an individual
sector or industry. Vringer 67 shows that if energy sectors (e.g.
coal, crude oil, natural gas, nuclear and hydro-electricity) are in-
cluded in the matrix, energy demand can be attributed to each
economic sector. The total energy requirement of final delivered
goods is calculated by applying mathematical operators to the ma-
trix in the form of the energy intensity vectors or physical intensity
vectors.
The use of input-output analysis to determine total energy
requirements was first applied in 1975 by Bullard and Herendeen 32
using 1967 data covering 357 USA sectors, and by Wright 68 using
1968 data covering 90 UK sectors. Other country specific energy
Table 2.4 Second order consumption for a unit of steel
To produce 0.1 units of steel 0.2 units of electricity Total
Steel 0.1× 0.1 = 0.01 0.2× 0.3 = 0.06 0.07
Electricity 0.1× 0.2 = 0.02 0.2× 0.4 = 0.08 0.10
To produce 0.3 units of steel 0.4 units of electricity Total
Steel 0.3× 0.1 = 0.03 0.4× 0.3 = 0.12 0.15
Electricity 0.3× 0.2 = 0.06 0.4× 0.4 = 0.16 0.22
24
§2.2
analyses have been performed as economic input-output data has
become available. Recent studies have also focused on: the energy
impacts of trade between nations (for example, Mongelli et al. 69);
quantifying CO2 emissions at the sector level (Rhee and Chung
70);
and evaluating energy-use in specific sectors such as household
consumption (Kok et al. 71).
The following limitations to input-output analysis have been
summarised from Boustead and Hancock,65 Bullard and Heren-
deen 32 and Wright:72
Monetary values Equating monetary transaction data with
physical fuel quantities is difficult. The conversion depends on
choosing accurate energy intensity values; these cannot account
for fuel price variances between sectors, large and small consumers,
and over time.
Available data Input-output data is collected separately from
other national statistical data. The process is both time consum-
ing and costly. Thus input-output data is often released several
years late and not all countries collect data, preventing energy
comparisons on a global scale.
Data accuracy Data is collected from industries and companies
using surveys. Incomplete sector coverage, variance in collection
methods and differing time periods lead to data error. Further in-
accuracies are introduced from companies which produce multiple
products of varying energy intensities. The approach also fails to
capture the embodied energy in capital goods and non-combusted
fossil fuels.
Trade The effect of imported and exported goods on energy-use
is difficult to quantify. Typically it is assumed that foreign tech-
nology has the same energy intensity as domestic technology in
the absence of accurate data from exporting countries.
Despite these limitations, input-output analysis has proved valu-
able for establishing sectoral trends for energy consumption.
25
§2.2
2.2.3 Index decomposition
Index decomposition—sometimes referred to an indexing or fac-
toral decomposition—is an approach used for isolating the drivers
for change in resource consumption, over a time period. The tech-
nique has been applied to food, water, transport, manufacturing
and household resource use. The recent trend to set and com-
pare national energy efficiency and carbon emissions targets has
resulted in numerous studies which decompose energy and carbon
emission indicators.
Hoekstra and van der Bergh 73 report that decomposition be-
gins with the identification of a suitable indicator, sub-groups and
a time period, for which the driving forces are to be examined.
Indicators for energy related decomposition include absolute en-
ergy use, energy intensity, carbon intensity and CO2 emissions.
Energy data may be disaggregated into sub-groups (sector, coun-
try, fuel type, etc.) according to the availability of data over the
selected time period. The collected data is then decomposed to
isolate the effect of drivers such as structural changes (e.g. the
shift from heavy industry toward commercial activities), demand
changes (increased overall resource consumption) and technology
changes (which result from improved efficiencies of processes), for
example see Liu and Ang.74 Decomposed data can also be reag-
gregated into new groupings to reveal more valuable information.
According to Schipper et al. 12 decomposition of changes in
energy use (E) can be described by the ASI equation:
E =
∑
Ai Si,j Ii,j (2.5)
where A represents overall sectoral activity (value added) in each
sector i, S is the structure of each sector i expressed as a share of
subsector j, and I represents the energy intensity of each subsector
j (in S).
If the dimension of fuel mix is introduced, changes in CO2
26
§2.2
emission (G) can be decomposed using:
G =
∑
Ai Si,j Ii,j Fi,j,k (2.6)
where F represents the carbon content of each fuel k, used in sub-
sector j of sector i. The Kaya identity (discussed in section 1.2) is
frequently used in climate change literature for decomposing CO2
emissions. Numerous decomposition models have been proposed;
these can be divided by their mathematical form into additive or
multiplicative models, as demonstrated by Hoekstra and van der
Bergh.73 A substantive review of energy and environmental de-
composition studies has been performed by Ang and Zhang.75 The
authors classify 124 studies by: application area (energy and emis-
sions), indicator type (quantity, ratio/index and elasticity) and de-
composition scheme (multiplicative/additive and specific method).
Decomposition studies are valuable for separating out the ef-
fects of various influencing variables. Nevertheless, the method
is limited by the accuracy of energy data (normally statistical or
input-output based) and the choice of drivers. Care must be taken
to avoid attributing changes to a single driver which in practice is
influenced by several others factors, or overlooking large increases
in one variable which are cancelled out by reductions in other
variables, resulting in almost no variation at the indicator level.
When used to decompose carbon emissions, the technique cannot
account for non-energy related emissions nor the use of industry
feedstock fuels.
2.2.4 Process analysis
For the three top-down approaches described above—statistical,
input-output and decomposition—vast coverage of energy con-
sumption comes at the expense of technological richness. Process
analysis attempts to capture this detail by breaking down complex
systems into a network of simple bottom-level operations. Using
this modular approach ‘all industrial processes, no matter how
27
§2.2
complex, can be subdivided into a sequence of operations linked
by a flow of materials’ (Boustead and Hancock 65(p.71)). The inten-
tion is to account for all embodied energy inputs to the system,
including contributions from for instance: fuels, raw materials,
capital, machinery, maintenance, prior processing of raw materi-
als, transportation of inputs and outputs and business overheads.
Process analysis begins with the selection of a target product,
which can be either a good or a service. Direct inputs required
to make the target product are listed, including both energy in-
puts (e.g. fuels) and non-energy inputs (e.g. raw materials and
machinery). For example, figure 2.3 shows four inputs to a target
product, labelled A through D. Next, the indirect inputs required
for the production of A are listed, and so forth for B, C and D.
The final process energy requirement can be calculated by sum-
ming energy contributions from all direct and indirect inputs to
the system.
Direct energy contributions are relatively easy to measure.
Companies normally record inputs (material and energy) and out-
puts (products and wastes) in both physical and monetary val-
ues. However, Lenzen and Dey 76(p.578) point out that indirect
(higher-order) energy contributions are ‘manifold, complex and
therefore difficult to assess’. Difficulties arise because of the en-
ergy interdependence between processes and industries. Quoting
from Boustead and Hancock, ‘The steel industry for example sup-
plies a proportion of its output to the electricity industry which in
turn feeds electricity to the steel industry’.65(p.67) Energy contri-
butions to capital items such as machinery are particularly prob-
lematic when the machine is made from the target product being
evaluated. Chapman 49 suggests some practical solutions to deal
with such feedback loops, which include estimating an approxi-
mate value and iterating or solving using simultaneous equations.
Some materials (e.g. steel and cement) and energy sources (e.g.
fuels and electricity) are, in practice, inputs to production process
in almost every sector.
28
§2.2
PRODUCTION
OF TARGET
PRODUCT
PRODUCTION
OF A
PRODUCTION
OF B
PRODUCTION
OF C
PRODUCTION
OF D
Inputs to
target
product
LEVEL 1 LEVEL 2 LEVEL 3
Inputs
to A
etc.
Figure 2.3 Levels in process analysis
In addition to complex feedback loops and interactions, higher-
order inputs to a given target product are theoretically limitless.
In theory, higher-order contributions diminish in importance al-
lowing truncation of the analysis at a level where additional in-
puts are insignificant in relation to the sum of all the energy inputs.
In practice, Lenzen and Dey 76 have found that for the manufac-
ture of basic iron and steel products in Australia, process analy-
sis underestimated the energy consumption by approximately half
(19 MJ/kg), in comparison to input-output analysis (40.1 MJ/kg).
Examples of process analysis in literature include: Bates et al. 77
who evaluate potential emissions reductions for the EU transport
29
§2.3
sector, Michaelis et al. 78 and Sakamoto et al. 79 who study the
iron and steel industry. Hybrid approaches have been proposed
by Bullard et al. 50 and Treloar 80 to limit errors due to trunca-
tion. The hybrid method makes use of the accuracy of process
analysis for direct and first-order inputs, and the wider coverage
of input-output analysis for higher-order contributions. Thus the
truncation errors of process analysis are replaced by smaller ag-
gregation errors of input-output analysis.
Process analysis delivers results that are accurate and specific
within a defined system boundary. Its main advantage is the in-
timate experience gained with the physical processes and equip-
ment which use energy, permitting energy saving opportunities to
be identified. However, the extensive energy and material flow
data for both direct and indirect inputs makes it unsuitable for
large energy studies. Truncation errors and higher-order energy
contributions are inherent problems.
Four methods for the allocation of energy flow data have been pre-
sented: statistical analysis, input-output analysis, index decompo-
sition and process analysis. If an appropriate unit of measure is
selected and energy flow is carefully allocated, then the activities
which use energy can be compared on an equal footing. The rel-
ative scale of energy use can then be evaluated and avenues for
improving energy efficiency explored.
2.3 Determining energy efficiency targets
Good efficiency targets are based upon sound estimates of the po-
tential savings from efficiency measures. This requires accurate
energy consumption data, organised into relevant groupings, and
a method for identifying and assessing potential efficiency gains.
Efficiency studies found in literature range from comprehensive
surveys of efficiency technologies to the review of a few isolated
30
§2.3
case studies, and from the tracking of top-down efficiency indica-
tors to detailed thermodynamic studies.
Four generic approaches to prioritising energy efficiency oppor-
tunities have been identified:
Comparative methods which compare energy use or carbon
emission data across countries, sectors or products
Top-down models which track historical trends in efficiency in-
dicators and extrapolate these in the future
Bottom-up models which survey best-practice efficiency mea-
sures and aggregate the savings
Physical models which calculate efficiency limits based on physics
and engineering principles
These approaches are described and critically reviewed below.
2.3.1 Comparative methods
A simple comparison of the energy use in different activities is help-
ful for identifying where efficiency measures are likely to deliver
the greatest gains. For example, some countries consume more
energy than others; these countries present an obvious place to
begin looking for energy savings. Comparisons are typically made
across very different groupings—international regions, countries,
industrial sectors, consumer products or time intervals—and are
sometimes based on alternative indicators such as carbon emis-
sions.
The Sankey diagram, first used by the Irish engineer Riall
Sankey in 1898,81 has become an important graphical tool for com-
paring the scale of energy flow. In these diagrams the quantity
of energy (or sometimes emissions) is traced through society as
arrow or lines, with the line width being proportional to energy
flow. This allows the dominant energy flows to be quickly identi-
fied. An early example, entitled Pathways to end uses maps the
31
§2.3
flow of energy in the United States Summers.82(p.150) More recent
examples include the Global energy flows diagram produced by the
IPCC83(p.259) and the Navigating the Numbers, GHG diagram by
the WRI which attributes the worldwide greenhouse gas emissions
to end-use activities.7(pp.4–5)
Energy and carbon emission data is also compared between in-
dustrial sectors and processes. For example, the UK DTI has pub-
lished energy consumption tables comparing 23 industrial sectors
against 9 end-use processes (e.g. lighting, motors, space heating,
etc.). The impact of energy reduction initiatives can be assessed
by monitoring changes in such indicators over time, and comparing
to benchmark figures. The USDOE 84 and the EU 85 among others,
publish best-practice energy technology case studies for compara-
tive purposes.
Comparisons between consumer products are frequently per-
formed using Life Cycle Analysis (LCA) principles. This offers
a methodology for counting energy inputs and outputs across all
life cycle phases of a product, and the ability to make compar-
isons with alternative products. The conclusions that result are
accurate between equivalent product systems, but are not ‘abso-
lute’ due to irregularities in boundary system definition. Cullen
and Allwood 86 suggest that LCA studies underestimate the im-
pact of indirect energy inputs (i.e. transport, equipment and cap-
ital goods) and introduce errors from overlapping product system
boundaries, especially between product use phases. A consequence
is that when LCA studies are used for prioritisation, they are in
danger of overemphasizing the use-phase impacts and overlooking
the impacts from indirect activities. For these reasons, Cullen and
Allwood warn practitioners to be wary of using LCA for prioritis-
ing action.
Recently, several studies have attempted to measure the mag-
nitude of end-use CO2 emissions from consumer activities. For
instance the Carbon Trust 62 compares the impact of ‘high-level
consumer needs’ in order to make consumers aware of the activ-
32
§2.3
ities that drive carbon emissions and point to possible emission
reduction options. Such measures are helpful for highlighting the
need for change, and for obtaining information about the latest
energy reduction technologies.
Comparative methods make no attempt to quantify the poten-
tial to reduce energy consumption, making them largely unsuit-
able for determining efficiency priorities. Furthermore, statistical
energy studies in their current form lack sufficient coverage and
technical detail to be useful as a basis for setting global efficiency
targets. Two specific problems are discussed in more detail: the
failure to trace energy completely from fuels to services and the
lack of focus on the technical areas where efficiency gains are found.
Fuel to service: Current statistical energy studies and Sankey di-
agrams stop short of tracing the entire length of each energy chain,
from fuels to services. It is these final services—a comfortable
thermal environment, the illumination of a work space, mobility
for people and goods—that satisfy human needs and desires, not
energy itself nor the complex network of energy chains. By termi-
nating the energy flows at the sector level, current analyses fail to
make a distinction between the devices which convert energy into
useful forms (e.g. engines, electric motors, furnaces, and light-
bulbs) and the energy systems which transform this energy into
final services (e.g. vehicles, buildings, and factory systems). Yet
devices and systems are interconnected, and energy savings in one
reduces the potential for savings in the other.
This idea is explained using an example from the climate change
literature. In their paper on stabilisation wedges, Pacala and So-
colow 87 suggest two efficiency measures to improve the operation
of the world’s 2 billion cars in 2054. The first wedge requires in-
creasing fuel economy in cars from 30 to 60 miles per gallon (mpg),
saving 1 billion tonnes of carbon ( Gt C). The second wedge in-
volves decreasing the average annual travel per car from 10,000 to
5,000 miles per year, also saving 1 Gt C. In each option, half of the
33
§2.3
total carbon emitted from cars is saved. Yet, if both wedges were
implemented perfectly, the reduction in carbon emissions would
not equal 2 Gt C—found by adding the savings from both wedges—
because this requires the 2 billion cars to produce no emissions at
all. Instead, the savings would be only 1.5 Gt C found by multi-
plying, not adding, the carbon savings. Such examples of over-
estimating energy and carbon savings are common in the energy
efficiency literature.
Significant reductions in energy demand and carbon emissions
are available from improving the systems which deliver energy ser-
vices. Increasing the insulation in buildings and reducing the mass
of vehicle bodies are just two tangible examples. However, the
separation between devices and systems is seldom mentioned in
literature and almost never used in the calculation of practical
efficiency limits. Nakicenovic et al. introduced the term ‘service
efficiency’, defined as ‘the provision of a given task with less useful
energy without loss of “service” quality’.88(p.422) The intention was
to separate efficiency measures, for example using a more fuel-
efficient engine, from conservation measures, such as improving
the flow of traffic or improving the car aerodynamics. They com-
ment that in many cases, the conversion of energy in upstream
devices is highly efficient, yet the ‘low efficiency of the last link
in the chain, namely the provision of energy services, drastically
reduces the overall efficiency’.88(p.435)
The Untied Nations Development Programme (UNDP) World
Energy Assessment report makes a distinction in theory between
conversion devices and the ‘technology producing the demanded
services’,89(p.176) and provides examples including building materi-
als, window systems, insulation, and light-weight vehicles. They
argue ‘. . . energy efficiency can be improved—and energy losses
avoided—during the often overlooked step between useful energy
and energy services’.89(p.175) However, in the detailed data anal-
ysis that follows the potential energy savings are still not sepa-
rated into devices and systems, but instead aggregated under the
34
§2.3
broader category of end-use efficieincy.
Technical focus: Current comparative studies will typically trace
primary energy through electricity generation, and then divide the
energy flows into broad commercial sectors (e.g. transport, build-
ings and industry) for which statistical data is readily available.
This approach proves useful for monitoring a sector’s energy use
over time or directing high-level energy policy, however it fails to
focus on the specific technical components in each energy chain,
from which efficiency gains are achieved. For example, electric
motors are not found in a single economic sector, but have numer-
ous applications across transport, industry and buildings. There-
fore, an efficiency gain in electric motors will translate into savings
across all sectors, and yet this is not implicitly clear from current
energy Sankey diagrams. Attempts to map energy flows through
technical devices have been made at the national level, most no-
tably the United States Department of Energy (USDOE) Energy
footprints for the industrial sector,90 however a technically focused
global diagram has yet to be published.
Comparative energy studies will continue to be the dominant
choice for energy analysts. Knowing the scale of energy flow is
critical for determining the potential of efficiency options. Yet for
the purposes of this research, current statistical energy analyses
fail to trace global energy flow completely from fuel to services,
and focus on economic sectors rather than the technical devices
and systems where efficiency solutions can be applied.
2.3.2 Top-down models
Scenario based projections of future energy use have become pop-
ular for energy policy decision making. Complex macroeconomic
models, using top-down analysis, are required for determining the
impact of factors such as economic growth, population growth,
35
§2.3
technology changes, scarcity of resources and climate changes.
Scenarios are created by adjusting important variables to eval-
uate the effect of possible policy interventions. Historical data
is collected over multiple years and is used to forecast future
trends. Major international energy agencies publish multiple sce-
nario studies including the IEA World Energy Outlook 2006,91
the Intergovernmental Panel on Climate Change (IPCC) Climate
Change 2007:The Physical Science Basis,4 and the Pricewater-
houseCoopers The World in 2050, prepared by Hawksworth.92
Top-down models are also used to track historical trends in ef-
ficiency indicators and extrapolate these into the future, to deter-
mine energy efficiency targets. By extrapolating historical trends
in energy indicators, estimates of future advances in efficiency can
be made independently of current technology options. For exam-
ple, in the World Energy Outlook reference scenario, the IEA pre-
dicts that the global average energy intensity (a measure of global
energy efficiency) will fall on average by 1.7% per year from 2004 to
2030, based on the past 30 year trend.93 Others make similar pro-
jections: IEA 53 predicts global energy intensity (primary energy
per Gross Domestic Product (GDP)) will fall by 1.5% per year un-
til 2030; and Pacala and Socolow 87 forecasts a baseline 1.96% per
year improvement in carbon intensity (carbon emissions per GDP)
over the next 50 years, based on USA goal announced in 2002; and
continuous improvements in energy and carbon intensity underpin
the projections in the IPCC 16 scenarios.
Long range forecasts are particularly sensitive to small changes
in such indicators. Given that historical trends in energy inten-
sity are only documented over short periods (20–30 years) it seems
imprudent to extrapolate these trends as far as 50 years into the
future. This raises the question of whether economists can ac-
curately model such trends over long periods and thus places the
accuracy of long-range forecasting in doubt. Two specific problems
with long-range forecasts are presented in further detail.
Firstly, the extrapolated efficiency target may be unachievable
36
§2.3
because it exceeds some theoretical or practical limit. An annual
improvement in efficiency of 1.7% equates to a 35% saving by 2030
and an impressive 55% by 2050. For many technical devices, such
gains may not be physically possible, leading to an exhaustion of
the innovation potential if alternative solutions cannot be found,
as discussed by Blok.94
Secondly, these models assume that the underlying structural
components of energy demand are stable and predictable over long
periods. In contrast, Craig et al. assert that:
‘[l]ong-run forecasting models generally assume that there exist
underlying structural relationships in the economy that vary in a
gradual fashion. The real world, in contrast, is rife with disconti-
nuities and disruptive events, and the longer the time frame of the
forecast, the more likely it is that pivotal events will change the
underlying economic and behavioural relationships that all models
attempt to replicate.’95(p.87)
For example, Raupach et al. 14 show that the declining trend in
global energy intensity from 1980 to 2000, has in recent years re-
versed, placing in doubt many predictions of future energy demand
and associated carbon emissions. The difficulty of making accu-
rate forecasts is also discussed by Farla and Blok,63 Karbuz 64 and
Focacci.96 In practice, future predictions based on extrapolation of
energy trends are rarely accurate, prompting a leading academic
in the field of energy, Vaclav Smil, to comment that ‘long-range
energy forecasts are no more than fairy tales’.31(p.154)
Despite advances in modelling techniques and computational
power, the engineer should avoid the temptation to view future
scenarios as factual. Scenario based approaches are of limited
value for setting efficiency targets because they do not assess the
potential for energy reduction nor highlight new technical oppor-
tunities. They are useful only for predicting short term macro-
economic trends.
37
§2.3
2.3.3 Bottom-up models
Bottom-up models survey best-practice efficiency technologies and
estimate their combined potential for reducing energy demand.
The identification of efficiency opportunities typically involves a
detailed review of emergent technologies within a sector or in-
timate knowledge of an energy system. Mitigation potential is
evaluated using energy, carbon and cost metrics, and the scope of
analysis can range from case studies to full global assessments.
The case study approach typically starts with an energy re-
duction target in mind (often appropriated from a scenario) and
then searches from the bottom-up for technologies with the po-
tential to reduce energy consumption. Identified efficiency options
are then analysed and ranked according to their energy reduction
potential. Finally, the individual energy reductions are summed,
or scaled up to be compared with the reduction target. Results of
case studies are sometimes published in popular science format, for
example: Factor Four: Doubling Wealth, Halving Resource Use by
von Weizacker et al.,97 and Heat: How to Stop the Planet Burning
by Monbiot.98 This format provides a valuable catalyst for public
debate.
Industrial based case studies are often confined to single sector
or a selection of individual efficiency options which are relevant
to the process operation. Worrell et al. state that in many cases
‘it is not possible to provide an all-encompassing discussion of
technology trends and potentials.’99(p.2) Instead, in their report
on emerging energy-efficient technologies in industry, they focus
on a number of selected key technologies: near net shape casting,
membrane technology, gasification, motor systems and advanced
cogeneration. This process of choosing technologies is valid for
industrial energy analysis, where time scales are short, but can
be biased towards current and emerging technologies and risks
overlooking a potentially valuable energy saving solution.
Wider scope assessments of efficiency options (and alternative
38
§2.3
mitigation options) have been performed by various international
and governmental organisations. Some recent examples include:
the EU Action Plan for Energy Efficiency,100 the IEA Energy
Technologies at the Cutting Edge,101 and Her Majesty’s Treasury
Energy Efficiency Innovation Review.102 These reports are sub-
stantial undertakings and typically cover a range of both supply
and demand technologies. The Action Plan for Energy Efficiency,
prepared by the EU,100 specifically targets technologies within 4
end-use sectors, allowing the projected energy saving per sector to
be compared (see table 2.5).
Table 2.5 Available energy savings from end-use sectors
Sector Current Business Potential
demand as usual savings
2005 2020 2020
EJ EJ EJ
Residential households 10.7 12.9 3.5
Commercial buildings 6.0 8.1 2.3
Transport 12.7 15.5 4.0
Manufacturing industry 11.4 14.6 3.6
Notes: data from EU.100 1 EJ = 1018 J = 26.1 Mtoe
A useful tool for visualising available energy or carbon sav-
ings is the abatement cost curve. These curves are constructed by
plotting the marginal cost of abatement, for example in £/t CO2,
versus the reduction potential in t CO2. Some well known ex-
amples of abatement curves include: the Global climate abatement
map by Vattenfall;103 the McKinsey Global Institute report, Curb-
ing global energy demand growth;104 the IPCC bottom-up analysis
for sectoral mitigation in 2030;83 the IEA marginal abatement cost
curves for sectors in the report Energy Technology Perspectives.1
Such studies provide a useful snapshot of current economic and
technological drivers, and show where efficient technologies can be
immediately applied.
39
§2.3
Nevertheless, bottom-up models in their current form are in-
complete for two reasons. Firstly, these models often ignore the
complex chains of technical devices and systems in the energy net-
work. Efficiency gains at different points in the network cannot
simply be added together, because a saving in one device often
reduces the potential for gain in a connected device. For exam-
ple, a more efficient electric motor requires less electricity for the
same load, reducing the demand for generation and therefore the
absolute benefit of efficiency gains in that upstream generation.
Secondly, bottom-up models assess only known or emerging
technologies that have evolved under today’s economic drivers and
technical conditions. Surveys of current efficiency options identify
mostly incremental gains to existing processes and tend to over-
look opportunities from novel disruptive technologies or divergent
development pathways, which are beyond the influence of indus-
try. If for instance, the cost of energy were to rise dramatically and
hold for several years, then a completely new set of efficiency tech-
nologies would emerge, and require the practical efficiency limits
to be revised. However, if an absolute measure which is indepen-
dent of today’s economic drivers is used, then the potential savings
from future, yet to be invented technologies, can be found.
2.3.4 Theoretical models
When efficiency performance targets are based on the potential
of existing technologies, they provide only one possible pathway
for future development. They therefore fail to consider alternative
pathways which are still unknown, and can become trapped in a
particular technology route. Instead, in theoretical models, the
targets are based on the theoretical limits to efficiency, derived
from fundamental physical laws. Using this approach, current en-
ergy use is compared, not to the potential of best practice available
technologies which will change with time, but to a fundamental
minimum energy requirement which is static. This helps to iden-
40
§2.3
tify the technical areas where further efficiency gains are likely to
be found.
Theoretical models define an absolute target by calculating an
upper efficiency limit based on thermodynamics. When using such
models it is impossible to set a target which is thermodynamically
impossible and the analysis is not constrained by currently known
technologies or industrial practice. The thermodynamic property
exergy (discussed in section §2.1.3 on quality units of measure)
shows how far each device is operating from its thermodynamic
ideal, allowing all energy conversion devices to be compared on an
equivalent basis.
Detailed exergy models exist for many individual conversion
devices and include useful breakdowns of exergy losses. However,
the use of exergy modelling has tended to be confined to energy
efficiency studies in industry. For example, de Beer et al. 42 cal-
culate the minimum theoretical energy required for production of
primary steel (using the blast furnace route) and secondary steel
(using the electric arc furnace route), as shown in table 2.6.
Table 2.6 Energy consumption for steel 1990
Specific energy required Primary Secondary
( GJ/t steel) steel steel
Best practice 19.0 7.0
Minimum theoretical 6.6 negligible
Minimum realistic †12.5 3.5
World-wide average 24
Notes: †A further reduction of up to 2.5 GJ/t steel may be achieved
using heat recover techniques from hot steel. World-wide average is a
weighted average of both production routes. Data from de Beer et al. 42
Large differences between best practice and minimum theoreti-
cal energy requirements are noted: 12.4 GJ/t steel and 7.0 GJ/t steel
respectively for primary and secondary steel. The minimum real-
istic values are based on the theoretical minimum, but include
additional energy requirements that are practically difficult to
41
§2.3
eliminate. For example, although it is theoretically possible to
make steel at room temperature, in practice steel is melted during
production. Therefore, the minimum realistic values for primary
and secondary steel include an additional 1.05 GJ/t steel required to
heat and melt the steel. The margins between current and min-
imum energy have been divided by de Beer et al. 42 into energy
loss groupings, in an attempt to qualify whether potential exists
to reduce the consumption from each group.
The USDOE 84 Industrial Technologies Program (ITP) aims to
improve the energy efficiency of industrial process in the USA.
‘Energy bandwidth studies’ have been published for the most en-
ergy intensive industries (for example, aluminium, cement, chem-
icals, forest products, mining, petroleum refining, and steel) cov-
ering 75% of all industrial energy consumption. An energy band-
width analysis ‘identifies the theoretical minimum amount of en-
ergy required for each major operation within a given industry, the
current amount of energy that is used in that operation, and the
difference between the two’. Sponsored reports are prepared for
each intensive industry which draw from the published work of aca-
demic and industry stakeholders. For example the Steel Industry
Energy Bandwidth Study prepared by Energetics 105 makes refer-
ence to reports by Stubbles,106 Energetics 107 and Fruehan et al..108
This programme has proven invaluable for improving energy effi-
ciency in industry.
Beyond the industrial sector, theoretical studies of entire so-
cieties are occasionally performed. The first exergy analysis of
an entire society was published by Reistad 109 and estimated the
overall efficiency of the United States to be 21%. A review paper
by Ertesvag 110 summarises a further 15 societal exergy studies,
including coverage of numerous countries, regions, and one global
study by Nakicenovic et al..111 Rosen et al. 112 stress that exergy
analysis has an important role to play in charting the increase of
energy efficiency in society, because it clearly identifies possible ef-
ficiency improvements and reductions in thermodynamic loss. The
42
§2.4
analysis by Nakicenovic et al. estimates the global efficiency of en-
ergy conversion in 1990 to be about 10% of the theoretical limit,
but the paper is highly technical and difficult to comprehend for
a non-expert reader. Although many exergy analyses have been
performed on individual conversion devices, these are also tech-
nical in nature and typically appear in specialist thermodynamic
journals. Attempts to aggregate exergy information for conversion
devices into an accessible global form are rare, and for this reason
theoretical models are often overlooked when determining research
priorities and creating energy policy.
Nevertheless, using a theoretical basis to assess energy conver-
sion devices provides an absolute basis for identifying and ranking
efficiency options. This requires comparing the current energy use
conversion devices with the theoretical minimum energy to provide
the same output. Using a purely theoretical measure of efficiency
promotes an ideal which may not be practically achievable, either
economically or technically. Yet it provides a useful theoretical
target and an absolute basis from which to measure progress.
The four current approaches are summarised in table 2.7.
2.4 Proposed framework for assessing efficiency gains
Previous efforts to assess the potential savings from efficiency mea-
sures are useful for identifying options and directing responses in
the short term. Yet, current efforts are unlikely to be accurate over
the times scales being negotiated in climate change policy because
of their reliance upon recent economic trends and known technical
options. Using an absolute measure of efficiency, such as exergy
analysis, avoids the uncertainty which results from the extrapo-
lation of economic trends and captures the potential of yet to be
discovered efficiency designs. However, the use of exergy analysis
for directing priorities has to date had limited application, due to
its perceived complexity and the lack of worldwide studies.
43
§2.4
T
ab
le
2.
7
C
u
rr
en
t
ap
p
ro
ac
h
es
fo
r
p
ri
or
it
is
in
g
en
er
gy
u
se
re
d
u
ct
io
n
A
p
p
ro
a
ch
D
es
cr
ip
ti
on
L
im
it
at
io
n
C
om
p
ar
at
iv
e
A
g
g
re
ga
te
d
en
er
g
y
d
a
ta
is
co
m
p
ar
ed
b
et
w
ee
n
re
gi
on
s,
se
ct
or
s,
or
p
ro
d
u
ct
s.
S
p
ec
ifi
c
or
av
er
ag
e
va
lu
es
ca
n
b
e
co
m
p
ar
ed
to
be
st
p
ra
ct
ic
e
te
ch
n
o
lo
gy
.
O
p
p
or
tu
n
it
ie
s
fo
r
en
er
gy
re
d
u
ct
io
n
ar
e
n
ot
as
se
ss
ed
.
F
ai
lu
re
to
tr
ac
e
en
er
gy
fr
om
fu
el
to
se
rv
ic
es
or
fo
cu
s
on
te
ch
n
ic
al
ca
te
go
ri
es
.
T
o
p
-d
ow
n
H
is
to
ri
ca
l
tr
en
d
s
in
ec
o
n
om
ic
an
d
en
er
gy
in
d
ic
at
or
s
ar
e
u
se
d
to
p
ro
je
ct
fu
tu
re
co
n
su
m
p
ti
on
an
d
en
v
ir
on
m
en
ta
l
im
p
ac
ts
.
M
o
d
el
li
n
g
a
p
p
ro
a
ch
es
a
ll
ow
p
ot
en
ti
al
n
ew
te
ch
n
ol
og
ie
s
or
p
ol
ic
ie
s
to
b
e
in
tr
o
d
u
ce
d
,
an
d
th
ei
r
im
p
ac
ts
to
b
e
ev
al
u
at
ed
.
T
ar
ge
ts
ca
n
ex
ce
ed
a
ec
on
om
ic
or
te
ch
n
ic
al
li
m
it
.
A
ss
u
m
e
k
ey
in
d
ic
at
or
s
(s
u
ch
as
en
er
gy
in
te
n
si
ty
)
ar
e
st
ab
le
ov
er
lo
n
g
ti
m
e
p
er
io
d
s,
w
h
ic
h
is
u
n
li
ke
ly
.
B
ot
to
m
-u
p
C
u
rr
en
t
a
n
d
em
er
g
in
g
te
ch
n
ol
og
ie
s
fo
r
re
d
u
ci
n
g
en
er
gy
u
se
ar
e
sy
st
em
a
ti
ca
ll
y
ev
a
lu
at
ed
.
C
as
e
st
u
d
ie
s
ar
e
p
er
fo
rm
ed
at
th
e
p
ro
ce
ss
o
r
se
ct
or
le
ve
l.
C
ar
b
on
an
d
en
er
gy
ab
at
em
en
t
cu
rv
es
su
m
m
a
ri
se
a
n
d
ra
n
k
p
o
te
n
ti
al
op
ti
on
s
ac
co
rd
in
g
to
p
re
d
ic
te
d
co
st
s
of
im
p
le
m
en
ta
ti
o
n
.
T
im
e
co
n
su
m
in
g.
B
as
ed
on
su
rv
ey
s
of
k
n
ow
n
te
ch
n
ol
og
ie
s
w
h
ic
h
ig
n
or
e
fu
tu
re
so
lu
ti
on
s.
O
v
er
lo
ok
th
e
co
m
p
le
x
en
er
gy
n
et
w
or
k
w
h
en
ad
d
in
g
effi
ci
en
cy
ga
in
s.
T
h
eo
re
ti
ca
l
C
u
rr
en
t
en
er
g
y
u
se
is
co
n
tr
as
te
d
w
it
h
th
e
th
eo
re
ti
ca
l
m
in
im
u
m
re
q
u
ir
em
en
ts
.
T
h
is
p
ro
v
id
es
an
ab
so
lu
te
m
ea
su
re
of
effi
ci
en
cy
w
h
ic
h
is
co
m
p
a
ra
b
le
b
et
w
ee
n
d
iff
er
en
t
en
er
gy
co
n
ve
rs
io
n
d
ev
ic
es
.
S
et
id
ea
li
st
ic
ta
rg
et
s
w
h
ic
h
ca
n
n
ot
b
e
ac
h
ie
ve
d
ec
on
om
ic
al
ly
or
te
ch
n
ic
al
ly
.
E
x
er
gy
an
d
en
tr
op
y
ar
e
n
ot
w
el
l
u
n
d
er
st
o
o
d
.
44
§2.4
A new framework is developed over the next three chapters to
address these limitations and answer the research questions listed
below, while the final chapter presents a discussion of the work.
Chapter 3
What is the global scale of energy flow, from fuel to final services?
How much energy flows through the technical components in the
energy network?
How should conversion devices and passive systems be separated?
How can the results be presented visually in an accessible way?
Chapter 4
What are the theoretical efficiency limits in conversion devices?
In which devices are the largest efficiency gains likely to be found?
By what mechanisms is energy lost from devices?
Chapter 5
What is a passive system?
What are the practical efficiency limits in passive systems?
Which systems result in the greatest loss of useful energy?
Chapter 6
What contribution has been made to the field of energy efficiency?
What new conclusions can now be made?
Where are the opportunities for further research?
The resulting framework is global in scope, technical in focus,
absolute in measurement and visual in presentation, and provides
a rational basis for assessing all future developments in energy
efficiency.
45
3 Tracing the global flow of
energy from Fuel to Service
Claude Summers, in his 1971 paper entitled The conversion of en-
ergy, comments, ‘A modern industrial society can be viewed as
a complex machine for degrading high-quality energy into waste
heat while extracting the energy needed for creating an enormous
catalogue of goods and services’.82(p.41) The outputs of this com-
plex machine are the final energy services demanded by human
society: transport, thermal comfort, illumination and sustenance,
to name a few. The inputs to this machine are the primary energy
sources—fossil fuels, such as oil, gas and coal, renewable sources
and nuclear energy. So complex is the energy network in between,
that the numerous chains of conversion devices and energy systems
are yet to be mapped at the global scale.
Without a complete map of the global energy network, it is
difficult to attribute the carbon dioxide emissions from fossil fuel
combustion to final energy services. Without a map, the overall
efficiency of the energy network cannot be calculated, nor can valid
efficiency comparisons be made between the technical components
in the network. Therefore, the starting point for this chapter is
to construct a technical map of global energy flow, from fuels to
final services. This allows the energy devices and systems which
are likely to deliver the largest efficiency gains to be identified.
3.1 Potential gains from energy efficiency
Finding the global improvement potential from energy efficiency
measures necessitates tracing the scale of energy flow along the
numerous energy chains that form the energy network, and calcu-
46
§3.2
lating the efficiency limits for the individual technical components
in each energy chain. Equation 3.1 is used to find the available
energy savings for each energy conversion device or system:
Potential for
saving energy
=
Scale of
energy flow
×
[
Target
efficiency
− Current
efficiency
]
(3.1)
where the energy terms are measured in joules (J) and the effi-
ciency terms in percentages (%).
The key motivation for this research is to calculate the im-
provement potential using an absolute physical basis, which is in-
dependent of drivers in today’s market, and also correctly maps
the flow of energy through technical components. In particular,
this chapter addresses the first term of equation 3.1, the scale of
energy flow, by mapping the technical devices, systems and energy
chains which form the global energy network.
To understand the complete picture of global energy use it is
necessary to trace the complex chains of energy flow from fuels
through to final services. The focus throughout should remain on
the technical conversion devices and subsequent energy systems
in each chain. This extension of the energy flow-path has been
described qualitatively, yet to date no attempt has been made to
map the global flow of energy in physical units, from fuels to the
delivery of final energy services.
3.2 Drawing a map of global energy flow
The flow of energy from fuel to service includes the transformation
of energy sources into refined fuels and electricity, and the conver-
sion of the refined energy into final services. The first transforma-
tion, typically refining oil into petrol or burning coal to generate
electricity, is well understood. However, in delivering the final
service this refined energy is typically converted again by some
end-use device into a useful form (mainly heat or motion) which
47
§3.2
drives the activity of a technical system (a car, fridge or house) to
deliver the required service (passenger transport, sustenance, or
thermal comfort).
In order to clarify the different stages of conversion the term
passive system is introduced here for the first time, and refers to a
system to which useful energy (in the form of heat, motion, light,
cooling, or sound) is delivered. Passive systems are the last techni-
cal components in each energy chain, and in contrast to conversion
devices, do not convert energy into another useful form, hence the
descriptor ‘passive’. Instead, useful energy is ‘lost’ from passive
systems as low-grade heat, in exchange for the provision of final
energy services. Examples of passive systems include a car (ex-
cluding the engine) which delivers transport, or a house (without
the boiler or lighting device) which provides thermal comfort and
illumination.
Defining the boundary between the conversion device and the
passive system is not always simple. For example, it could be as-
sumed that the filament in a light bulb is the conversion device and
the surrounding glass bulb is the passive energy system. However,
the light (and unwanted heat) delivered into the bulb envelope is
not yet in a usable form and must pass through the glass bulb and
into the illuminated space before it can be considered useful en-
ergy. Therefore, the entire light bulb is defined as the conversion
device, and the illuminated space as the passive system. Similarly,
in a refrigerator, the rotational energy from the electric motor is
of no practical use until it is converted in cooling. Therefore, the
complete refrigeration system is defined as the conversion device
and the insulated cold-box as the passive system.
The novel distinction between conversion devices and passive
systems is shown schematically in figure 3.1. The flow of energy
can be traced from energy sources (left) to final services (top-
right) through three key conversion stages: fuel transformation;
electricity generation; and end-use conversion. At each conversion
stage the energy is upgraded into a more usable form, resulting in
48
§3.2
significant energy ‘losses’ (as low-grade heat with little practical
use).
The challenge in constructing a map of global energy flow is
to breakdown the generic energy flows in figure 3.1 into individ-
ual energy chains made up of technical components. For example,
the flow through ‘conversion devices’ needs to be divided accord-
ing to the different types of engines, furnaces and electrical de-
vices; ‘passive systems’ should be broken down by various types
of vehicles, industrial systems and building spaces. The aim is to
select a manageable number of similar sized categories (approxi-
mately ten) which cover the entire energy flow, for each step in the
flow-path. It is through mapping the connections between these
technical categories in Summer’s ‘complex machine’,82 that poten-
tial opportunities for improving energy efficiency can be identified.
The remainder of this section describes the process of allocating
the global energy supply to conversion devices, passive systems
and final services.
Figure 3.1 The flow-path of energy
49
§3.2
3.2.1 Energy sources
Energy enters society from fossil fuel reserves, biomass matter,
uranium deposits and renewable sources. However, Lightfoot 113
explains that the scales used to measure energy supplies differ be-
tween international data sources. The main differences arise from
the way energy is calculated for electricity generated from renew-
able and nuclear energy, and the varied groupings for ‘combustible
renewables and waste’. In an attempt to avoid unnecessary errors
in this analysis, Lightfoot’s recommendation to use one data source
with an absolute basis for measuring energy is followed.
Energy supply data is taken from the 2005 Balance Table for
the World, available from the International Energy Agency (IEA),114
and divided into the energy source categories listed in table 3.1
(renewable energy is technically not a ‘fuel’ but included here for
completeness). This source also provides the basis for allocat-
ing energy supply between direct fuel uses and electricity gener-
ation. The IEA category of non-energy—which consists of non-
combusted chemical feed-stocks (e.g. nitrogen fertilisers and plas-
tic products) and raw materials used directly for their physical
properties (e.g. lubricants, bitumen, carbon black)—is omitted
from this analysis as it has only a small effect on overall carbon
emissions. Direct carbon emissions associated with fossil fuel en-
ergy supply for 2005 are taken from the IEA Key World Energy
Statistics.115(p.44)
Fossil fuel energy data is typically published in joules ( J) based
on the standard enthalpy of combustion. These energy values are
converted into exergy values (also in J) which provide a measure
of the maximum work which can be extracted from the fuel. Using
exergy provides a more equitable basis for comparing fossil fuels
with uranium supplies or electricity, and for comparing heat with
motion or light, because all forms of energy are measured by the
same scale, their ability to perform work. In practice, using exergy
as a measure increases marginally the fossil fuel energy values (4
50
§3.2
Table 3.1 Energy sources and transfer mediums
Type Description
Energy source
Oil Crude oil and petroleum products
Biomass Combustible plant/animal products and municipal/
industrial waste
Gas Natural gas and gas works
Coal Hard coal, lignite and derived fuels (e.g. coke, blast
furnace gas)
Nuclear Heat equivalent of electricity (at 33% efficiency)
Renewable Electricity/heat from hydro, geothermal, solar, wind,
tide, and wave energy
to 11% across the sources, from Ertesvag and Mielnik 48(p.959) to
account for the additional energy content of the post-combustion
water vapour (lower heating value) and the flue-gas components.
3.2.2 Conversion devices
The grouping of conversion devices includes both upstream devices
(fuel refineries and electricity generation facilities) and end-use de-
vices (engines, furnaces and light bulbs). The IEA 2005 Balance
Table for the World114 gives conversion efficiencies for fuel trans-
formation and electricity generation. These have been transformed
into equivalent exergy efficiencies. Most energy studies, including
those of the IEA proceed to allocate the energy in refined fuels
and electricity (secondary flows) to broad commercial sectors such
as transport, industry and buildings. Yet, technical advances in
energy efficiency are not found in these sectors, but instead are
found in examining conversion devices such as engines, motors,
burners and light bulbs.
In contrast, for this analysis, secondary energy flows have been
allocated to the list of end-use conversion devices in table 3.2.
These devices are chosen to be technically distinct and of signif-
icant scale. The allocation of energy to each conversion device is
51
§3.2
based on the study Regional and global energy and energy efficien-
cies by Nakicenovic et al..111(tab. 3.3) Minor corrections are made to
match these fractions to the chosen device categories, and to re-
flect some structural changes which have occurred since the study
was published. For example, for the allocation to transport fuels,
the recent trend to switch from petrol to diesel powered cars is cor-
rected using 2005 world refinery production data from IEA.115(p.20)
3.2.3 Passive systems
The listing of passive systems in table 3.3 is novel. Each passive
system is chosen from within three broad categories—vehicles, fac-
tories and building—to be technically discrete but also of sufficient
scale in terms of energy flow. It is within these systems that useful
energy in the form of motion, heat, light, cooling and sound, is lost
as low-grade heat, in exchange for final energy services.
In previous studies, industrial facilities involved in manufactur-
ing materials and goods have been treated as final energy services.
For example, in Goldemberg 89(p.76) ‘steel making’ sits alongside
‘illumination’ and ‘food storage’ in the final row of energy ser-
vices. However, humans desire the structural properties of steel
rather than steel itself, and could in many cases be equally satis-
fied using an alternative such as aluminium. Thus, a distinction is
required between the material, steel or aluminium, and the final
service, structure. In this study, the energy delivered to facto-
ries has been divided into eight material production groups as
described in table 3.4. The allocation is based upon the 2005 in-
dustrial energy data from IEA 1(pp.476–7) and the conversion device
breakdown from USDOE 90(pp.13–16) after accounting for upstream
generation and fuel losses.
3.2.4 Final services
The key consideration when creating a list of final services is to
select a small number of distinct but comparable categories, for
52
§3.2
Table 3.2 End-use conversion devices
Conversion
device
Description
Motion
Diesel engine Compression ignition diesel engine: truck, car,
ship, train, generator
Petrol engine Spark ignition otto engine: car, generator,
garden machinery (incl. two-stroke)
Aircraft engine Turbofan, turboprop engine
Other engine Steam or natural gas powered engine
Electric motor AC/DC induction motor (excl. refrigeration)
Heat
Oil burner Oil combustion device: boiler, petrochemical
cracker, chemical reactor
Biomass burner Wood/biomass combustion device: open fire,
stove, boiler
Gas burner Gas combustion device: open fire, stove, boiler,
chemical reactor
Coal burner Coal combustion device: open-fire, stove, boiler,
blast furnace, chemical reactor
Electric heater Electric resistance heater, electric arc furnace
Heat exchanger Direct heat application: district heat, heat from
CHP
Other
Cooler Refrigeration, air con.: industry, commercial,
residential
Light device Lighting: tungsten, fluorescent, halogen
Electronic Computers, televisions, portable devices
which physical data is available or can be inferred. Eight final
energy categories are chosen for this study as listed in table 3.5.
The physical values for final energy services are estimated using
two methods. Where possible, bottom-up calculations from lit-
53
§3.2
Table 3.3 Passive energy systems
Passive system Description
Vehicle
Car Light-duty vehicle: car, mini-van, SUV, pick-up
Truck Heavy duty vehicle: urban, long-haul, bus
Plane Aircraft: jet engine, propeller
Ship Ocean, lake and river craft: ship, barge, ferry
Train Rail vehicle: diesel, diesel-electric, electric, steam
Factory
Driven system Refrigerator, air compressor, conveyor, pump
Steam system Petrochemical cracker, reactor, cleaning facility
Furnace Blast furnace, electric arc furnace, smelter, oven
Building
Hot water
system
Fuel and electric immersion boilers
Heated/cooled
space
Residential/commercial indoor space
Appliance/
equipment
Refrigerator, cooker, washer, dryer, dishwasher,
electronic, mechanical
Illuminated
space
Residential/commercial, indoor/outdoor space
Table 3.4 Materials and products
Material Description
Steel Iron and steel production
Chemical Chemicals and petrochemicals (excl. non-energy)
Mineral Non-metallic minerals
Paper Paper, pulp and printing, and wood products
Food Food, beverages and tobacco
Machinery Machinery and transport equipment
Aluminium Aluminium and non-ferrous metals
Other Textile, leather, mining, construction, non-specified
54
§3.2
erature of the global final service in physical units are used. For
example, Gantz et al. 116 estimate the size of the digital universe
in 2007 (a measure of the throughput of digital information) to
be 281 exabytes (281×1018 bytes) and the IEA calculates that
133 petalumen-hours (480×1018 lm s) of light was consumed in
2005.117(p.33) For structural materials, global production in tonnes,
is combined with material ‘strength’ properties (yield strength for
steel, aluminium and plastic; compressive strength for concrete,
from Ashby 118(p.452)) to give an estimate of the total structural
strength of all materials.
Where bottom-up estimates are not available, published phys-
ical indicators (in energy use per final service output) are matched
with global energy use (accounting for the conversion efficiency as
required), to provide an estimate of the final service. For the provi-
sion of transport services, indicators are taken from IEA 119(p.427) in
MJ/tonne-km and MJ/person-km. A weighted average of trains,
trucks and ships is used for freight transport, and of cars and
planes for passenger transport. For thermal comfort, the specific
heat capacity of air (1.2 kJ/m3K) is used to infer the total volume
and temperature change of air as a result of heating and cool-
Table 3.5 Final services
Final service Description
Passenger transport Number of people transported by car and
plane
Freight transport Tonnes of goods transported by truck, train
and ship
Structure Materials used to provide structural support
Sustenance Preparation, storage and cooking of food
Hygiene Clothes washing/drying, hot water , appli-
ances
Thermal comfort Heating and cooling of air in buildings
Communication Digital and written communication
Illumination Provision of light
55
§3.3
ing. This departs from the thermal comfort indicators used in
literature, for example in Schipper et al.,12 which take the hous-
ing floor area multiplied by the average temperature difference
( MJ/m2 degree-day). However, the chosen indicator is more rep-
resentative of the actual quantity of heating and cooling achieved,
rather than a proxy based on available data in collected statis-
tics. The same approach is used for cooking and refrigeration of
food, and the provision of hot water, using 3.0 kJ/kgK for food
and 4.2 kJ/kgK for water. The remaining energy use in buildings
provides mainly rotational work in many different devices. Rather
than divide these further, they are left under the hygiene service
category and measured in Newton metres ( N m) of mechanical
work.
In the absence of a global breakdown in literature, the allo-
cation of materials to final services is based on regional product
end-use data from: EUROFER 120 for steel; IEA 121 for chemi-
cals; BCA 122 for minerals; FAOSTAT 123 for paper; and IAI 124
for aluminium. For example, the IAI 124 divide aluminium prod-
ucts (by final energy use) into five applications: engineering ca-
bles (18%), packaging (13%), building (25%), transport (28%) and
other (16%). Based on this breakdown, energy use has been re-
allocated to the final services as follows: engineering cables and
building are assumed to be part of the ‘structure’ service; packag-
ing is allocated to ‘sustenance’; transport is split evenly between
‘freight transport’ and ‘passenger transport’ services; and other is
divided evenly between ‘structure’ and ‘communication’. A similar
allocation procedure is performed for all the material categories.
3.3 Results and discussion: what do we now know?
The energy data is presented in Sankey diagram form, in figure 3.2.
The global flow of energy is traced along each individual energy
chain from left to right, through four technical grouping: energy
sources, conversion devices (including fuel transformation, elec-
56
§3.3
trical generation and end-use devices), passive systems (including
materials) and final services. The thickness of each line represents
the scale of energy flow, with colour used to distinguish differ-
ent types of flow, and the vertical lines indicating where energy is
reallocated into new categories. Energy values are reported in ex-
ajoules ( EJ = 1018 J) and direct carbon emissions associated with
the primary fossil fuels are shown in red circles in billion tonnes
of carbon dioxide ( Gt CO2 = 10
9 t CO2).
Having traced the flow of energy from fuel to services and iden-
tified the technical steps in each energy chain, what can we now
say about the energy use in society? How should the energy map
be interpreted and how does it help us identify the areas in which
efficiency technologies will deliver benefit? To answer these ques-
tions it is useful to view the energy map in two ways:
Vertical from which meaningful comparisons of the scale of en-
ergy flow through technical components can be made within each
of the four vertical slices
Horizontal for which alternative technical options for providing
final goods and services can be compared if each horizontal energy
chain is traced completely from fuel to final service
These two views are explored below, followed by a brief comment
on the uncertainty of the analysis.
3.3.1 A vertical perspective of the energy map
The problem of adding, rather than multiplying, potential effi-
ciency gains from sequential steps in the energy flow, has already
been discussed on page 33, using the example of the Pacala and
Soclow stabilisation wedges. This conflict also applies to abso-
lute energy flows in the four vertical slices of the Sankey diagram:
energy sources (including fossil fuels and electricity); conversion
devices; passive systems (including the manufacture of materials
and products); and final energy services. For example, more than
57
§3.3
F
ig
u
re
3.
2
T
ra
ci
n
g
th
e
gl
ob
al
fl
ow
of
en
er
gy
fr
om
fu
el
to
se
rv
ic
e
58
§3.3
a third of the world’s energy is used to generate electricity, a third
is converted into heat, and a third is used in factories to make
materials—but these three thirds do not add up to the whole, be-
cause they come from different vertical slices. Thus the absolute
energy flows and potential improvements in efficiency can only
be compared within each vertical slice, as shown in table 3.6. To
add together energy flows or efficiency gains from different vertical
groupings ignores the sequential flow of energy, and could poten-
tially lead to exceeding the total energy supply, or an efficiency
savings of greater than 100%.
Despite the current focus on low-carbon energy sources, ta-
ble 3.6 shows that fossil fuels still dominate the first vertical slice
of energy sources. Transportation is almost entirely powered by
crude oil, and the majority of electricity is generated by burn-
ing coal and natural gas. Low-carbon sources (nuclear, biomass,
and renewables) currently make up 20% of energy supply, and are
dominated by nuclear, hydropower and biomass. With the excep-
tion of nuclear power, it will be difficult to expand supply of any
renewable source to the scale of supply from fossil fuels. The re-
maining renewables—wind, solar, tide and geothermal—account
for less than 1% of energy supply, thus de-carbonising the energy
supply remains a difficult challenge when compared with alter-
native gains from energy efficiency. Efforts should be focused on
improving combustion processes (as over 90% of energy sources
are fuels which are combusted), and exploring technical options
for converting the chemical energy of fuels, directly to electricity,
heat or motion.
Conversion devices that produce heat and motion are shown
to be important in the second vertical slice. Efficiency gains are
more likely to be found in heaters, burners and engines, than in
lighting devices, electronics and aircraft engines, due to the scale
of energy flow through these devices. For instance, efforts aimed at
promoting compact fluorescent light bulbs and reducing electronic
standby losses are useful for raising public awareness of efficiency
59
§3.3
T
ab
le
3.
6
T
ec
h
n
ic
al
co
m
p
on
en
ts
ra
n
ke
d
b
y
th
e
sc
al
e
of
en
er
gy
u
se
E
n
er
gy
so
u
rc
e
E
J
C
o
n
ve
rs
io
n
d
ev
ic
e
E
J
P
as
si
ve
sy
st
em
E
J
F
in
al
se
rv
ic
e
E
J
O
il
15
2
D
ie
se
l
en
gi
n
e
58
A
p
p
li
an
ce
s/
eq
u
ip
m
en
t
88
T
h
er
m
al
co
m
fo
rt
90
C
oa
l
12
7
E
le
ct
ri
c
h
ea
te
r
58
H
ea
te
d
/c
o
ol
ed
sp
ac
e
86
S
u
st
en
an
ce
84
G
a
s
97
E
le
ct
ri
c
m
ot
or
55
F
u
rn
ac
e
67
S
tr
u
ct
u
re
68
B
io
m
as
s
54
B
io
m
a
ss
b
u
rn
er
49
D
ri
ve
n
sy
st
em
56
F
re
ig
h
t
tr
an
sp
or
t
64
N
u
cl
ea
r
30
G
a
s
b
u
rn
er
47
C
ar
40
P
as
se
n
ge
r
tr
an
sp
or
t
64
R
en
ew
ab
le
s
1
5
P
et
ro
l
en
gi
n
e
41
T
ru
ck
38
H
y
gi
en
e
56
C
o
o
le
r
33
S
te
am
sy
st
em
31
C
om
m
u
n
ic
at
io
n
29
C
o
al
b
u
rn
er
31
H
ot
w
at
er
sy
st
em
23
Il
lu
m
in
at
io
n
19
O
il
b
u
rn
er
28
Il
lu
m
in
at
ed
sp
ac
e
18
H
ea
t
ex
ch
an
ge
r
20
P
la
n
e
10
L
ig
h
t
d
ev
ic
e
18
S
h
ip
10
E
le
ct
ro
n
ic
16
T
ra
in
8
A
ir
cr
a
ft
en
gi
n
e
11
O
th
er
en
gi
n
e
10
D
ir
ec
t
fu
el
u
se
27
2
H
ea
t
23
3
B
u
il
d
in
gs
21
5
E
le
ct
ri
ci
ty
1
8
3
M
o
ti
on
17
5
F
ac
to
ry
15
4
H
ea
t
2
0
O
th
er
67
V
eh
ic
le
10
6
T
o
ta
l
4
75
T
o
ta
l
47
5
T
ot
al
47
5
T
ot
al
47
5
60
§3.3
issues, but will have little effect on global energy consumption.
Similarly, future improvements in aircraft engine efficiency will
lead to weight and cost benefits, but will have only a small im-
pact on global carbon emissions. Thus, if the scale of energy flow is
considered, devices such as light-bulbs, electronics and aircraft en-
gines can be given less emphasis in policy initiatives because they
cannot deliver the required large reductions in carbon emissions.
The challenge for passive systems is to design technologies that
make better use of energy, by preserving and recovering the heat in
buildings, the materials in products, and the momentum in vehi-
cles. For buildings, space heating and cooling is predictably at the
top of the priority list, with a significant fraction of energy used
to maintain a temperature difference between the building interior
and exterior. Reducing heat transfer through the building fabric,
by insulating and preventing air leaks, remains a priority especially
for existing building stock. However, the high ranking for energy
use in appliances and goods is surprising and requires further in-
vestigation because of the diverse nature and much shorter life of
products in this grouping. Almost one third of energy is attributed
to the production of materials and goods in industry. Options for
reducing energy use in material production have been surveyed by
Allwood et al. 125 including improving material efficiency through
substituting less energy intensive materials, light-weighting prod-
ucts and designing for reuse and recycling. Advances in vehicles,
such as reducing aerodynamic drag and friction losses, should be
applied to cars and trucks in preference to planes, ships and trains.
Improvements in the fourth vertical slice can only be made
by reducing the demand for final services, through behavioural
and lifestyle changes. Nevertheless, it is helpful to examine these
services because the entire energy network exists solely for their
provision. Passenger and freight transport, when added together,
dominate the final services. The provision of sustenance is the
single largest category, because modern methods of growing (with
fertiliser), distributing, preparing and cooking food are energy in-
61
§3.3
Figure 3.3 Delivering passenger transport using alternative energy chains
tensive. Thermal comfort ranks high on the list and can be tar-
geted by reversing the practice of using high quality fossil fuels
to supply low temperature heat. Significant savings are available
from the wider use of heat pump technology and improving the
insulation of buildings.
3.3.2 A horizontal view of the energy map
It is through the process of mapping the complex global energy
network and comparing the scale of energy flow within the four
vertical slices, that technical priorities for improving energy ef-
ficiency can be identified. However, energy use or potential ef-
ficiency gains cannot be aggregated between vertical groupings.
Instead, to make comparisons between alternative horizontal en-
ergy flows, the entire energy chain from fuel to service must be
considered. This concept of improving energy efficiency by se-
lecting alternative horizontal energy chains is illustrated using the
example of delivering passenger transport, in figure 3.3.
Swapping conversion devices and systems within their verti-
cal slices leads to alternative energy chains, and potential savings
in energy. For example, switching all petrol engines (∼12% ef-
ficiency) to diesel engines (∼20% efficiency) would save approxi-
mately 4 EJ worldwide. However, switching one component in an
energy chain will often force changes to the components upstream,
resulting in new component efficiencies at every step along the en-
ergy chain. For example, if a petrol driven car is replaced with an
electric driven train, the flow of energy through the motor drive
62
§3.3
and electricity generation must also be considered. Yet this simple
concept is often overlooked in comparative energy studies, where
fuel efficiency values for vehicles are based on the volume of fuel
( L/km), irrespective of the type of fuel (diesel or petrol) and the
upstream energy losses associated with the fuel choice. The spec-
ification of electrical vehicles, in kWh/km from the socket, which
ignores the upstream efficiency losses from electricity generation,
is potentially even more misleading.
Tracing each alternative chain back to primary energy (and
carbon emissions) enables meaningful comparisons to be made be-
tween the scale of energy use, the impact of associated carbon
emissions, and the overall efficiency of the energy chain. Reduc-
tions in energy use for passive systems are particularly attractive,
because any saving in energy is compounded in the upstream steps,
resulting in a larger overall energy reduction. These compound
savings can only be identified when passive systems are separated
from conversion devices.
3.3.3 Data accuracy
All energy data is at best a good estimate, being dependent on the
accurate completion of energy surveys and the time delay between
collection and analysis. Significant differences of opinion exist over
how to measure primary energy supply, according to Lightfoot,113
and energy institutions do not publish error analyses with their
data. Rigorous data for the allocation of energy to conversion de-
vices, passive systems and final services is more difficult to obtain
due to the lack of global studies. Therefore, in the absence of
any specific uncertainty analysis for IEA data, the energy values
reported in this analysis are rounded to the nearest EJ.
Despite these limitations, the accuracy of the global energy
map is sufficient for determining the scale of energy flow through
the energy network. Patterns of energy consumption are certain
to change in the future, driven by structural changes, energy effi-
63
§3.4
ciency improvements and human behaviour. However, in the long-
term, the actions taken by society to improve energy efficiency are
likely to dwarf any data inaccuracies in this study. It is important
to use the best available data to direct priorities now, rather than
wait for more accurate data in the future.
3.4 Conclusion
The energy map presented in figure 3.2 provides a framework for
assessing the global scale of opportunity for energy efficiency mea-
sures. The analysis makes four unique contributions to our under-
standing of energy efficiency by:
• tracing the global flow of energy from fuels to final services in
Sankey diagram form
• focusing on the technical steps, rather than economic sectors,
within each chain of energy
• clearly defining the distinction between conversion devices and
passive systems
• identifying the key areas where technical innovation is likely to
deliver the greatest efficiency gains
The next two chapters, calculate the technical potential for
energy efficiency gains in conversion devices (§4) and passive sys-
tems (§5). The target efficiencies for individual technical devices
are then overlaid back onto the global map of energy flow to pro-
vide an absolute physical measure of improvement potential.
64
4 Theoretical efficiency limits
in Conversion Devices
Using a theoretical basis to assess energy conversion devices pro-
vides an absolute framework for identifying and ranking efficiency
options. This requires comparing the current efficiency of con-
version devices with their theoretical minimum, while considering
the complex interactions between technical devices in the global
energy network. Inevitably, using a purely theoretical measure of
efficiency promotes an ideal which may not be practically achiev-
able, either economically or technically. However, such an ap-
proach provides a useful theoretical target to direct priorities and
and a absolute basis from which to measure progress.
This chapter attempts to answer three key questions:
• how can the efficiencies of energy conversion devices be compared
on an equivalent basis?
• in which conversion devices are the greatest efficiency gains likely
to be found?
• how does categorising the avoidable losses according to energy loss
mechanisms help understanding?
4.1 Constructing a map of global energy efficiency
The section constructs a visual map of global energy efficiency,
which allows options to be identified and compared according to
an absolute basis, independent of benchmarks based on economic
or technical limitation. Three components are required to create
65
§4.1
such a map. The first is to determine the global scale of energy
flow through conversion devices, which is provided in chapter 3.
The second, requires determining the theoretical efficiency limit
for each type of conversion device, and superimposing these onto
the device energy flows. Finally, it is important to present the
results in a visually accessible format—such as a Sankey diagram—
permitting the maximum savings from efficiency measures to be
visualised.
4.1.1 Selecting a consistent measure of efficiency
To calculate the theoretical efficiency limit for each conversion de-
vice an appropriate measure of energy efficiency is required. Con-
ventional energy efficiency, which is based on the first-law of ther-
modynamics, is typically defined for a conversion device as:
η =
energy output (useful)
energy input
(4.1)
A natural gas power plant operating at 40% efficiency, an elec-
tric motor that is 95% efficient, and an air conditioner with a
Coefficient of Performance (COP) of 1.8, are all typical exam-
ples of reported first-law efficiencies. However, this measure of
efficiency is of limited use when comparing different types of con-
version devices because it is possible to have a maximum efficiency
greater than 100%, and the quality of energy is not considered. For
example, in space heating applications, a typical ‘high-efficiency’
gas burning furnace has a first-law conversion efficiency of 95%,
and an electric heating system is 100% efficient. Based on these
figures, it could be assumed that space-heating devices are already
approaching their maximum efficiency limits. However, a typical
heat pump has a COP of 3 (equivalent to an efficiency of 300%)
and under ideal conditions can approach 10 (or 1000%).
Such large variances in efficiency result from the failure of con-
ventional efficiency definitions to consider the quality of energy—
66
§4.1
electricity and mechanical work are more valuable energy carriers
than low temperature heat. Conventional energy efficiency (based
on the first law of thermodynamics) does not take into account
this difference in quality and hence is not an objective basis for
evaluating energy conversion devices.
In contrast, exergy efficiency (based on both the first and sec-
ond laws of thermodynamics, and similar in concept to effective-
ness or availability) provides a more equitable measure of conver-
sion efficiency. It uses mechanical work rather than energy as the
basis for comparing devices with each other and their thermody-
namic ideal. Exergy efficiency is defined for a device as:
 =
exergy output
exergy input
=
work output
maximum possible work output
(4.2)
By definition, the theoretical limit of exergy efficiency for an indi-
vidual device or a chain of multiple conversion devices, is always
unity.
Mechanical work is chosen because it is the highest quality,
lowest entropy form of energy. Electricity, which can be perfectly
converted into mechanical work, is another high quality form of
energy. Thus for a device which converts one form of mechanical
energy to another (e.g. gearbox), or electrical energy to mechani-
cal energy (e.g. electric motor), exergy efficiency and energy effi-
ciency are almost the same. However, when the input or output of
the device is heat (e.g. space-heater), the energy value of the heat
must be downgraded into equivalent units of mechanical work.
The importance of using an absolute measure of efficiency is
explained using an example of lighting devices. It is sometimes
argued that replacing incandescent light bulbs with more efficient
compact fluorescent bulbs saves little energy, because the build-
ings space heating requirements are offset by the bulb’s waste heat
production. Ignoring the fact that in many climates space heating
67
§4.1
is not required in summer, and that waste heat from the bulb may
compete with air-conditioning systems, the argument is flawed be-
cause it ignores the ‘quality’ of the energy. According to the first
law of thermodynamics, 100% of the electricity input to the bulb
is converted to either light or waste heat. Yet, from a second-
law perspective the electricity is high quality energy (it can be
converted into work almost completely), whereas the bulb’s waste
heat is a low quality form of energy (it is at low temperature, so is
difficult to convert to mechanical work). If a more efficient lighting
device was installed, the electricity saved could be used to run a
high efficiency device like a heat pump that could deliver 3 times
more of the same low quality heat than the light bulb (assuming
a typical COP of 3). Clearly, not all forms of energy are equal in
quality or usefulness, and therefore a consistent measure such as
exergy is needed to equate device efficiencies.
Exergy efficiency can be calculated directly, by finding the ra-
tio of the output to input exergy flows through the device, but
in practice this is complicated. Instead, if the conventional en-
ergy efficiency (η) is known, then the exergy efficiency () can be
estimated using:
 = η × ν (4.3)
where a dimensionless quality factor (ν) is used to correct for the
loss of energy quality in the conversion process, which results from
two sources. Firstly, the chemical exergy in a fuel is marginally
higher than the standard enthalpy of combustion due to the ad-
ditional contribution of the post-combustion water vapour (lower
heating value) and the flue-gas components. Ertesvag and Miel-
nik 48(p.959) give values called ‘exergy factors’ which vary by be-
tween 4 to 11% across typical fuel sources. Secondly, where energy
is converted into heat, the heat output must be downgraded to be
measured as mechanical work, using the thermal efficiency defined
68
§4.1
by a reversible Carnot engine (defined as |T−T0
T
|, where T is the
heat carrier and T0 is the ambient temperature, both in Kelvin).
4.1.2 Calculating efficiency limits in conversion devices
Creating a map of global energy efficiency requires assigning av-
erage efficiencies to each conversion device in the energy network,
including fuel transformation, different modes of electricity gener-
ation and end-use applications. It is important to select efficiency
values that are representative of the global device average, cal-
culated in a consistent way, and are from credible sources. The
input and output energy flows for the upstream conversions—fuel
transformation and electricity generation—are well defined in the
energy literature, allowing efficiencies to be deduced. However,
global energy flow data is not available for end-use conversion de-
vices, instead the efficiency values must be found by a survey of
literature.
The conversion efficiencies for fuel transformation and elec-
tricity generation are calculated from the 2005 Balance Ta-
ble for the World, produced by the International Energy Agency
(IEA).114 This table provides values for the global energy supply
broken down by fuel type, and for the ‘final’ energy delivered to
consumers in the form of refined fuels and electricity. Thus the
average energy and exergy efficiencies for fuel transformation, elec-
tricity generation and heat production, can be inferred from these
flows and other literature sources, as shown in table 4.1.
Some minor differences are found between efficiency for com-
bustion based electricity generation (ν<100%). The input to these
devices is increased when it is changed from energy to chemical
exergy, while the electricity output remains unchanged. Thus the
ratio of electricity output to chemical exergy input is reduced, by
the factor ν, and the exergy efficiency is lower. For CHP and
Utility heat plants, the difference between energy and exergy ef-
ficiency is larger because the heat output of these devices must
69
§4.1
be downgraded to mechanical work. In contrast, no difference is
found for fuel transformation (ν = 100%) because the inefficiency
relates to material losses during processing, nor is there differ-
ence for nuclear and renewable sources because the device input
remains unchanged.
Finding representative efficiency values for the global stock of
end-use conversion devices is difficult. Efficiencies cannot be
inferred from statistical studies of global energy flows, as this data
is not available for end-use devices. Instead published values for
Table 4.1 Energy and exergy efficiencies for conversion devices
Device Description η ν 
% % %
Electricity generation from:
Oil Crude oil and petroleum
products
37a 94 35
Biomass Combustible plant/animal
products and
municipal/industrial waste
25b 90 23
Gas Natural gas and gas works 40a 96 38
Coal Hard coal, lignite and derived
fuels (e.g. coke, blast furnace
gas)
34a 94 32
Nuclear Nuclear fission (heat
equivalent of electricity)
33c 100 33
Renewable Hydro, geothermal, solar,
wind, tide, and wave energy
80b 100 80
Fuel
transformation
In petroleum refineries, gas
works, coal preparation,
liquefaction, distribution and
own use
93d 100 93
CHP Combined heat and power
plants (all fuels)
56d 62 35
Heat Utility heat plants (all fuels) 85d 24 20
Notes: η=energy efficiency, ν=quality factor, =exergy efficiency
a IEA 126(p.73), b estimated, c IEA 127(p.138), d calculated from IEA 114
70
§4.1
energy and exergy efficiency must be used, but these vary consid-
erably depending on the technology and vintage of the equipment
surveyed, the chosen system boundary for each device and the
geographical scope of the study. Table 4.2 presents a review of
10 studies, covering the last 40 years, that list conversion device
efficiencies. The review indicates whether the values are first (en-
ergy) or second law (exergy or equivalent), the number of devices
categories given, and describes the scope of the study.
The study by Nakicenovic et al. 111 is easily the most compre-
hensive and consistent analysis of global energy and exergy effi-
ciency values. Unlike other studies, which use device case studies
to estimate best practice values, Nakicenovic et al. aggregate data
from 11 sub-regions and across 6 fuels types, to create average
global values of energy efficiency (η) and exergy-quality factors
(ν). The analysis also allocates global energy flows for 1990 to the
selected devices, which helps to verify the efficiency values. The
list of end-use conversion devices includes:
Residential/commercial sector cooking, washer/dishwasher, space
heating, hot tap water, space cooling, refrigeration, mechanical en-
ergy, lighting, Electronic Data Processing (EDP)/television, other
household appliances
Industry process heat (low and medium temperature), high tem-
perature heat/electrolysis, mechanical energy, other industrial uses
Transport bus/truck (Diesel), car/truck (Otto), airplanes, inter-
nal navigation (by water), rail, other
However, three adjustments are required to bring the Nakicenovic
et al. efficiency values into a form which is suitable for directing
technical priorities today.
Firstly, some individual device efficiencies are updated to re-
flect changes in system boundaries. The energy efficiencies for ‘me-
chanical energy’ devices (relating to electrical motors) are reported
as 70% in industry and 54% in residential. However, USDOE 90
71
§4.1
calculate an average efficiency of 45% for the entire motor driven
system, including the pump or compressor. This lower value more
Table 4.2 Survey of conversion device efficiencies
Reference and scope F
ir
st
-l
aw
S
ec
on
d
-l
aw
N
o
.
o
f
d
ev
ic
es
Summers 82(p.151)
Chart of best available technology, by the type of
energy conversion
X 25
Reistad 109(p.431)
Table of US data, including electricity generation at
38%
X X 23
Ford et al. 29(p.50)
Table, system boundary includes upstream electricity
generation
X 17
O’Callaghan 128(p.108)
Chart,system boundary includes upstream electricity
generation
X 38
Culp 129(p.33)
Chart of typical operational efficiencies
X 28
Gilli et al. 130(p.11)
Chart of end-use devices, with range of efficiencies
shown
X X 17
Nakicenovic et al. 111(p.228)
Global values for energy and exergy efficiency,
categorised by fuel
X X 20
Hammond and Stapleton 131(pp.152–157)
Charts and tables, by domestic, commercial,
industrial and transport applications
X X 11
USDOE 90(p.1)
Table of US power generation and industrial
equipment
X 14
Warr et al. 132(pp.34–35)
Charts of UK devices
X 10
72
§4.1
accurately describes the boundary system for a end-use conversion
device used in this research—the device output is measured in its
final useful form, in this case fluid motion. Ayres et al. 133(p.1117)
provides a breakdown of industrial electricity use in the United
States, which is used to separate refrigeration (6% of total net de-
mand) from the broader Nakicenovic et al. category of mechanical
energy. The efficiency for internal navigation (by water) is also
applied to transport by international marine vessels, and biofuel
powered engines are assumed to have an efficiency of 10%.
Secondly, the reported efficiency values represent 1990 technol-
ogy and are therefore outdated. Thus the efficiencies are scaled to
match historical improvements in global energy intensity using the
IEA reported sector improvements (cumulative from 1990–2004,
updated for 2005): buildings 13.3%, industry 22.7% and transport
8.2%.134 Applying a uniform efficiency improvement across the de-
vices in each sector might lead to device efficiencies greater than
100%. Instead the historical scale factors are applied uniformly to
the loss from each conversion device, grouped by economic sectors.
Thirdly, the devices presented by Nakicenovic et al. are grouped
by economic sectors instead of individual technologies. For exam-
ple, the electrical motor could be listed as a distinct technology,
but instead is included in four different categories: ‘mechanical
energy’ (both industry and residential/commercial), ‘other house-
hold appliances’ and ‘other industrial uses’. The diesel engine is
also hidden in these same four categories, and in four additional
transport categories: ‘bus/truck’, ‘internal navigation’, ‘rail’ and
‘other’. The selected device categories also vary considerably in
scale of energy flow, from as low as 0.1% of global energy demand
for ‘washer/dishwashers’ to greater than 16% for ‘space heating’.
It is preferable to organise the efficiency data into technically dis-
crete categories of conversion devices, of approximately equal scale
of energy flow.
Therefore, the efficiency values and quality factors reported
in Nakicenovic et al. 111 are adjusted and regrouped into the 14
73
§4.1
end-use conversion devices shown in table 4.3.
Table 4.3 Energy and exergy efficiencies of end-use conversion devices
End-use device Description η ν 
% % %
Motion 26 90 24
Diesel engine Compression ignition diesel
engine (truck, car, ship, train,
generator)
22 95 21
Petrol engine Spark ignition otto engine
(car, generator, machinery)
13 99 12
Aircraft engine Turbofan, turboprop engine 28 99 27
Other engine Steam or natural gas powered
engine
47 53 25
Electric motor AC/DC induction motor (excl.
refrigeration)
60 93 56
Heat 58 24 14
Oil burner Oil combustion device (boiler,
petrochemical cracker, reactor)
61 25 15
Biomass
burner
Biomass combustion device
(open fire/stove, boiler)
34 20 7
Gas burner Gas combustion device (open
fire/stove, boiler, reactor)
64 21 13
Coal burner Coal combustion device
(open-fire/stove, boiler, blast
furnace, reactor)
59 31 19
Electric heater Electric resistance heater,
electric arc furnace
80 30 24
Heat exchanger Direct heat application
(district heat, CHP)
87 15 24
Other 60 14 8
Cooler Refrigeration, air conditioning
(commercial, residential)
104 6 7
Light device Lighting (tungsten,
fluorescent, halogen, etc)
13 90 12
Electronic Computers, televisions,
portable devices
20 30 6
All devices 51 50 25
Notes: η = energy efficiency, ν = quality factor,  = exergy efficiency
74
§4.1
4.1.3 Grouping energy losses by engineering mechanisms
A perfect conversion device has no energy loss and therefore is
considered reversible (both the system and surroundings could be
returned to their original state). O’Callaghan 128 states that for a
process to be reversible it must be:
• adiabatic (no heat exchange between the system and surroundings)
• isothermal (system temperature is constant)
• fully resisted (no unrestrained expansion or throttling of gases or
liquids, and no ‘paddle-work’)
However, for any real process (and thus energy conversions)
there are always thermodynamic irreversibilities present. To pro-
vide further insight into how energy is lost, and therefore what
strategies could prevent energy loss, the losses from global conver-
sion processes are aggregated into ten engineering loss mechanisms
as described in table 4.4.
There is no single study which provides a breakdown of global
energy losses across the range of conversion devices considered. In-
stead a number of exergy analyses of individual conversion devices
are consulted: Dunbar and Lior 135 and Prins and Ptasinski 136 for
generic combustion processes (applicable to engines, heaters and
fossil fuel based electricity generation); Dunbar et al. 137 and Dur-
mayaz and Yavuz 138 for electricity generation using nuclear fission;
Ertesvag and Mielnik 48 for hydroelectricity; Rakopoulos and Giak-
oumis 139 for diesel engines; Ford et al. 29 for petrol engines; Turgut
et al. 140 for aircraft engines; Mecrow and Jack 141 and USDOE 90
for electric motor drives; Kotas 142 for refrigeration. The exergy
breakdowns do not always correlate directly with the conversion
device categories or efficiencies used in this study. In these cases
scale factors, interpolation and estimation were used to complete
the data.
75
§4.1
Table 4.4 Energy loss mechanisms
Combustion
Internal heat
exchange
Heat transfer between product molecules leaving
the reaction site (with kinetic and photon energy)
and neighbouring unreacted molecules, leads to
unrecoverable exergy loss. Internal heat exchange
can be avoided if the reactant and product streams
are separated.
Oxidation Chemical interactions (intra-molecular, radiation,
thermo-mechanical) result from the reaction of
oxygen and fuel, producing irreversible changes of
energy. Conversion of chemical energy to a useful
form without combustion, for example in fuel cells,
can prevent some of this loss.
Mixing Spontaneous mixing of reactants in the
pre-combustion stage, and products in the
post-combustion stage, cannot be reversed without
additional energy input. It is difficult to avoid
mixing in combustion processes.
Heat transfer
Heat
exchange
Heat transfer through a finite temperature produces
irreversibilities (e.g. from combustion gases to
steam). Minimising the temperature difference
reduces losses, but increases the heat exchanger
costs. Avoiding the use of high temperature fuel
combustion for low quality applications (space and
water heating), and cascading heat can reduce
losses.
Exhaust Thermal and chemical potential of stack and
tailpipe emissions. Extracting heat from water
vapour (condensing boilers) and completely
oxidising fuel can prevent some loss.
Heat loss Heat transfer from equipment to the environmental
reference state. Losses can be minimised using
insulation, preventing leaks of hot gas and liquids,
and ensuring reactants and products leave the
system at the surrounding temperature.
Continued. . .
76
§4.1
Table 4.4 Energy loss mechanisms (continued)
Other
Electrical
resistance
Resistivity (I2R), eddy currents and magnetic
hysteresis losses in devices (e.g. power distribution,
electric motor, light bulb, electronic). Can be
minimised by selecting superior materials/metals
for electrical components, and by reducing the
length of electrical wires through miniaturisation of
electronics and localisation of electricity supply.
Friction Friction (sliding and fluid flow), inelastic
deformation and unrestrained
compression/expansion leads to non-recoverable
energy losses (e.g. in motors, turbine, engine, pump
and pipe). Losses are reduced by using lubricants,
reducing fluid flow velocities, and resisting
expanding gases.
Fission
(nuclear)
Highly irreversible fission and heat transfer
processes result in losses. Can be partially reduced
by using fossil-fuel fired superheat and reheat units
in the downstream steam system.
Fuel losses Transformation, own-use, distribution and
transmission of primary fuels results in physical
losses (e.g. oil and gas leaks from pipelines). These
can be reduced with good design and maintenance,
or by using a more localised energy source.
4.1.4 Results
The global map of energy conversion efficiency is presented in fig-
ure 4.1. Energy flow is traced from primary energy sources (left),
through fuel transformation, electricity generation and end-use de-
vice conversion, to useful energy (top-right). The thickness of each
line represents the scale of energy flow, with the use of colour to
help distinguish different energy flows. Useful energy, in the form
of heat, motion, light, sound and cooling, is collected in the top-
right corner and indicates the energy required if the current con-
version devices were all to operate at their theoretical maximum
77
§4.1
efficiency. The vertical lines show where energy is converted in a
new form, with any loss of energy being separated from the main
flow and collated in the bottom-right corner. Energy values are
reported in exajoules ( EJ = 1018 J) and direct carbon emissions
associated with fossil fuels are shown in the red circles in billion
tonnes of carbon dioxide ( Gt CO2 = 10
9 t CO2) (based on 2005
data from the IEA Key World Energy Statistics 115(p.44)).
4.1.5 Data accuracy
Rigorous data for estimating conversion device efficiencies and al-
locating energy losses, is not readily available. Few global stud-
ies exist, therefore national and sector publications are used to
build a detailed picture of energy use. Energy allocation varies
considerably between countries and energy efficiency differs be-
tween devices depending on the age, operation and type of device.
Although some energy loss breakdowns are available for specific
devices, the methodology used differs between studies and it is
difficult to translate this data into a consistent global analysis.
For these reasons, and in the absence of any specific uncertainty
analysis for the collected energy data, the energy values reported
in this analysis are rounded to the nearest EJ.
However, one simplifying factor in this analysis is the alloca-
tion of energy use directly to the physical devices which convert
energy. There is no need to embed the energy associated with
upstream conversion processes such as electricity generation, or
non-direct energy inputs such as transport and capital equipment.
These energy inputs are allocated directly to the conversion device.
This avoids the complex boundary issues associated with other
energy analysis methods, such as Life Cycle Assessment, where
the allocation of non-direct impacts is subject to truncation and
double-counting errors, as discussed by Cullen and Allwood.86
Despite these known imperfections in data accuracy, the use
of best available energy data provides a much needed basis for
78
§4.1
F
ig
u
re
4.
1
T
h
e
gl
ob
al
m
ap
of
en
er
gy
co
n
ve
rs
io
n
effi
ci
en
cy
79
§4.2
prioritising action in the area of energy efficiency. It is anticipated
that over time new studies will provide more accurate efficiency
data for energy conversion devices, which can be used to build
further upon this research.
4.2 Discussion
Having mapped the theoretical efficiency limits for conversion de-
vices onto the global energy network, what can now be inferred
about the efficiency with which society uses energy? How do the
efficiencies of different conversion devices compare? How can we
interpret the map of energy efficiency, in order to direct priori-
ties for researchers, designers and engineers working in the field of
efficiency?
4.2.1 How efficient are current conversion devices?
Individual device efficiencies from different parts of the diagram
cannot be compared directly with each other. To state that an
electric motor is more efficient than a diesel engine, ignores the
larger upstream energy losses from electricity generation and dis-
tribution that are linked to the electric motor. Instead, a com-
pound efficiency (c) can be calculated for each energy chain, by
multiplying consecutive device efficiencies together along the en-
tire chain length:
c = f × e × d (4.4)
The subscripts used to indicate the type of conversion device
are taken from the map of energy flow shown in figure 3.1: c =
compound efficiency, f = fuel transformation; e = electricity gen-
eration and distribution; d = device conversion (end-use).
The resulting compound efficiencies for energy chains are shown
in table 4.5, organised by the end-use conversion devices. These
80
§4.2
Table 4.5 Comparing the efficiency of conversion devices
Energy chain Conversion efficiencies
f e d c
% % % %
Aircraft engine 93 100 27 25
Diesel engine 93 100 21 20
Other engine 92 78 25 18
Electric motor 93 32 56 17
Petrol engine 93 100 12 12
Motion average 93 77 24 17
Coal burner 90 100 19 17
Oil burner 93 100 15 14
Gas burner 91 100 13 12
Electric heater 93 32 24 7
Biomass burner 95 100 7 6
Heat exchanger 93 17 13 2
Heat average 93 76 14 10
Light device 93 34 12 4
Cooler 93 33 7 2
Electronic 93 32 6 2
Other average 93 33 8 2
Overall Average 93 70 18 11
Notes:  = exergy efficiency, with subscripts, f = fuel transformation; e
= electricity generation; d = end-use device conversion; c = compound
efficiency
indicate the theoretical efficiency limit for each chain, from fuel to
useful energy, irrespective of any particular combination of con-
version devices. The table shows that the conversion of fuels to
useful energy is typically inefficient, averaging only 11% across all
devices. The efficiency of conversion devices has improved only
marginally over the last 15 years, when compared with the 10%
calculated by Nakicenovic et al..111 This small absolute improve-
ment in average device efficiency places into sharp contrast the
reported and acclaimed 15% relative improvement in global en-
ergy efficiency between 1990 and 2005.126 Furthermore, the com-
pound efficiencies (c) for energy chains in 2005 range from 2–25%
81
§4.2
suggesting any device operating above an efficiency of 20% is con-
verting energy in an efficient manner.
Most of the inefficiency can be traced to the poor conversion
of energy in end-use conversion devices (d), which average only
18%. Looking specifically at this column, it can be seen that en-
gines, which deliver motion, typically operate with relatively high
efficiencies (12–27%) due to intense development motivated by eco-
nomic drivers to reduce the weight of both fuel and the engine in
transport vehicles. This is particularly the case for aircraft en-
gines where weight constraints have resulted in highly efficient de-
signs. Electric motors are even more efficient (56%), because the
upstream conversion losses from combustion are included in the
intermediate conversion step of electricity generation (e). In con-
trast, devices which combust fuels to provide heat operate at lower
device efficiencies (7–19%), with the variance depending primarily
on the temperature at which heat is delivered. This explains why
natural gas, a high quality fuel used in many low-grade applica-
tions such as space heating, is combusted at lower efficiencies than
coal, which has many higher temperature industrial applications
such as steel production. Cooling, lighting and electronic appli-
cations have low efficiencies (6–12%), and additional losses result
from the conversion of fuel to electricity, at an efficiency of 32%.
However, the efficiencies calculated in table 4.5 are not in them-
selves sufficient for ranking conversion devices. To be consistent,
the analysis needs to consider both the device efficiency limit and
the scale of energy flow. For example, it would be illogical to focus
efforts on improving the low efficiency of steam engines (included
under other engines), when this technology is no longer in com-
mon use. The resulting efficiency gains would not translate into
significant reductions in energy use or carbon emissions because
the application lacks scale.
82
§4.2
Table 4.6 Theoretical energy and carbon savings
Energy 1− c Energy Energy Carbon Carbon
Chain demand savings emissions savings
% EJ EJ Gt CO2 Gt CO2
Electric heater 93 58 54 3.4 3.1
Diesel engine 80 58 47 4.1 3.3
Electric motor 83 55 46 3.2 2.6
Biomass burner 94 49 45 0.0 0.0
Gas burner 88 47 41 2.6 2.3
Petrol engine 88 41 36 2.9 2.5
Cooler 98 33 33 1.9 1.9
Coal burner 83 31 26 2.7 2.2
Oil burner 86 28 24 1.9 1.7
Heat exchanger 98 20 20 1.2 1.2
Light device 96 18 17 1.0 1.0
Electronic 98 16 15 0.9 0.9
Other engine 82 10 8 0.7 0.6
Aircraft engine 75 11 8 0.7 0.5
Heat 90 233 210 11.7 10.4
Motion 83 175 145 11.6 9.6
Other 98 67 65 3.9 3.8
Total 89 475 420 27.2 23.8
Notes: c = compound exergy efficiency; potential for saving energy ≡
conversion losses
4.2.2 Theoretical energy and carbon savings
Theoretical energy savings can now be calculated for each complete
energy chain, from fuel to useful energy. Using equation 3.1, the
target efficiency is set to unity and the current efficiency equals
the compound efficiency for each chain (c), from table 4.5. The
corresponding savings in carbon emissions are calculated by equat-
ing the fossil fuel energy inputs with their direct carbon emissions,
and are reported in table 4.6. This allows alternative energy chains
to be compared and ranked, based on the potential for energy
savings, and for responses to be directed towards the conversion
devices with the greatest improvement potential.
Table 4.6 shows that 85% of conversion losses can be attributed
83
§4.2
to the provision of heat and motion (10.4 and 9.6 EJ respectively,
out of a total 23.8 EJ). The top half of the table is dominated
by heater, burners and engines, and efforts should be focused on
improving the efficiency of these devices. Lighting devices, elec-
tronics and aircraft engines together account for less than 10%
of global loss. Efforts aimed at promoting Compact Fluorescent
Light-bulb (CFL) and reducing electronic standby losses, present
easy gains due to their relatively low efficiencies and help raise pub-
lic awareness of efficiency concerns, but will not make a significant
impact on energy consumption. The conversion efficiency of air-
craft engines is already high (27%), suggesting that improvement
in engine efficiency will be difficult to achieve, and the available
energy savings at the global level are small. Thus few technical
options remain to improve the energy efficiency of flying, so a re-
duction in carbon emissions from this sector can only be achieved
by a reduction in the number of flights.
4.2.3 Understanding how energy is lost
The global map of energy conversion (figure 4.1) shows that only
a small fraction of the available energy supply is converted to use-
ful energy in conversion devices . This fraction represents the
theoretical minimum amount of energy that is required to pro-
vide the same amount of final service (assuming the downstream
passive system does not change). The remaining energy is ‘lost’
to the environment as non-recoverable energy in the form of low-
temperature heat, and is equal to the theoretical energy savings
calculated above. The breakdown of loss mechanisms presented
in this chapter is the first known attempt to collate and rank
global conversion losses by technical categories. Understanding
how energy is lost helps to direct research priorities and technical
innovation for engineers and technical designers. Therefore ta-
ble 4.7 shows how the energy lost from conversion devices, shown
in figure 4.1, is reallocated to ten loss mechanisms.
84
§4.2
Table 4.7 Categorisation of energy loss by mechanism
Loss mechanism Fuel Electricity Device Total
EJ conversion generation conversiona Loss
Internal heat exchangeb 0 25 51 76
Heat exchange 0 24 49 73
Exhaust 0 7 47 54
Electrical resistance 0 15 34 49
Heat loss 0 19 26 45
Oxidationb 0 15 29 44
Fuel loss 34 0 0 34
Friction 0 10 12 22
Fission 0 15 0 15
Mixingb 0 2 6 8
Heat transfer 0 50 122 172
Combustion 0 42 86 128
Otherc 34 40 46 120
Total 34 132 254 420
Notes: a in end-use devices; b in combustion processes; c includes fric-
tion, electrical, fission, fuel losses
Heat transfer processes are identified as the most significant
source of loss (at 172 EJ, more than 40%). This stems from the
irreversible nature of heat transfer across a finite temperature dif-
ference, and reflects the ill-considered use of high quality energy
sources (fossil fuels and electricity) for low temperature appli-
cations. Combustion processes are a significant source of losses
(128 EJ, 30%), especially from internal heat exchange when cold
reactants mix with hot combusted products. The majority of com-
bustion losses cannot be avoided without separating the reactant
and product streams, suggesting that long-term technical oppor-
tunities lie in devices which convert chemical energy directly to
electricity. Surprisingly, friction does not figure prominently in
the analysis indicating that the research activity in the fields of
lubrication and tribology, though important for preventing mate-
rial wear and hence reducing equipment costs, have limited scope
for reducing energy use.
85
§4.3
4.3 Conclusion
Developing more efficient energy conversion devices is essential if
efforts to reduce carbon emissions are to be successful. The global
map of energy efficiency presented in this chapter allows conversion
devices to be ranked according to their theoretical improvement
potential. The analysis makes three novel contributions to our
understanding of energy efficiency by:
• determining the average global conversion efficiency of devices
along each individual energy chain, and presenting this analysis
in a visually accessible format
• combining the scale of energy flow and the theoretical limits to
efficiency to identify key areas where technical innovation is likely
to deliver gains
• allocating, for the first time, global energy losses to engineering
loss mechanisms to direct priorities
Simple options for improving device efficiency include moving
the average global efficiency towards best practice and reducing
excess capacity from over-design. For example, the average effi-
ciency of energy use in light devices is 4%, still far below advanced
technologies such as CFL and light emitting diodes Light Emit-
ting Diodes (LED) with efficiencies above 20%. Similarly, elec-
tricity generation in advanced gas-turbine plants is approaching
efficiencies of 60%, yet the global average is nearer to half this
value. Many conversion devices are also over-designed for excess
capacity so operate well away from their optimal efficiency point.
This is the case with vehicle engines, which at normal cruising
conditions operate well below their optimum efficiency, because of
the requirement to have reserve power for acceleration. Designs
which avoid or smooth out these peaks in power demand, such as
hybrid power systems in vehicles, deliver much higher conversion
efficiencies.
86
§4.3
How much of the theoretical efficiency improvement could be
realised in practice? Beyond the simple gains described above, it
is necessary to consider the technical barriers preventing advances
in energy efficiency and look for alternative technology chains to
deliver useful energy. In practice, there are many technical fac-
tors that prevent designers from approaching theoretical efficiency
limits. For example, combustion processes, because they convert
fuel into heat, are constrained by Carnot’s Law and the adiabatic
flame temperature of the fuel. This means that the efficiency of
power generation is unlikely to rise much above 65% and current
efforts to improve efficiency—for example, increasing the heat ad-
dition temperature by using novel materials, preheating combus-
tion reactants, extracting mechanical work from turbines prior to
steam production—will give only incremental gains. To approach
the thermodynamic limit would require avoiding combustion al-
together, by converting the chemical energy in fuels directly into
electricity (and then motion) in devices such as fuel cells.
Nevertheless, the overriding lesson from this analysis is to begin
focusing research initiatives and directing efficiency policy towards
the technical devices in which the greatest gains can be found.
Only 11% of primary energy is converted into useful energy, thus
the theoretical gains available are substantial.
87
5 Practical efficiency limits
in Passive Systems
Efforts to improve energy efficiency can be divided into two techni-
cal approaches. The first is to improve the efficiency of energy con-
version devices which upgrade energy into more useful forms such
as motion, heat and light. For example, efficiency improvements
can be made to the internal combustion engine, which converts
the fuel’s chemical energy into the more useful form of vehicle mo-
tion. Finding the theoretical efficiency limits in conversion devices
is described in the previous chapter 4.
The second approach, which is less often considered, requres
making better use of the upgraded energy to deliver more final ser-
vice (such as transport, thermal comfort and illumination). Not
only can the efficiency of the engine be improved, but also the
aerodynamic design of the vehicle can be adapted to deliver more
transport service. However, for passive systems no conversion of
energy takes place and therefore a theoretical efficiency limit can-
not be calculated. For example, a perfectly insulated building
would require no heat input from a space heating device to main-
tain thermal comfort, giving an infinite efficiency limit which is
nonsensical. Instead, only a practical efficiency limit can be de-
fined for passive systems, measuring the minimum energy required
to deliver a unit of final service.
This chapter explores the practical energy savings available in
passive energy systems. A practical efficiency limit is found for
each type of passive system, and then multiplied by the global
scale of energy flow to determine where technical efforts should
be directed. This enables energy researchers and policy makers to
predict where the largest gains in passive systems are to be found.
88
§5.1
5.1 Methodology: assessing the practical energy savings
The solution proposed here is to assess the practical efficiency
limits, using physical models for each passive system. This permits
the energy losses from each system to be understood and realistic
estimates of energy savings to be made. The challenge is to collate
and simplify existing engineering models to obtain sensible values
for the efficiency gains achievable, and overlay these onto the global
energy flow through each passive system.
5.1.1 The utilisation ratio
To avoid confusion with the various definitions for efficiency, a new
term is defined for passive systems called the utilisation ratio:
UR = Utilisation Ratio =
Energy input
Final service output (physcial)
(5.1)
which is typically measured in MJ/km for transport and MJ/kg
of material for steel or aluminium. A lower value of the utilisa-
tion ratio indicates the energy is being used more effectively. The
practical energy savings available from a passive system can be
described by:
Potential for
saving energy
=
Scale of
energy flow
×
[
1− Practical UR limit
Current UR
]
(5.2)
The scale of energy flow is found by allocating global energy
supply in exajoules ( EJ = 1018 J) to 12 passive system categories
as shown in table 5.1. The term within the brackets gives a per-
centage measure of the practical savings available in each passive
system, up to a maximum limit of 100%. The ratio is dimen-
sionless as long as the terms are calculated using the same energy
input and final service output.
89
§5.1
5.1.2 More final services for less energy input
Passive energy systems are located at the end of each energy chain,
and are needed to transform useful energy—in the form of motion,
heat, cooling, light and sound—into final services. Although con-
version devices are necessary to upgrade energy into more useable
Table 5.1 Global energy supply allocated to passive systems
Passive system Description EJ
Building 215
Appliance/equipment Refrigerator, cooker, washer, dryer,
dishwasher, electronic devices,
mechanical systems
88
Heated/cooled space Residential/commercial indoor space 86
Hot water system Fuel and electric immersion boilers 23
Illuminated space Residential/commercial indoor
space, outdoor space
18
Factory 154
Furnace Blast furnace, electric arc furnace,
smelter, oven
67
Driven system Refrigerator, air compressor,
conveyor, pump
56
Steam system Petrochemical cracker, reaction
vessel, cleaning facility
31
Vehicle 106
Car Light-duty vehicle: car, mini-van,
SUV, pick-up
40
Truck Heavy duty vehicle: urban delivery,
long-haul, bus
38
Plane Aircraft: jet engine, propeller 10
Ship Ocean, lake and river craft: ship,
barge, ferry
10
Train Rail vehicle: diesel, diesel-electric,
electric, steam
8
Total 475
Notes: passive system categories and energy data from chapter 3
90
§5.1
forms, it is passive systems which deliver the final services that sat-
isfy human needs and desires. It is these services, a comfortable
thermal environment or the illumination of a work space which are
sought, not energy itself. The value of final services derives from
the loss of useful energy (high quality, low entropy) in their cre-
ation. They typically consist of some deviation of state from the
surrounding environmental reference, for example, in temperature,
pressure, composition or potential/kinetic energy.
The division of final services is shown in table 5.2. Eight dis-
tinct but comparable categories are chosen, for which physical data
is available or can be inferred from literature. Categories such as
‘recreation and leisure’ or ‘culture’ used in alternative studies, for
example Carbon Trust,62 are avoided because the method of allo-
cating energy to these non-physical services is somewhat arbitrary.
To improve passive systems it is necessary to ask: to what level
can the energy input to the passive system be reduced, while still
providing the same service? In this analysis, behavioural changes
which imply a degree of austerity or service loss are avoided, but
a careful examination of the final service is undertaken to identify
where energy savings might be found. For example, the service
provided in passenger transport is to move the mass of passengers
over some distance. Theoretically, this service can be achieved
using no energy, as the net energy difference between the initial
and final position is zero providing there is no change in gravita-
tional position. However, such an idealised case is not practically
achievable, as all entropy production would need to be eliminated,
requiring an infinitely long trip duration.
In contrast, current passenger transport in cars requires the
movement of an additional 2000 kg of vehicle mass as well as spend-
ing time idling in traffic jams, which are non-essential to the provi-
sion of the service. To find the practical minimum energy required
to deliver the service, which lies between today’s excessive use and
the idealised case, some additional judgement is required. This is
91
§5.1
achieved by modelling the passive system using scalar equations
and varying the equation coefficients within practical values.
The main body of this chapter involves building or adapting
physical models for each passive system to to estimate the utilisa-
tion ratios. Basic scalar engineering equations are used to describe
each system and show how energy is lost. By varying the equation
coefficients, within practical limits as defined in literature, the im-
provement potential in each passive system can be calculated. In
this way, the calculated practical energy savings are based on fun-
damental physical laws, and can be used with confidence to direct
engineering design choices. The analysis begins with buildings,
followed by factories and then vehicles.
Table 5.2 Current utilisation ratio for final services
Final service Description Value
Passenger transport Person-kilometres
travelled by car and
plane
23×1012 p km
Freight transport Tonne-kilometres of
goods by truck, train
and ship
46×1012 t km
Structure Materials used to
provide structural
support
28×109 MPa2/3 m3
Sustenance Preparation, storage
and cooking of food
28×1018 kg K
Hygiene Clothes
washing/drying, hot
water use
1.5×1012 kg K (wa-
ter)
Appliances 2.8×1018 N (work)
Thermal comfort Heating and cooling
of air in buildings
30×1015 m3 K
Communication Digital and written
communication
280×1018 byte
Illumination Provision of light 480×1012 lm s
92
§5.2
5.2 Practical energy savings in buildings
Energy is used in buildings to deliver many varied final services,
including warmth in winter and coolness in summer, hot and cold
water, light and entertainment, and the storage, preparation and
cooking of food. Or as Ford et al. 29(p.5) states, ‘buildings provide
an environment for the occupants’. The equivalent of 215 EJ of
primary energy is lost from the passive building system as low
temperature heat, in exchange for these services. Baumert et al. 7
calculate that buildings contribute almost 20% of energy related
carbon dioxide emissions, some 6.4 billion tonnes of CO2, with
approximately two thirds coming from residential buildings and
the remainder from commercial buildings.
In this section, physical models have been developed for the
passive systems in buildings, using a domestic house as a reference
for heated and cooled space, hot water systems and appliances,
and an office for illuminated space. Heated and cooled spaces
are examined first, followed by hot water systems and illuminated
spaces. The appliance category is examined last, despite using the
largest fraction of building energy, because of the complexity and
heterogeneous nature of this grouping.
5.2.1 Heated spaces in buildings
The provision of thermal comfort in buildings uses 86 EJ of pri-
mary energy worldwide. Heating the internal air space accounts
for 84% of this energy use, with the remainder used for cooling.
Thermal comfort could hypothetically be provided with much less
energy if only the occupants, instead of the building space, were
heated. People could wear extra layers of clothing when inside, as
they already do when outside in winter, or drink more cups of tea.
However, such measures involve major behavioural changes which
are outside the scope of this chapter. The practical compromise
is to heat the occupants and the building space around them, but
eliminate energy misuse such as the heating of spaces while the
93
§5.2
windows are open. The aim is to produce a well designed passive
building system which delivers thermal comfort with the minimum
practical energy use.
The following assessment of the practical energy savings avail-
able in heated spaces begins by building a model from scalar equa-
tions to describe the physical basis for heat flows in a typical house
and defining ‘thermal comfort’ for the occupants. Practical lim-
its are found for the coefficients in the model and the effects of
the geographic location of the house are considered. Finally, all
houses are assumed to be constructed at the practical limit for
reducing heat loss and the additional heat required to maintain
thermal comfort is calculated for each geographical climate zone.
Figure 5.1 shows the heat inputs and outputs for a typical heated
building space. The heat inputs and heat outputs can be balanced
and expressed as:
QH +QZ +QI = QS +QV +QT (5.3)
Figure 5.1 Balancing heat loads in a heated building space
94
§5.2
where Q are the heat transfer rates ( kW) with the subscripts for
the heat inputs: H = heater, Z = solar and I = internal, and for
the heat losses: S = shell, V = ventilation and T = thermal mass.
Scalar equations have been derived for each of the heat losses and
gains in equation 5.3.
Shell loss occurs whenever the interior space of a building is
warmer than the outside environment. Heat is conducted through
the building shell (walls, roof, floor, windows and doors), according
to the equation:
QS =
∑
i
(UiAi)(Tinside − Toutside) (5.4)
where U is the overall heat transfer coefficient ( W/m2K) of a given
building shell section (i.e. wall), A is the external surface area of
the section ( m2), and T is the temperature. The U-value is the
inverse of the thermal resistance through the shell section, and is
calculated for a typical wall by summing the inverse of the surface
resistances (R), and the thermal resistance of each material in the
wall:
Uwall =
1
Rinside
+
kplaster
wplaster
+
kinsulation
winsulation
+
kbrick
wbrick
+
1
Routside
(5.5)
where k is the thermal conductivity ( W/mK) of the material com-
ponent and w the component thickness ( m). Thus heat transfer
through the building shell can be described using only the physical
dimensions and material properties of the shell components, and
the inside-outside temperature difference.
Ventilation is required to prevent the build-up of carbon diox-
ide, toxic gases and odours in the interior building space. Fresh air
can be provided passively by opening windows or through leaks in
the building shell, or actively using a mechanical ventilation sys-
tem. The temperature of the incoming cold air must be raised to
95
§5.2
the inside temperature, and the required heat load can be calcu-
lated using:
QV = q ρCp(Tinside − Toutside)(1−R) (5.6)
where q is the air flowrate ( m3/s), ρ is the air density ( kg/m3), Cp
is the specific heat capacity of air ( kJ/kgK), and R is the fraction
of heat recovered in a heat exchanger (%).
Thermal mass measures the ability of the building to store en-
ergy or act as a buffer against temperature changes. It represents
a heat loss, when the average building temperature is increasing,
or a heat gain when the building temperature is falling. It is cal-
culated using:
QT = mCp
dT
dt
(5.7)
where m is the building system mass ( g), with a specific heat
capacity Cp ( J/gK) and subject to a change in temperature T with
time t. For a house, the aim is to maintain the indoor temperature
within a comfortable range. Although energy is required to heat
the thermal mass initially, once the desired temperature is reached
no additional heat is required. The temperature of the thermal
mass will fluctuate around some equilibrium and the heat absorbed
and released during each cycle will cancel. Designing buildings
with large internal thermal mass helps to even out temperature
fluctuations.
Solar heat gains result from the radiant heat of the sun passing
through the transparent components of the building shell (mainly
windows) and heating the building interior. The irradiation reach-
ing a surface at an angle (Hθ in Wh/m
2) is estimated for ge-
ographic locations, taking into account the average cloud cover,
hours of daylight and incident angle of the sun. The solar heat
96
§5.2
gain coefficient (CS) is used to estimate the proportion of the ra-
diant heat that passes through the windows (glazing area Aw in
m2) to the building interior. Thus:
QZ = AwCSHθ (5.8)
Interior heat gains come from the building occupants, each
producing approximately 90 W at rest, and the waste heat from
various conversion devices, including lights, refrigerators, cookers
and electronics. The heat input from the space heater (gas boiler,
stove, open fire or electrical heater) is characterised by the mass
flowrate and specific heat of the combustion fuel, or by the de-
mand for electricity. The heat input from the space heater, and
to some degree the solar heat gain, are varied in a typical house
to balance the heat loss and maintain the internal temperature at
a comfortable level.
Occupant comfort: The minimum inside temperature to main-
tain thermal comfort is defined using the European Standard for
indoor building environments EN15251 143 and the discussion pa-
pers on this standard by Nicol and Humphreys 144 and by Ole-
sen.145 The central equation used to define a comfortable occu-
pant temperature (Tcomf in
◦C) for buildings without mechanical
cooling, is:
Tcomf = 0.33Trm + 18.8 (5.9)
where Trm is the running mean of daily outside temperature, for
10 ◦C ≤ Trm ≤ 30 ◦C. This equation incorporates the latest re-
search in adaptive comfort theory which models how occupants
modify their behaviour in response to temperature changes. Nicol
and Humphreys 144 explain that in buildings without HVAC sys-
tems (heating, ventilating and air conditioning) the range of ac-
97
§5.2
ceptable temperatures is much greater, because people respond
‘adaptively’ to temperature change by making clothing changes or
opening windows. The occupants’ expectation of a tightly con-
trolled thermal environment is also relaxed.
For the occupant to feel comfortable, the occupant tempera-
ture (Toccupant) must lie within some variation from the comfort
temperature. The variance allowed on this comfort temperature
is given as ±3 K for a ‘normal’ expectation in new build and reno-
vations (used for this analysis), and ±4 K for a ‘moderate’ expec-
tation in existing buildings. However, the occupant temperature
is affected by both the air temperature and the average radiant
temperature of the surrounding surfaces. If indoor air speeds are
below 0.1 m/s and the air moisture content ignored, then the con-
tribution from the air and surfaces can be assumed equal. Thus
the occupant temperature can be expressed as the average of the
air and surface temperatures.
For an outside temperature of 10 ◦C and a normal expectation
of comfort (±3 K), the minimum acceptable inside temperature is
calculated as 19.1 ◦C using equation 5.9. This is a lower tempera-
ture than would normally be specified in a heated space, because
of the adaptive comfort assumptions. The aim of the practical
minimum house design is to trap solar and internal heat gains
within the building enclosure, and therefore bridge the difference
of 9.1 ◦C between the inside and outside temperatures. If the out-
door temperature falls below 10 ◦C then equation 5.9 no longer
applies, and the minimum indoor temperature is taken as 19.1 ◦C.
Equations 5.3–5.9 are used to estimate the practical energy
savings available in heated spaces. This involves three steps: (1)
defining the physical parameters for the model house which oper-
ates at the practical design limit; (2) categorising building loca-
tions by climatic zone to quantify the effects of solar heat gains
and outdoor temperature; (3) performing a heat balance by cli-
matic zone to determine the additional heat requirements for the
model house.
98
§5.2
Model house: A typical two-storey detached house is used as
a basis for modelling the heated space. The house measures 6.5 m
wide by 7.0 m long with the overall height of both floors measuring
6.0 m, and has 6.0 m2 of vertical windows facing the sun. The sam-
ple house is occupied by 3 people, each producing approximately
90 kW at rest. The selected practical limits for the equation co-
efficients are based on the PassivHaus design,146 and are given in
table 5.3.
More than 7000 buildings across Europe have reached the Pas-
sivHaus standard, each with an annual space heating and cool-
ing load below 15 kWh/m2 (54 MJ/m2). Low energy requirements
are achieved by specifying high levels of thermal insulation, mak-
ing use of solar and internal gains, and providing excellent air-
tightness using mechanical ventilation systems with heat recov-
ery. Typical wall and roof constructions use more than 300 mm
of cavity insulation with minimal thermal bridging, and triple-
glazed windows are standard. For this analysis, the only adjust-
ment made to the PassivHaus standard is to reduce the design
U-value for the exterior shell components (roof, wall and floor).
The standard recommends a maximum U-value of 0.15 W/m2K
but in practice a U-value of 0.10 W/m2K can be achieved.
Table 5.3 Typical and practical limit constants for heated spaces
Description Symbol Typical Practical limit1 Units
U-values
Roof Uroof 0.4–2 0.10 W/m
2K
Walls Uwall 0.5–2 0.10 W/m
2K
Floor Ufloor 0.4–2 0.10 W/m
2K
Windows Uwindow 3–5 0.80 W/m
2K
Ventilation rate q 10–50 7 l/s/person
Solar heat gain CS 0.4–0.6 0.5
Heat recovery R 0 0.8
Notes: 1 based on Passivhaus design standard.146
99
§5.2
Table 5.4 Climatic data for temperate and cold geographical zones
City Oct Nov Dec Jan Feb Mar Year
Barcelona, Spain (temperate zone)
H90◦ kWh/m
2 3.72 3.21 3.17 3.26 3.33 3.68 3.19
Toutside
◦C 18.0 12.6 9.6 9.2 9.7 12.0 16.2
Cambridge, England (temperate zone)
H90◦ kWh/m
2 2.26 1.57 0.94 1.39 1.97 2.25 2.19
Toutside
◦C 11.7 7.5 4.9 4.7 5.4 6.8 10.7
Oslo, Norway (cold zone)
H90◦ kWh/m
2 1.57 0.88 0.58 0.64 1.46 2.15 2.03
Toutside
◦C 6.8 1.7 -2.4 -3.2 -2.6 -0.3 6.2
St Petersburg, Russia (cold zone)
H90◦ kWh/m
2 1.64 0.65 0.36 0.76 2.10 3.18 3.08
Toutside
◦C 6.1 -0.2 -4.4 -5.7 -6.2 -2.5 5.5
Notes: H90◦ = daily solar irradiation for a surface at 90
◦, Toutside = 24
hour average outside temperature
Climatic data: the solar conditions and outside temperatures
vary according to geographic location. Therefore, the world pop-
ulation is divided into four climatic zones—tropical (24%), desert
(17%), temperate (44%), cold (14%)—using country level data
from the CIESIN database.147 Solar irradiation levels (H90◦) and
outside temperatures (Toutside) are taken from the database man-
aged by PVGIS,148 for two representative cities in each of the colder
zones. The 24 hour average outside temperature is used in place
of the running mean temperature suggested in equation 5.9. Ole-
sen 145 comments that little separates these two measures, but that
on average the running mean gives slightly wider fluctuations in
temperature. However, it is assumed that the model house has suf-
ficient thermal mass to even out any daily fluctuations in temper-
ature above or below the average. Table 5.4 collates this climatic
data for each city in the winter months.
Heat balance: the model house parameters and climatic data
100
§5.2
Figure 5.2 Heat requirement in temperate and cold climatic zones
are used in equations 5.3–5.9 to assess the additional space heating
requirements. The comfort temperature variation in equation 5.9
is assumed to be ±3 K, a normal expectation for a new build or
renovation. The results are shown in figure 5.2, with a mean daily
heat requirement ( MJ) greater than zero, indicating that addi-
tional heat input is required.
The graph shows that for a model house located in a temperate
zone (for example, Barcelona or Cambridge) the solar and internal
heats gains are sufficient to maintain thermal comfort throughout
the winter months. In fact the windows may need to be opened
on occasion to prevent overheating. For a model house located in
St Petersburg the maximum daily heat from the space heater is
about 20 MJ. This equates to 1.6 GJ of additional energy over the
101
§5.2
year to maintain thermal comfort (the area between the curve and
zero), which is less than 5 kWh/m2 (18 MJ/m2) per year.
Currently 14% of the world’s population, some 850 million peo-
ple, live in the cold climatic zone147 for which additional heating
of the model house is needed. Angel et al. 149 reports the global
mean floor area per person as 18 m2 with a range from 4–69 m2,
which makes the model house floor area of 30 m2 per person con-
servative. Although this floor area does not include commercial
buildings it is assumed that when people are working in offices
they occupy an equivalent floor area, and that as the same time
their home space is unheated. Thus, if all houses were constructed
equal to the model house, the global requirement for space heating
equals:
QH = 850×106 × 18 MJ/m2 × 30 m2 = 0.5 EJ. (5.10)
which can be compared to the current global energy used for space
heating of 28 EJ (at a weighted first law conversion efficiency of
40% from primary energy to delivered heat). The conclusion is
that if all buildings were designed at their practical limit, the
energy required for space heating would be reduced by 98%.
5.2.2 Cooled spaces in buildings
The strategy in cooler climates is to minimise heat loss from build-
ings by insulating and controlling ventilation, so that solar and
internal heat gains are trapped to maintain a comfortable temper-
ature. In hot climates, this strategy is reversed. Solar heat gains
are minimised by providing shading—in the form of an extended
roof overhang, shutters and deciduous trees—and insulating the
building exterior to keep heat out. Ventilation rates are increased
by opening windows and doors to remove the now unwanted inter-
nal gains from devices and from the occupants. Thermal mass can
102
§5.2
Figure 5.3 Balancing heat and cooling loads in a cooled building space
be used to even out temperature fluctuations, preventing the build-
ing from heating up too quickly, and water features and plants can
provide additional evaporative cooling. Additional space cooling
is provided by air-conditioning equipment which currently is re-
sponsible for 14 EJ of primary energy use worldwide. An overview
of these heating and cooling flows is shown in figure 5.3.
Modelling the cooled space is much simpler than the heated
space. The maximum inside temperature that is comfortable for
the occupants is calculated as 31.7 ◦C using using equation 5.9.
The upper range limit for the outdoor temperature is taken as
Trm =30
◦C, and a normal expectation of variance from the com-
fort temperature (±3 K) is assumed. In a well designed building
solar heat gains are reduced to zero and internal gains can be
removed by increased ventilation. Therefore, it can be assumed
that the inside temperature is always less than the outside tem-
perature, and that the occupants will be comfortable when the
running mean outside temperature remains below 31.7 ◦C.
Table 5.5 shows the average outside temperatures for the two
representative cities in each of the warmer climatic zones, from
103
§5.2
PVGIS 148 and Hoare.150 The maximum recorded temperature is
27.7 ◦C, which is 4 ◦C less than the limit for occupant comfort.
This suggests that if all houses were designed at the practical
limit and the specification for thermal comfort was relaxed, air-
conditioning would never be required, even in the hottest climates.
Thus 100% of the energy used to cool building spaces could, in
practice, be saved. Once again, it is assumed that the 24 hour
average temperature and the running mean temperature are in-
terchangeable, and that the model house has sufficient thermal
mass to even out the daily fluctuations in temperature. Addi-
tional cooling provided by water features and plants has not been
considered.
5.2.3 Hot water systems
The provision of hot water in buildings accounts for 23 EJ of pri-
mary energy consumption. The IPCC 151 report lists six options
for reducing the energy used to heat water: (i) water efficient fix-
tures and appliances; (ii) more efficient and better insulated water
heaters; (iii) tankless ‘point-of-use’ water heaters; (iv) heat recov-
ery from waste water; (v) air-source heat pumps; (vi) solar water
heaters, and estimate the combined effect of all these measures to
approach 90% energy savings. However, the purpose of the passive
hot-water system is not to heat the water, but rather to store and
Table 5.5 Climatic data for tropical and desert zones
City Toutside (
◦C)
Tropical Oct Nov Dec Jan Feb Mar Year
Singapore 27.2 26.8 26.3 26.9 27.3 27.7 27.1
Accra, Ghana 26.0 26.9 27.1 27.4 27.6 27.6 26.3
Desert May Jun Jul Aug Sep Oct Year
Tunisi, Tunisia 18.3 22.6 25.5 26.3 23.9 19.5 17.7
Rabat, Morocco 17.3 20.1 22.3 22.6 22.1 19.4 17.6
Notes: Toutside = 24 hour average outside temperature
104
§5.2
distribute hot water to the point of use. Therefore several options
can be eliminated such as improving the combustion efficiency of
the water heater, switching to a solar or heat pump options, or
upgrading the appliances which use hot water. Demand reduction
measures, such as installing low-flow shower heads and faucets, are
not assessed as they are considered behavioural changes. Better
insulation is a valid option for systems which store hot water, but
greater gains are available by removing the cylinder altogether and
installing point-of-use water heaters. Therefore, only the practical
energy savings from tankless hot water systems and heat recovery
are assessed.
The heat balance for the hot water system can be modelled
using:
QH = mCp(Tout − Tin) +QL (5.11)
where the heat transfer rate QH ( kW) must be sufficient to heat
the water—where m is the mass flowrate of water ( kg/s), Cp is
the specific heat capacity of water ( J/gK), and Tout and Tin the
hot water outlet and cold water inlet temperatures ( ◦C)—and
to balance QL, the loss of heat during storage and distribution
of hot water from the cylinders, fixtures and distribution pipes.
There are two benefits from using a tankless point-of-use system
which dispenses with the storage and distribution of hot water:
QL can be reduced to zero and the outlet water temperature can
be reduced.
The heat loss from water cylinders can be calculated using
equation 5.4. For example, a typical 210 l cylinder at 65 ◦C and
with 2.5 cm thick insulation (U =1.1 W/m2K) loses about 110 W.
The same cylinder stores approximately 40 MJ of energy, relative
to an inside building temperature of 20 ◦C, giving a typical storage
time of 120 h. The rate of heat loss is relatively slow allowing
the water to be heated intermittently. Hot water in distribution
105
§5.2
pipes cools more quickly due to the higher surface area to volume
ratio, however the overall heat loss is less than for the cylinder
because of the small volume of hot water stored in the pipes. The
USDOE 152 website states that insulating pipes can raise the water
temperature by 1–2 ◦C at the point of use. Ford et al. 29 estimate
that about 30% of the energy used in a storage system is lost from
the cylinder, fittings and piping. For a point-of-use hot water
system losses can be ignored, leading to a practical energy saving
of 30%.
Hot water storage systems are typically required to maintain
a minimum temperature of 60 ◦C throughout the system to pre-
vent the growth of legionella bacteria. This means water heaters
typically have a set temperature of 65 ◦C. Yet, the dishwasher is
the only household application which requires a water tempera-
ture above 50 ◦C and is often fitted with a built-in heating coil to
boost temperature. Mixing hot water with cold water to obtain
the correct temperature makes little sense from a thermodynamic
viewpoint as entropy is increased. Fortunately, a temperature of
50 ◦C is sufficient to avoid bacteria growth if the water is used im-
mediately after heating. Reducing the hot water temperature Th
in equation 5.11 from 65 ◦C to 50 ◦C, saves a further 19% of the
water heater energy.
Further gains are available using a drain-water heat recovery
system, as shown in figure 5.4. In recovery systems, heat from
the waste water is captured and used to preheat the cold water,
reducing the heat requirements of the water heater. Storage of
the heat is needed for applications where hot water is used in
batches rather than continuously, for example, in bathtubs, sinks
and clothes washers. According to the USDOE 152 80–90% of the
energy used to heat the water is lost down the drain. The propor-
tion of energy recovered can be estimated using a simple counter-
current heat exchanger model.
If it is assumed that no heat is lost to the surrounding environ-
ment, then the heat balance for the the exchanger can be written
106
§5.2
Figure 5.4 On-demand hot water and heat recovery from waste water
as:
Qh = mhCp(Tin,h − Tout,h) = Qc = mcCp(Tin,c − Tout,c) (5.12)
where the subscripts h and c denote the hot and cold fluids. For
the heat recovery system, the mass flowrate (m) and the specific
heat capacity (Cp) are assumed equal. Under these conditions the
temperature drop in the hot water (4Th) must equal the temper-
ature rise in the cold water (4Tc).
The heat exchanger equation is defined as:
QX = fUA4TLM(= Qh = Qc) (5.13)
where f is a temperature correction factor based on the heat ex-
changer design, U is the overall heat transfer coefficient ( W/m2K),
A is the heat transfer area ( m2) and TLM is the log-mean tempera-
ture difference ( ◦C), measuring the average temperature difference
between the hot and cold fluids.
107
§5.2
The proposed heat recovery system is assumed to be purely
counter-flow, allowing f to be set to unity. Typical U -values for
water to water heat exchangers range from 800–1500 W/m2K, so
the upper limit is selected. The general aim is to minimise the
temperature difference TLM so that the outlet temperature of the
cold water (Tout,c), is as close as possible to the inlet temperature
of the hot water (Tin,h). Temperature approaches as low as 1
◦C
are possible with advanced heat exchanger designs, with fine chan-
nels giving very large heat transfer areas, however these are only
suitable for non-fouling applications.
A more conservative temperature difference of 10 ◦C is chosen
for the heat recovery system to avoid fouling from the hot waste
water, giving a heat transfer area of 3.1 m2. If a compact plate
heat exchanger is selected with a typical surface area density of
400 m2/m3,153 then the unit would fit in a cube of side length 0.2 m,
which is credible.
The reduced heat requirement of the point-of-use water heater,
with heat recovery installed, can now be calculated. Hot water is
supplied from the water heater at 50 ◦C. Average losses during
the use of the water (15%) reduce the drain-water temperature to
44 ◦C, which is used as the exchanger inlet temperature (Tin,h).
A temperature difference of 10 ◦C in the exchanger gives a cold
water outlet temperature (Tout,c) equal to 34
◦C. Thus the wa-
ter heater only needs to raise the temperature 16 ◦C, bringing the
overall practical savings to 80% for the passive hot water system,
as summarised in table 5.6. This agrees well with the range heat
recovery rates, 50–80%, given for hot water heat transfer applica-
tions in USDOE.90(p.126)
5.2.4 Illuminated space
Artificial lighting is responsible for 18 EJ of primary energy use,
with 99% of lighting devices connected to the electrical grid. The
potential for saving energy in lighting applications is large, ranging
108
§5.2
from 25 to 80% in case studies reported in the International Energy
Agency (IEA) ‘Light’s labour lost’ report,117 however most solu-
tions focus on improving the efficiency of the lighting device. Less
attention is given to improving the passive system surrounding the
light device, from which significant gains can also be achieved.
The key variable for the passive lighting system is luminous
flux, which measures the perceived light emitted from a light source,
in lumens ( lm). It is found by calculating the total light ‘power’
(or radiant flux) from a light source and adjusting for the sensi-
tivity of the human eye to different light wavelengths. The aim is
to minimise the lumens required (and therefore the energy) from
the light device, without compromising the illumination service.
Illuminance (E) measures the final service delivered by the
lighting system. It is the amount of light (or luminous flux) in-
cident on a plane per unit area, expressed in units of lux ( lx
= lm/m2). Minimum illuminance levels are specified in lighting
codes for different building spaces, ranging from 54 lx for hallways
and utility rooms to 430 lx for office lighting, in IEA.117 The human
eye is designed to cope with a wide range of lighting levels—the
illuminance for an overcast day is typically 50,000 lx, which is 100
times greater than a typical office space.
The practical energy savings for the passive lighting system can
Table 5.6 Practical energy savings in hot water systems
System configuration Tin Tout Savings
1
◦C ◦C %
Storage water heater 10 65
Point-of-use, no storage 10 65 30
Point-of-use, lower temp. 10 50 49
Point-of-use, heat recovery 34 50 80
Practical energy savings available 80%
Notes: 1 cumulative savings from reference case (storage water heater)
109
§5.2
be found by minimising the luminous flux from the light device
(LFdevice or source-lumens) while still maintaining a specified level
of illuminance, using the equation:
LFdevice =
E × A
UF × LLF (5.14)
where E is the average illuminance, UF is the light utilisation
factor, and A is the horizontal illuminated area which is assumed
to be constant for a given lighting application.
LLF is the light loss factor, which accounts for the degraded
performance of the lighting system with time under real condi-
tions, in comparison to laboratory tests. It can be divided into
non-recoverable components, such as the physical degradation of
the lamp, fixture and room reflectance over time, and recoverable
components, such as cleaning and maintenance. Although some
technical improvement in LLF is possible, for example by using
materials that do not ‘yellow’ or designing self cleaning systems,
the largest gains are found in the organisation of maintenance
practices. For this reason the light loss factor is also kept con-
stant in the model. This leaves three strategies for improving
the passive lighting system: avoiding over-capacity in light design,
focusing light on the task area (both which reduce the average
luminance level (E)) and increasing the utilisation factor (UF ).
Firstly, illuminance levels in real applications are frequently
specified much higher than the application requires. This re-
sults from an historical trend in lighting design to specify ever
brighter working spaces, and the need to provide extra capacity
to cover variable light distribution and future degradation of illu-
minance levels. IEA 117(tab 4.6) estimates the average illuminance
levels in commercial offices to be 775 lx for Organisation for Eco-
nomic Co-operation and Development (OECD) countries, whereas
the Illuminating Engineers Society of North America (IESNA) rec-
ommends a level of 430 lx. Thus commercial lighting applications
110
§5.2
Figure 5.5 Reduction in luminous flux from using task lighting
are currently 44% over-designed.
Secondly, a uniform level of light is not required over the entire
floor area, if light is instead focused where the task occurs. This
is illustrated using the simple model shown in figure 5.5, which is
based around the hypothetical example given in IEA.117(tab 4.6)
For the reference office it is assumed that a person occupies a
floor area of 9 m2, at a uniform illuminance of 430 lx. For an office
designed at the practical limit, only 1 m2 is illuminated at the task
level, with the light level in the surrounding area reduced by 80%.
Although the specified minimum uniformity factors (the variance
of illuminance levels across a space) in lighting codes range from
0.33 to 0.8, a factor of 0.2 is considered the practical limit for
occupant comfort. In comparison, Ford et al. 29 have calculated
that humans are comfortable with natural light levels that can
vary by more than 7 orders of magnitude. This redistribution of
light results in a practical saving of more than 70%.
Thirdly, modifications can be made to the luminaire (or light
housing) and the room to improve the utilisation factor (UF ).
Ford et al. 29 estimate that only half of the light from a typical
device reaches the horizontal surface, whereas the IEA 117 believe
111
§5.2
Table 5.7 Practical energy savings in passive lighting systems
Configuration E UF Efficacy1 LPD Savings2
Wm2 lm/W lm/m2 %
Reference3 775 0.3 50 15.5
Correct E limit 430 0.3 50 8.6 44
Task lighting 125 0.3 50 2.5 84
Improved UF 125 0.9 150 0.8 95
Practical energy savings available 95%
Notes: LPD = average lighting power density, E = illuminance lm/m2
1 Overall lighting system efficacy 2 cumulative savings from reference
office. 3 based on estimated data for commercial buildings in the OCED
2000, from IEA.117(tab 4.6)
the figure is much lower at 30%. The remaining light is trapped
in the luminaire or lost to the ceiling and walls where it is not
needed. The luminaire output ratio (LOR) accounts for losses in
the luminaire and ranges from 0.3 for uncleaned painted surfaces
to 0.96 for mirrored surfaces. Minimum LOR of 0.7 in commercial
lighting and 0.5 in residential lighting are recommended for new
installations, and the current average LOR is likely to be less than
0.5. Thus, a two-fold improvement in LOR is realistic if the most
advanced light housings are used.
The other contributions to the utilisation factor are the dimen-
sions of the room, reflectance of the ceilings and walls, position of
the light fixtures and height of the task surface. In the absence of
statistical data to quantify these contributions, an overall UF of
0.9 is assumed for the practical limit.
Table 5.7 presents the results of minimising the illuminance
levels (E) and the utilisation factor (UF ), giving an overall prac-
tical energy saving of 95%. The model is based on commercial
office lighting, which according to the IEA 117 accounts for 43% of
global energy use in lighting. Although the illuminance levels for
residential (31%), industrial (18%) and outdoor (8%) lighting are
112
§5.2
generally lower, this is counteracted by typically larger illuminated
floor areas and lower utilisation factors. Therefore, the percentage
gains for commercial lighting are applied across all lighting appli-
cations. The reference is based on commercial lighting in OCED
countries, which is likely to give conservative energy savings in
comparison to offices in the developing world.
Efficacy is the ratio of light produced to electrical power con-
sumed, in lumens per watt ( lm/W). It is specified here for the
overall lighting system, and therefore reflects the change in the
UF . However, the efficacy of the lighting conversion device and the
efficiency of the upstream electricity generation is left unchanged
in the model, so that only the energy savings from the passive
system are assessed. The IEA 117 states that daylighting (from the
sun) might already be offsetting 25% of artificial lighting needs
in commercial buildings and reported savings in daylighting case
studies range from 15–80% for offices. However, these potential
gains have not been included in the analysis.
5.2.5 Appliances
The appliance category uses a larger fraction of primary energy
than any other building category, some 88 EJ. This energy is con-
sumed in the delivery of three broad services: sustenance (55 EJ),
hygiene (17 EJ) and communication (16 EJ). Nevertheless, the ap-
pliance category is the last building category to be analysed for
three reasons. Firstly, the grouping covers a wide range of differ-
ent applications including cooking, washing, cooling, drying and
processing information. Secondly, there is more variety of techni-
cal designs within each application, in comparison to other passive
systems. For example, food can be cooked using a wood fuelled
open fire, a gas oven, or a microwave. Thirdly, energy data for ap-
pliances in literature is not easily divided between the conversion
device and the passive system. This makes it time-consuming to
develop a single model for each appliance.
113
§5.2
Table 5.8 Practical energy savings in appliance systems
Appliance Elec Fuel Primary energy Savings1
EJ EJ EJ % %
Cooker 8 272 35 40 80
Refrigerator/freezer 15 0 15 17 88
Consumer electronic 11 0 11 13 0
Washing machine 4 0 4 4 91
Dishwasher 2 0 2 2 91
Clothes dryer 2 0 2 2 65
Other 11 8 19 22 59
Total 53 35 88 100 67
Practical energy savings available 67%
Notes: Data from Nakicenovic et al. 111 and Ja¨ger-Waldau.154
1Percentage energy savings available in each appliance. 2Includes 18 EJ
of biomass
Table 5.8 shows a breakdown of global energy used in appli-
ances and the practical energy savings available in each type of
appliance. To follow is a description of how the practical savings
are calculated for each appliance type.
Cookers: The global energy required for cooking food (35 EJ) is
comparable to the energy used in the transportation of goods by
truck. Cooking food changes its flavour, texture, appearance, and
nutritional properties, and according to Warwick and Doig 155 is
required to make 95% of staple foods edible. More than half the
global energy used for cooking comes from biomass sources (such
as wood, dung, crop), another quarter from electricity, one fifth
from gas, and the remaining few percent from oil and coal. Signif-
icant energy savings are available from improving the combustion
efficiency and the heat transfer between the heat source and the
vessel, but are not considered here as they relate to the conversion
device. Only the passive system, the container or vessel in which
the food is cooked is analysed.
114
§5.2
Food is typically heated by one or some combination of the
following heat transfer mechanisms:
Conduction where the food is in direct contact with the vessel,
for example when frying
Convection where heat is transferred to the food by the move-
ment of the fluid (air, water or oil) surrounding the food, for ex-
ample when boiling
Radiation involving the direct transfer of heat or microwave radi-
ation to the food, for example when grilling or using the microwave
Heat loss occurs through the shell of the vessel, when the heating
fluid (air or steam) leaks from the vessel and when the heat stored
in the thermal mass of the system is not recovered. Thus the heat
required (Q in kW) to cook the food can be written as:
Q = QS +QV +QT (5.15)
QS is the heat loss through the shell of the vessel, QV results from
the venting of cooking fluid and QT is the heat required to change
the temperature (T ) of the thermal mass of the food, cooking fluid
and vessel. Normally the thermal mass term would act as a buffer
to changes in temperature, absorbing and releasing heat. However,
for the model it is considered a loss, because the absorbed heat
is typically lost after the cooking is finished, as the cooking vessel
cools down and the fluid escapes or is discarded.
Three models have been created to estimate the practical sav-
ings available in passive cooking systems: a pot on a stove top, a
conventional oven and a microwave oven. The breakdown of heat
losses, into the shell, ventilation and thermal mass components is
inferred from literature or estimated, due to the lack of complete
energy studies of cooking in literature.
115
§5.2
Figure 5.6 Heat balance for cooking vessel (pot)
Table 5.9 Enthalpy change for heating and boiling a litre of water
Stage Temperature Enthalpy change
◦C kJ/kg %
Heating 20 to 100 335 13
Evaporation (latent heat) 100 2,270 87
Total (water to steam) 20 to 100 2,605 100
Notes: Enthalpies at standard conditions. Density of water equals
1000 kg/m3. Specific heat capacity of water equals 4.187 kJ/kgK.
The stove-top model describes cooking on a gas, electric or
biomass stove (or hob) using a container such as a pan or pot
as shown in figure 5.6. Cooking food in a pot of simmering wa-
ter allows heat to be transferred quickly to the food due to high
heat transfer rates (conduction from metal container to water, and
water to food) and the convection currents present when heating
water. The drawback is that significant heat is required to heat
the water to boiling temperature and then evaporate the water
during boiling, as shown in table 5.9. Therefore, preventing the
loss of water vapour (QV ) from the cooking vessel is a priority.
Brundrett and Poultney 156 have shown that when boiling wa-
116
§5.2
ter without a lid, 80% of the heat from the stove is used for evap-
orating the water. The simple action of using a lid, reduces the
steam losses by a factor of 100 (practically zero), and eliminates
the heat loss due to escaping steam (QV ). The energy used to
evaporate the water is recovered when the vapour re-condenses
on the walls and lid of the pot. In the absence of empirical data
about the use of lids when cooking, it is assumed that pot lids are
used for only half the time when cooking. Thus in the stove top
model, two-thirds of the heat is used for ventilation. Other stove
top cooking methods such as grilling or frying are not analysed
but are likely to be even less efficient due to the heat loss to the
surrounding air.
The remaining heat input (one-third) is required to overcome
the shell and thermal mass heat losses. Heat loss from the shell is
calculated using equation 5.4, with the pot surface area equal to
0.05 m2, and a temperature difference of 80 ◦C. The heat transfer
is limited by the interface between the bare metal exterior surface
and surrounding air, with a typical surface resistance value (R) of
40 W/m2K in natural air currents, resulting in a heat flux of 160 W.
The food, perhaps a staple such as rice or potato, is assumed to be
cooked in 1 L of simmering water for 15 minutes, giving an overall
shell heat loss of 144 kJ. This is about 30% of the one off heat
input of 335 kJ to raise the water to boiling temperature. Thus
the estimated heat loss fractions for the average stove top pot are:
67% ventilation, 23% thermal mass and 10% shell.
Three changes are made to the stove-top model to assess the
practical limit. Firstly, it is assumed that the pot is sealed with a
lid, eliminating all ventilation heat loss. Secondly, the walls and
lid of the pot are insulated with 30 mm thick fibreglass with a
thermal conductivity (k) of 0.03 W/mK. This lowers the average
shell U-value from 40 to 1 W/m2K, reducing the shell heat losses
98%. Thirdly, the thermal mass of the system is reduced by 50%,
based on an assumption that half of the thermal mass is retained
in the cooked food, primarily as absorbed water. This can be
117
§5.2
practically achieved by using heat resistant materials instead of
metal for the inside of the pot and reducing the quantity of water
for boiling.
The oven model is based on a typical 50 l electric or gas oven
operating at 200 ◦C with a cooking time of 60 minutes. Food takes
longer to cook in an oven than in boiling water due to the lower
heat transfer rates through air. Goorskey et al. 157 report that only
6% of the heat input to a oven is absorbed by the food. Current
oven designs have a double-skin cavity, filled with loosely packed
fibreglass insulation to prevent the oven exterior from reaching
high temperatures. The shell heat loss is calculated to be 150 W,
assuming 50 mm of fibreglass insulation (k = 0.05 W/mK) which
gives an overall U-value of 1 W/m2K. Heat is lost when hot air is
vented or escapes from the oven. The ventilation heat loss is found
using equation 5.6 assuming the air inside the oven is replaced
every 10 minutes during the cooking time. This gives a ventilation
heat loss of 15 W, a factor of 10 less than the shell heat loss.
Yet, these heat losses are small in comparison to the loss from
thermal mass. This is because the typical oven contains 15–20 kg
of steel, which requires approximately 1.5 MJ of energy to heat to
200 ◦C. This energy, averaged over the 60 minute cooking time,
gives a heat loss of 415 W. Thus the estimated heat loss fractions
for the current oven are: 72% thermal mass, 26% shell and 2%
ventilation, which differ remarkably from the stove-top model.
To assess the practical limit for the oven model three modifica-
tions are made. Firstly, the oven is sealed throughout the cooking
time, reducing the ventilation heat loss to zero. Secondly, the
oven fibreglass insulation is packed tightly (k = 0.04 W/mK) and
increased in thickness from 50 to 100 mm, reducing overall heat
transfer coefficient (U) to 0.4 W/m2K and preventing 60% of the
shell heat loss. In theory, there is no limit to the thickness of the
insulation and negligible shell heat losses can be attained. A in-
sulation thickness of 1 m would result in a 95% reduction in shell
heat loss, but would be difficult to use in a normal kitchen. There-
118
§5.2
fore, for oven model (and also the stove top model) the volume of
insulation has been limited to 2.5 times the volume of the cooking
vessel. Thirdly, the thermal mass of the oven is reduced by remov-
ing the steel, leaving only a thin non-conducting barrier to protect
the thermal insulation. The thermal mass heat loss is reduced to
6%, which is the energy absorbed by the food.
Microwave ovens differ significantly, because the microwave
radiation heats the food directly and not the air surrounding the
food. Heat loss through the oven shell and from ventilation oc-
curs indirectly as a result of the cooking food—these losses are
assumed negligible in the model. LBNL 158 have calculated the
overall efficiency for a typical microwave oven to be 56%, with the
majority of energy loss occurring in the magnetron (the conver-
sion device). The practical energy savings in the microwave oven
(passive system) are negligible.
A summary of the practical energy savings for cooking is given
in table 5.10. The heat losses for each cooker type have been nor-
malised so that the sum of the loss components (shell, ventilation
and thermal mass) for the current model, add to equal one.
Refrigerators and freezers: Domestic refrigerators and freezers
consume 15 EJ of primary energy every year, for the purpose of
chilling or freezing of food and drinks. Food is kept cold for two
reasons: to keep the food fresh over an extended period of time,
and because it is preferable to eat some foods at a cold tempera-
ture.
To calculate the practical energy savings available in refrig-
erators requires an estimate of the current energy use per unit
volume for the average global refrigerator (the current utilisation
ratio). Rosenfeld 159 shows the average annual energy consump-
tion of new refrigerators in the US has fallen from 1,800 kWh in
1974 to 450 kWh in 2001, driven largely by changes in the federal
standards. The fall in energy consumption is approximately linear
during this period. The average size of a new refrigerator is also
119
§5.2
Table 5.10 Practical energy savings in passive cooking systems
System (fi) w k U QS QV QT Q
mm W/mK W/m2K % % % %
Stove top (76%)
Current 0 0 40 10 67 23 100
Practical 30 0.03 1 0 0 12 12
Practical energy savings available 88%
Oven (16%)
Current 50 0.05 1 26 2 72 100
Practical 100 0.04 0.4 11 0 6 17
Practical energy savings available 83%
Microwave (8%) Practical energy savings available 0%
Overall (100%)2 Practical energy savings available 80%
Notes: 1includes biomass stoves, 2weighted average, by distribution of
energy use (fi). QS= shell, QV = ventilation, QT= thermal mass, Q =
total heat loss from passive system
shown to have stabilised at 0.6 m3 since about 1980.
The average age of a refrigerator in Norway is calculated by
Strandbakken 160 to be 7.9 years old. This is rounded up to 10
years (a 1995 refrigerator as the energy data is collected for 2005)
to compensate for the older stock of refrigerators in other coun-
tries. The average energy consumption in 1995 for a new refrig-
erator, from the chart by Rosenfeld, is 700 kWh per year (equal
to an average use of 80 W). For an initial comparison, a similar
sized RF19 refrigerator from Sun Frost (who claim to manufac-
ture the most efficient refrigerators in the world) consumes just
120 kWh per year (14 W).161 This 80% improvement in efficiency
comes from modifications to both the conversion device and the
passive system, such as:
• the compressor is mounted at the top of the refrigerator, instead
of the bottom, which shifts the heat generated by the compressor
120
§5.2
away from the condenser coils, improving the cooling efficiency
• for refrigerator/freezer models, two independent compartments
with separate cooling systems and temperature controls are used,
allowing the refrigerator section to be cooled by a higher than
normal evaporator temperature
• the refrigerator box is insulated with 50–100 mm thick polyurethane
foam insulation
• defrosting is achieved without using heating coils
Models have been created for the refrigerator and freezer to
separate out just the efficiency contributions from the passive sys-
tem. These give even greater energy savings than the Solar Plus
refrigerator because there is no economic constraint placed on the
model. The models are based on the equation developed for the
cooking passive system (equation 5.15), which divides the heat
loss from the passive system into contributions from shell (QS),
ventilation (QV ) and thermal mass (QT ). The key difference is
that instead of adding heat to balance the system heat loss, the
refrigerator and freezer produce cooling to balance the system heat
gain. Improving the passive system requires eliminating or reduc-
ing these heat gains. A summary of the practical energy savings
in refrigerators and freezers is provided in table 5.11.
The cooling input to the refrigerator box (Q) is calculated as
34 W, using the Rosenfeld example and assuming a typical re-
frigeration conversion device efficiency of 50% and a smaller re-
frigerator volume of 0.5 m3. An equal sized freezer will typically
consume twice the energy of a refrigerator, thus the cooling input
to the freezer is calculated as 67 W. Overall refrigeration energy
use is divided evenly between refrigerators and freezers, assuming
chilled space accounts for twice the volume of freezer space.
Heat gains through the refrigerator shell have been calculated
using the standard expression for heat conduction (equation 5.4).
121
§5.2
Table 5.11 Practical energy savings in refrigerators and freezers
System (fi) w k U QS QV QT Q
mm W/mK W/m2K W W W W
Refrigerator (50%)
Current 15 0.05 2.9 29 1 4 34
Practical 200 0.03 0.15 2 0 4 6
Practical energy savings available 83%
Freezer (50%)
Current 20 0.05 2.2 56 1 10 67
Practical 200 0.03 0.15 4 0 11 5
Practical energy savings available 92%
Overall (100%)2 Practical energy savings available 88%
Notes: 1net cooling required for frozen goods defrosted in refrigerator
2weighted average, by distribution of energy use (fi). QS= shell, QV =
ventilation, QT= thermal mass, Q = total heat gain to passive system.
Both the refrigerator and the freezer are assumed to have a vol-
ume of 0.5 m3 giving a heat conduction surface area (A) of 0.63 m2
based on a cube shape. The refrigerator holds the chilled food
at a constant temperature of 4 ◦C, while the freezer maintains
the frozen food at -20 ◦C, with an average ambient temperature
of 20 ◦C. The surface area and temperature differences are kept
constant in the model. The overall heat transfer coefficient (U)
is dominated by the thermal resistance of the insulation mate-
rial. Values for the thermal conductivity (k), insulation thickness
(w) and corresponding U-value, for the current and practical limit
cases, are shown in table 5.11. The insulation thickness for the
practical limit is restricted to 200 ml to remain below a practical
maximum ratio of 2.5 for the insulation volume to cooled space
volume (assumed for the cooking vessel models). The percentage
reduction in shell heat gain in both models exceeds a factor of ten.
Ventilation heat gains resulting from hot air leaking into the
refrigerator box must be cooled down. However, they are relatively
122
§5.2
insignificant (less than 1% of the passive system loss) due to the
low volumetric heat capacity of air and because only a fraction of
the refrigerator volume (the air, but not the food) escapes when
the door is opened. To calculate QV , it is assumed that 2.5 m
3
of warm air infiltrates the refrigerator every day, equivalent to
opening the door 20 times when the refrigerator is three-quarters
full. In the practical model, ventilation heat gains are reduced
to zero. This can be achieved by compartmentalising the cooled
space, sealing each compartment, and using horizontal drawers
or a chest type refrigeration unit to prevent the cool air leaking
downwards when accessing food.
Cooling is required to lower the thermal mass of the food that is
deposited in the refrigerator or freezer. Table 5.12 gives the change
in enthalpy required to cool and freeze a litre of water, used as a
proxy for food. The thermal mass components in table 5.11 are
found by closing the heat balance, using equation 5.15. For the
refrigerator, the value of 4 W is the equivalent of chilling 6 litres
of water per day. The thermal mass component is higher in the
freezer model, 10 W, but due to the greater enthalpy change, is
equal to freezing 2 l of water. Both these values appear credible
in light of typical refrigerator and freezer use. The thermal mass
Table 5.12 Enthalpy change for cooling and freezing a litre of water
Stage Temperature Enthalpy change
◦C kJ/kg %
Cooling (refrigerator) 20 to 4 67 15
Cooling (freezer) 4 to 0 17 4
Freezing (latent heat) 0 333 73
Cooling (freezer) 0 to -20 41 9
Total (water to ice) 20 to -20 458 100
Notes: Enthalpies at standard conditions. Density of water equals
1000 kg/m3. Specific heat capacity of water equals 4.187 kJ/kgK, and
of ice equals 2.05 kJ/kgK.
123
§5.2
of the refrigerator and freezer is ignored because once cooled it
remains at constant temperature in operation.
The thermal mass of the food in the refrigerator remains un-
changed for the practical limit calculation. However, it is reduced
for the freezer model. Domestic freezers maintain food at approxi-
mately -20 ◦C, well below the freezing temperature, extending the
shelf life of food products to between 2 and 12 months. Only a
small fraction of food products are consumed at this cold temper-
ature, for example ice cream. The majority of foods are defrosted
and frequently cooked before consumption. Thus food enters the
freezer at room temperature and is consumed at or above room
temperature, suggesting there is potential to recover the energy
used for freezing the food. In the practical model it has been
assumed that all frozen food is defrosted in the refrigerator, re-
sulting in a net cooling requirement equal to refrigeration (instead
of freezing).
Washing machines, dishwashers and dryers: So called wet appli-
ances use approximately 8 EJ of primary energy to deliver clean
and dry clothes and crockery. For washing machines and dishwash-
ers approximately 85% of this energy is used for heating water,
according to Goorskey et al..162 Heated water is normally sup-
plied from the house hot-water system, however some dishwashers
have in-built booster heaters to deliver the higher temperatures
required for killing germs and softening fatty deposits.
Practical savings in energy use result from three strategies:
lowering the water temperature, recovering energy from the hot
waste water, and using less hot water. The wash temperature
for clothes can be reduced for most loads to 20 ◦C with the use
of ‘cold’ water detergents. This reduces the energy required for
heating the hot water by about 80%. For clothes washers and
dishwashers, which require hot water, similar practical savings can
be attained by recovering the heat from waste water, using the
model developed for hot water systems in section 5.2.3. Using
124
§5.2
a counter-flow heat exchanger, with a temperature gap of 10 ◦C
and a waste water temperature 60 ◦C, gives an equivalent 80%
potential savings in energy.
Traditional vertical axis washing machines require significantly
more water than horizontal axis machines, because the clothes
must be completely submerged during washing. Horizontal axis
washing machines also have faster spin speeds, reducing the en-
ergy required for drying. Goorskey et al. 162 state that advanced
washing machines and dishwashers can use 30–60% less water than
conventional models. Therefore, it has been assumed for the prac-
tical limit that water usage in washing machines and dishwashers
can be reduced by a further 50%.
The practical utilisation limit is found for washing machines
and dishwashers by multiplying together the hot water savings
(80%) and water usage savings (50%), and overlaying these onto
the fraction of energy used in these appliances for heating water.
Practical UR limit = (1− 0.8)× (1− 0.5)× 0.85% = 9% (5.16)
Therefore, the practical energy savings available in these appli-
ances equals 91%.
Clothes dryers work by passing warm air through the clothes as
they are rotated in a drum. The warm air evaporates and absorbs
moisture from the clothes before it is vented outdoors or passed
through a heat exchanger where the water vapour condenses. The
reason for condensing the water vapour is not to save energy, but
to avoid the need for a hole through the external wall of the build-
ing. It is more convenient, especially in apartments to discard the
extracted water down the drain. In both options the enthalpy in
the water vapour is lost.
Using a heat exchanger, the latent heat of vaporisation for
water (see table 5.9) could be recovered. Possible options for re-
125
§5.2
covery include: preheating the cool drying air, mechanically re-
compressing the vapour for drying, preheating water in a hot-water
system, or converting the energy to electricity using a heat pump.
Palandre and Clodic 163 report energy savings of 50% for a mechan-
ical vapour re-compression system with a temperature approach
of 20 ◦C, in comparison to conventional condensation drying. In
comparison, USDOE 90(p.126) give an upper limit of heat recovery
in steam systems of 60% and domestic condensing boilers are re-
ported as recovering up to 75% of the water vapour energy from
the boiler exhaust gases. Therefore, a practical energy saving of
60% is assumed for clothes dryers.
Consumer electronics: The use of consumer electronic devices,
such as televisions, DVD players, radios, computers, printers, mo-
bile phones and office equipment, accounts for 13% of the primary
energy demand for buildings. Some ambiguity exists over the ser-
vice they provide—some options include entertainment, data pro-
cessing, communication, and record archiving—and the separation
of the conversion device from the passive system is challenging. In
this analysis, electronics have been treated in much the same way
as the light bulb. The various flows of energy and conversion pro-
cesses are of little use until the point where the information is
displayed on the screen or projected audibly through speakers. It
is the output of the electronic device, as light or sound, that in-
terests the user and can be measured. The more subtle functions
of storing data are more difficult to quantify.
Consequently, the passive system is defined as the space into
which the information, as light or sound, is delivered. This is
akin to the illuminated space for artificial light. However, the
three strategies used for improving the utilisation of light—task
lighting, reducing over-design and improving the luminaire—are
difficult to apply to electronic devices. The light and sound from
consumer electronics is normally focused toward the user so task
lighting principles do not apply. Brightness and volume are easily
126
§5.3
adjusted by the user to the correct level and in battery operated
devices there is a natural driver to avoid over-design. Reflecting
the light off the surrounding surfaces does not make a computer
screen easy to read. Reflecting sound causes echoes which can
render speech unintelligible. Therefore, it is concluded that there
are no practical energy savings available in the passive systems of
consumer electronics.
Other appliances and equipment: This category contains many
diverse types of appliances and other equipment used in domes-
tic and commercial buildings. Most energy use is attributed to
mechanical energy systems, including: water pumps, ventilation
systems and generators used in commercial buildings; and motor-
mowers, chain saws and other small petrol driven engines used
in workshops and gardening. Without any specific breakdown of
appliances and equipment, it is impossible to attribute practical
energy savings within any accuracy. Instead practical energy sav-
ings of 59% are applied based on the analysis of industrial driven
systems in section 5.3.3.
Combining the practical energy savings from each appliance group
as a weighted average by energy use leads to an overall potential
reduction of 67% for appliances as shown previously in table 5.8
on page 114.
5.3 Practical energy saving in factories
The manufacture of materials and products in factories uses 154 EJ,
almost a third of the world’s primary energy. The energy used
to make products contributes to all the final service catergories:
structure (68.4), sustenance (28.0), hygiene (17.1), communication
(13.2), freight transport (11.2), passenger transport (10.6), ther-
mal comfort (4.3) and illumination (1.2). The modelling of prac-
127
§5.3
tical energy savings in industry is challenging due to the complex
and varied use of energy, the wide range of manufactured products,
and the embedded energy content of materials. This sections starts
with a discussion of the influence of embedded energy, before devel-
oping engineering models for furnaces, steam systems and driven
systems. The potential gains from material efficiency strategies—
reducing material scrap in processing (yield gain), extending the
service life of products, and recovering waste material—have not
been included in this assessment of practical limits.
5.3.1 Embedded energy
Materials and products differ from the services provided by vehi-
cles and buildings. Whereas light, heat and motion last only for a
brief time, the energy input to materials can remain embedded for
many years. This energy is carried with the product, as embedded
energy, and accounts for the chemical and physical changes made
to the product material during processing. Chemical changes to
the composition of the material account for most of the embedded
energy—materials processes such as mixing, separating, crushing
and deforming, require comparatively smaller quantities. For ex-
ample, the enthalpy of reaction (∆Hr) for the reduction of iron ore
(in the form of hematite Fe2O3) to iron (Fe) equals 7.4 MJ/kgFe.
This energy is embedded in the steel, and could theoretically be
recovered from the steel at a later stage. The enthalpy of reac-
tion is also, from an energy perspective, the theoretical minimum
energy required to make the steel.
Embedded energy is not to be confused with the term em-
bodied energy (as Ashby 164 explains) which includes all the energy
inputs to a product, including the ore, feed-stock, fuel and electric-
ity consumed in making a material. This remaining energy—the
energy input to the process less the enthalpy of reaction—is lost
as heat from processing equipment (for example, furnaces, pumps
and steam systems) during the manufacture of the product. It is
128
§5.3
this wasted energy which is targeted for reduction.
The difference in energy between the reactants and products
defines the theoretical energy requirement for a chemical reaction
(∆Hr). Endothermic reactions (∆Hr<0) require an energy input
and thus the theoretical energy requirement is positive, whereas
exothermic reactions (∆Hr>0) have a negative theoretical energy
requirement. This is important because some material production
processes are exothermic, and should release energy, yet in practice
they still consume process energy. In these cases, the potential
savings in energy may be greater than the fuel and electricity input
to the process.
Table 5.13 gives a breakdown of global energy use in indus-
try, with the five most energy intensive materials shown. Moving
across the columns from left to right, the total primary energy (T )
has been allocated to the three main passive systems in industry—
furnaces (F ), driven systems (D) and steam systems(S). The to-
tal primary energy has been divided by global material production
figures to give the energy intensity in units of MJ/kg This is con-
trasted with the theoretical energy requirements per unit mass,
taken from literature.
The theoretical minimum energy value for steel is a weighted
average of 60% primary production with ∆Hr=7.4 MJ/kgFe,
165
and 40% secondary production with no enthalpy change. It is
noted that for secondary production of steel from scrap no re-
action takes place, and theoretically the steel does not need to
be melted to be formed into a new product (the deformation en-
ergy is assumed negligible). Similarly, for aluminium the min-
imum energy is a weighted average of 50% primary production
with ∆Hr=31 MJ/kgAl,
166 and 50% secondary production with
no enthalpy change.
For chemicals, Neelis et al. 167 have calculated the average reac-
tion enthalpy change from a survey of 68 chemical processes, cov-
ering 63% of the primary energy use in the chemical industry. The
energy minimum is negative due to the many exothermic chem-
129
§5.3
Table 5.13 Energy use in industry
Industry Primary energy1 Global Energy Theoretical
T F D S prod. intensity minimum
EJ EJ EJ EJ Mt MJ/kg MJ/kg EJ
Steel2 36 26 8 2 1100 32 4 5
Chemical 24 10 5 10 500 47 -31 -15
Cement 13 10 3 0 2400 6 2 4
Paper 11 0 2 9 380 29 3 1
Aluminium3 6 5 1 1 40 155 16 1
Other 64 17 38 9
Top 5 90 50 19 21 4420 20 -4
Total 154 67 56 31
Notes: 1Primary energy: T = Total, F = Furnace, D = Driven system,
S = Steam system. Breakdown by industry and passive system from
USDOE 90 with non-combusted energy excluded. 2Steel production:
60% primary, 40% secondary. 3Aluminium production: 50% primary,
50% secondary.
ical reactions especially in the petrochemicals industry. Taylor
et al. 168 report the thermodynamic minimum energy requirement
for the calcification reaction to produce cement as 1.8 MJ/kg of
clinker.
Kinstrey and White 169 calculate a theoretical minimum for
pulping and paper making of 15.7 MJ/kg. However, this value in-
cludes contributions from items such as ‘powerhouse losses’ from
the cogeneration of electricity and the evaporation of the water in
paper-making—the value is not a true theoretical minimum as it
is constrained by today’s known technology. In contrast, de Beer
et al. 170 states the theoretical energy to make a flat sheet of pa-
per from pulp is almost negligible as the only energy needed is for
aligning and bonding the fibres. Therefore, only the minimum en-
ergy for pulping (lime kiln and liquor evaporation) from Kinstrey
and White 169 has been included in the table.
In the last column of table 5.13 the absolute theoretical mini-
130
§5.3
mum energy has been calculated for each of the top five materials,
representing nearly 60% of global industrial energy use. The total
of this column shows net embedded energy in materials of minus
4 EJ; the energy released from the exothermic reactions to produce
chemicals is greater than the energy input to the endothermic reac-
tions for metals, cement and paper. Assuming that this conclusion
holds for the remaining processes in the other category, then the
waste energy lost from factories is at least equal to, if not slightly
more than the total inputs of fuel and electricity. Therefore, in the
engineering models developed in the next section, the embedded
energy in materials is ignored.
5.3.2 Furnace
Furnaces account for 44% of industrial energy use worldwide and
are used to deliver medium and high temperature heat, either
directly or indirectly. The furnace category includes fired heat-
ing systems (including furnaces, dryers, calciners, reactors and
evaporators) which combust fuel, electrical furnaces, and electro-
chemical cells (where electricity instead of heat is used to drive a
chemical reaction). Engineering models of furnaces typically in-
clude the combustion of fuel in the conversion device, the losses
from the heated space and the distribution and transfer of heat.
For the passive system analysis only the heated space or vessel is
considered. Many of the principles governing heat loss from the
passive furnace system have already been developed in the previ-
ous section on buildings, for example the domestic oven.
According to the first law of thermodynamics, the heat deliv-
ered to the furnace system must be balanced by the sum of the
heat losses:
Q = QS +QV +QT (5.17)
131
§5.3
QS is the heat loss from the shell (walls and roof) of the fur-
nace and is calculated using the standard heat transfer equation
(see equation 5.4). The magnitude of heat loss depends primarily
on the thickness (w) and thermal conductivity (k) of the insu-
lation material, from equation 5.5. QV results from the leakage
of hot air primarily when the doors are opened to charge or dis-
charge material from the furnace. It depends upon the volume
and temperature of the lost air, according to equation 5.6, but is a
comparatively small heat loss due to the low heat capacity of air.
QT is the energy required to heat the furnace shell and raise
the temperature of the product material (the thermal mass of air
is ignored). Ashby 164 explains that the for small batch furnaces,
with frequent cycles of heating and cooling, significant energy is
used to heat the furnace shell to the operating temperature. The
heat input can be found using the average temperature of the shell:
QT,shell = mCp (
Tout − Tin
2
) (5.18)
where m is the mass of the shell and CP the specific heat capac-
ity. An optimum shell thickness can be found which minimises
heat loss, by trading off a thick well-insulated shell (QS) against a
thin shell of lower thermal mass (QT,shell). Ashby shows that this
optimum wall thickness is found when (kCpρ)
1
2 is minimised.
When a furnace is operated continuously, as is the case for most
larger furnaces, the heat absorbed by the shell can be ignored. In
this case, the heat input to the furnace equals the heat loss, with
no accumulation of heat within the system. Thus, heating the
thermal mass of the product material is the only contribution to
QT .
The prediction of current energy use in furnaces is based on a
study of two 250 t/h reheating furnaces in a Taiwanese hot strip
steel mill, performed by Chen et al..171 Each furnace is designed to
reheat steel slabs (1250 mm wide and gauge 250 mm) from 25 ◦C
132
§5.3
Table 5.14 Practical energy savings in the furnace
Furnace w k U QS QV QT Q
mm W/mK W/m2K % % % %
Current 100 1.0 10 25 31 43 100
Practical 250 0.5 2 4 6 28 38
Practical energy savings available 62%
Notes: QS= shell, QV = ventilation, QT= thermal mass, Q = total
heat loss from passive system
to 1250 ◦C prior to being rolled into steel coil. The furnace is oil
fired with a rectangular insulated shell of effective dimensions of
40 m long by 10 m wide by 0.25 m high, and a door at each end to
allow continuous operation.
Chen et al. have published a breakdown of the heat losses cov-
ering both the conversion device and passive system components
of the furnace.
• 17.7% for heat losses from the furnace shell (QS) and the opening
of the doors when slabs are charged and discharged (ignored)
• 42.3% as enthalpy in the discharged steel slab (QT )
• 31.4% in the exhausted flue gas (QV )
• 7.7% in the water used to cool the furnace and equipment (QS)
• 0.9% as enthalpy in the removed oxidation scale (QT )
This distribution has been used to find the heat loss terms
from equation 5.17 (as indicated), leading to the current energy
use breakdown shown in table 5.14. No specific details are pro-
vided about the insulation of the furnace shell, although refrac-
tory bricks typically have thermal conductivities ranging from
0.5 to 1.5 WmK at 800 ◦C.172 An average value of 1 WmK has been
used in the current model.
For the practical limit model, three modifications have been
made. Firstly, the insulation thickness (w) is assumed to have
133
§5.3
increased by 150 mm and the thermal conductivity (k) is lowered
from the average to the lowest value for Zircoa 172 refractory brick.
This results in an 83% reduction in shell heat loss. Secondly, heat
is recovered from the exhaust flue gas at an efficiency of 80%,
based on the energy recovery rates given in USDOE.90
Thirdly, some of the enthalpy contained in the hot discharged
slab (900 kJ/kg at 1250 ◦C) is recovered as steam. de Beer et al. 42
give an example of direct transfer to steam in a boiler where the
slab enters the boiler at 900 ◦C and exits at 300 ◦C. The boiler
produces 40 bar steam at 450 ◦C, recovering 320 kJ/kg in the steam
at a recovery rate of 36%.
This recovery efficiency is lower than for liquid or gas heat ex-
changers, because transferring heat from solids restricts the range
of available heat exchangers and the maximum working tempera-
ture of steam is much lower than the slab discharge temperature.
In the practical limit model a recovery rate of 36% is used, giving
an overall practical energy saving of 55%, which is applied across
all types of furnaces.
5.3.3 Driven systems
Motor driven systems consume 55 EJ of primary energy and ac-
count for approximately one third of global electricity use. The
category includes all industrial use of: pumps, fans, material pro-
cessing and handling equipment, compressors and refirgeration.
The majority of motor systems are powered by electrical motors
(90%), with natural gas being the second most used energy source
(5%).
A generic pumping model has been developed to calculate the
practical energy savings in motor driven systems. Although each
driven system is likely to have distinct characteristics which differ
from the pumping application, the lack of specific data on energy
losses in these systems has necessitated the use of a more general
model. This pumping model is described first, followed by a de-
134
§5.3
scription of the the specific adaptions or exceptions to match the
model to the other categories of driven systems.
Pumping applications can be divided into two categories de-
pending on the service they deliver. The first application aims
to transport a fluid (liquid, gas or slurry) over some distance or
change in height, through some distribution system (piping, duct-
ing or channels), to the point of use. The distribution of potable
water from a source to the residents in a city is a common exam-
ple. Material handling systems such as conveyors belts achieve the
same aim for solids. The purpose of the second application, is to
raise the pressure of the liquid to provide mechanical work—for
example, the propulsive force of a water jet or hydraulic pressure
in a piston—or to deliver the liquid into a high pressure vessel.
Compressed air and material processing systems (grinders, crush-
ers and mixers) operate in a similar fashion.
The liquid pressures in the pumping system can be calculated
using the mechanical energy equation, sometime referred to as the
extended Bernoulli equation:
Pin+
ρν2in
2
+ρgzin+ρWshaft = Pout+
ρν2out
2
+ρgzout+ρWloss (5.19)
where P is the static pressure (inlet and outlet), ρ is the fluid
density, ν is the fluid velocity, g is the acceleration due to gravity,
z is the elevation height, Wshaft is the net shaft energy per unit
mass of fluid, and Wloss is the loss due to friction. For pumping
systems, the fluid velocity terms can be cancelled out if a constant
volumetric flow rate and pipe diameter are assumed. The height
terms balance if no net change in height is required—this is a valid
assumption for most pumping applications.
For pumping liquids over a distance (transport applications)
the pressure terms cancel out as the fluid enters and exist at at-
mospheric pressure. The increase in pressure imparted by the
135
§5.3
pump is eventually lost due to friction in the piping and fittings,
and the all the shaft energy is lost as friction, Wshaft = Wloss. If
a higher pressure is required at the outlet then the required shaft
work is Wshaft = Wloss + Pout − Pin.
The termWloss includes contributions from friction losses in the
pump (suction, impeller and discharge), valves (throttle and con-
trol), piping, and fittings. Friction losses in the pump and throttle
valve are normally grouped together with the electric motor and
drive coupling losses, to calculate the conversion device efficiency.
The USDOE 90 estimates these losses to be 40% of the electrical
input to the electric motor. Passive system friction losses result
from the fluid flowing through the pipes and fittings. This de-
lineation between conversion device and passive system is logical,
given that the fluid is not in a useful form (at the correct pressure)
until after it has been conditioned by the throttling valve. This
separation is also consistent with the common practice of replac-
ing the throttling valve with a variable speed drive (VSD) on the
pump motor, to save energy in the conversion device.
The pressure loss in piping and fittings due to friction is cal-
culated using the DarcyWeisbach equation:
∆Ppiping = ρWloss = f
L
D
η2
2g
(5.20)
where f is the dimensionless Darcy friction factor, L is the effective
pipe length (which includes the equivalent length of piping for
fittings) and D is hydraulic diameter of the pipe (which equals the
diameter for circular pipe). Substituting the volumetric flow (q)
per unit cross-sectional wetted area for the fluid velocity η, gives:
∆Ppiping = f
L
D
1
2g
(
4q
piD2
)2
=
8fLq2
pigD5
(5.21)
136
§5.3
The variables f , L and D can be adjusted to reduce the piping
pressure loss (and pumping energy), but the greatest gains are
achieved by increasing the pipe diameter. For example, doubling
the pipe diameter will decrease the pressure by 1 − 1
2
5
or 97%.
However, this must be balanced against the additional piping cost
and greater floor area.
For the practical limit a 25% increase in the pipe diameter
has been assumed, giving a 67% reduction in pumping energy.
This agrees well with industrial case studies reported by Lovins 173
where the redesign of piping systems reduced the friction loss by a
between 67% and 83%. A similar example provided by the IEA 1
shows the piping efficiency increasing from 60% to 90%, equating
to a two-thirds reduction in piping losses. The calculated 67%
energy reduction from increasing the pipe diameter is applied to
the motor driven systems which transport materials: fan system,
material handling systems and pump systems (the fraction of en-
ergy used for transporting fluids is assumed to be 80% as no value
could be found in literature).
The practical energy saving in refrigeration systems (88%) have
been based on the domestic refrigerator and freezer model de-
scribed in section 5.2.5. The remaining motor driven systems are
used to provide direct mechanical energy: compressed air systems,
material processing systems and 20% of pumps. The practical lim-
its for these devices are sourced from the United States Depart-
ment of Energy (USDOE) report Energy use, loss and opportuni-
ties analysis,90 which provides a breakdown of the delivered usable
work versus the energy lost for these applications. It is assumed
that 50% of the energy lost in these mechanical processes can be
practically avoided. The resulting practical savings in energy use
are shown in table 5.15.
137
§5.3
Table 5.15 Energy use in driven systems
System Primary energy Savings
EJ % %
Pump 14 25 621
Material processing 12 22 45
Compressed air 9 16 40
Fan 8 14 67
Material handling 7 12 67
Refrigeration 4 7 88
Other 2 4 67
Total 56 100 59
Practical energy savings available 59%
Notes: 1weighted average of transportation and mechanical energy ap-
plications
5.3.4 Steam
Steam systems consume 31 EJ of primary energy every year, de-
livering heat at lower temperatures (150–550 ◦C) than furnaces.
Conceptually, steam systems differ from furnaces because the heat
is not transferred directly to the heated space. Instead, steam acts
as a heat transfer medium and is typically distributed from a cen-
tralised boiler, via a piping network, to the place of application.
The indirect heat transfer also means that further heat cannot be
recovered from the combustion flue gas.
To calculate the practical energy savings from the steam sys-
tem, the ventilation term QV in equation 5.17 has been replaced
by a distribution term QD. According to the USDOE
90 steam dis-
tribution accounts for 15% of the overall fuel input to the boiler,
which equates to 29% of the overall heat loss (after taking into
account the conversion efficiency of the boiler). Table 5.16 shows
the modified distribution of heat loss components for the current
energy use.
For the practical limit, the shell (QS) and thermal (QT ) heat
138
§5.4
Table 5.16 Practical energy savings in steam systems
Steam system QS QD QT Q
% % % %
Current 27 28 45 100
Practical 4 0 29 33
Practical energy savings available 66%
Notes: QS= shell, QD= distribution, QT= thermal mass, Q = total
heat loss from passive system
losses are reduced by the same percentage as for the furnace model.
Although, the absolute energy savings are much lower for the
steam system, because of the reduced operating temperature, it
is credible that the percentage reductions will be similar. The
distribution heat losses are completely eliminated using the same
arguement that was developed for point-of-use hot water heaters
in section 5.2.3. The resulting practical energy savings for the
steam systems are marginally higher than for the furnace, at 66%.
5.4 Practical energy savings in vehicles
Transporting people and goods in vehicles results in 106 EJ of pri-
mary energy use. This energy is ultimately lost from the vehicle
passive system as low temperature heat to the surrounding envi-
ronment, in exchange for the final service of transport. Half of the
energy is used to move goods (freight transport) and the other half
to move people (passenger transport). Improving the utilisation of
energy in the vehicle passive system requires reducing the resistive
forces where energy is lost.
Figure 5.7 shows the generic forces which act on a vehicle,
in this case a car. The same schematic can be used for trucks,
ships and airplanes, with some minor adjustments. The resistive
drag forces, and therefore efficiency, are highly dependent on ve-
locity, as illustrated in the seminal paper by Gabrielli and von
Ka´rma´n 174 using a graph of specific resistance (resistive force di-
139
§5.4
vided by weight) versus velocity. The Ka´rma´n-Gabrielli line was
plotted on this graph as practical efficiency limit for vehicles, irre-
spective of the transport mode. However, it has since been demon-
strated that large ships and freight trains have passed through this
barrier, as reported by Yong et al..175 Practical velocity ranges for
different vehicles can by classified by the type of force supporting
the vehicle: buoyancy force for ships (<50 km/h), reaction force
for cars, trucks (<200 km/h) and trains (<350 km/h), and lift force
for airplanes (<800 km/h). As a consequence, the drag forces and
utilisation ratio increase as the vehicle changes from ships, to cars
and trucks, to trains, to planes.
Although the models presented below include velocity as a
variable, for the comparisons between current utilisation and the
practical limit the average velocity for each class of vehicle has
been held constant, to avoid unnecessary changes to delivered fi-
nal transport service. In the section to follow the model for the
passenger car is explained followed by a summary of the specific
differences for the other transport modes.
5.4.1 Cars
The use of passenger cars to transport people and goods results in
40 EJ of primary energy consumption. The car engine and drive
system (both conversion devices) convert fuel energy to vehicle
Figure 5.7 Forces on a car
140
§5.4
motion (passive system). The thrust force provided by the engine
is resisted by two drag forces, rolling resistance (mechanical) and
aerodynamic drag. According to Newton’s law, when the thrust
and drag forces are balanced the car moves at constant speed.
However in practice additional thrust is required to overcome iner-
tia and accelerate the car. Conversely, reducing the engine thrust,
or braking, produces a negative force which causes the vehicle to
decelerate.
Equation 5.22 describes these forces as a scalar equation for a
car. Forces are measured in Newtons ( N), or in joules per metre
( J/m or kJ/km) which conveniently are the same units as for the
utilisation ratios described in equations 5.1 and 5.2. At constant
speed, the thrust force (F ) is counteracted by the mechanical drag
(FM), aerodynamic drag (FA) of the car, and during acceleration,
inertia (FI). Thus, energy is finally dissipated as low temperature
heat to the road, air and braking system.
F = FM + FA + FI = µmg +
1
2
ρν2CDAf +m
dν
dt
(5.22)
Mechanical drag in the car is almost entirely due to rolling
resistance, as energy dissipated during deformation of the tyre.
(Mechanical friction in the engine, drive-train and wheel bearings
is considered part of the conversion device.) Ford et al. 29 explains
that tyre rubber is visco-elastic, resulting in one-fifth of the en-
ergy needed to flex the tyre being converted into heat by internal
hysteresis. Losses are primarily dependent on the vehicle weight
(mg), although some models include minor contributions from ve-
locity terms. Micro-slippage between the tyre and road and the
aerodynamic fan effects of the rotating wheel make up the remain-
ing resistance (less than 10%). For a typical car tyre the friction
coefficient µ ≈ 0.015.
Aerodynamic drag results from shear stresses created as air
flows around the car body, in both laminar and turbulent flow
141
§5.4
conditions. Santin et al. 176 describes the complex aerodynamic
system as having two principle mechanisms: 1) frictional drag,
which is proportional to the wetted area of the vehicle; and 2)
pressure drag, caused by flow separation around blunt vehicles,
the generation of lift and the boundary layer pressure loss. Aero-
dynamic drag is modelled adequately by the classical fluid-drag
equation, where ρ is the mass density of air (1.225 kg/m3), ν is
the average car velocity ( m/s), Af is the car’s projected frontal
area ( m2), and CD is the dimensionless drag coefficient, which
ranges from 0.2 to 0.6 for cars.
Inertia, the resistance of mass, must be overcome to accelerate
the car. Vehicle engines are therefore designed with significant
reserve power to accelerate the car, causing the engine to operate
at lower efficiencies during normal load. The energy expended
during acceleration dominates during urban driving, where the
car is forced to brake and accelerate in frequent cycles, and when
climbing gradients. However, inertia energy is ideally recovered by
decelerating without braking (coasting) or by using a regenerative
braking system.
On a gradient, the force of inertia due to gravity is described
by:
FI = mg sin θ ∼= mgθ (5.23)
where mg is the weight of the car and θ is the grade angle. For
the car acceleration, both the linear and rotational inertia are
combined to give:
FI = m
′a =
[
m+
4Iw
r2w
]
dν
dt
(5.24)
where the effective mass (m′) includes both the car mass, the ro-
tational inertia of the engine and wheel assemblies, and the accel-
142
§5.4
eration is found by differentiating the changes in velocity profile
for the driving pattern, with respect to time. The wheel assembly
inertia is typically small, but Ford et al. 29 claims that the en-
gine inertia, in the lowest gear ratio, can be more than 1.5 times
the vehicle mass. However, under cruising conditions the rota-
tional inertia is typically less than 5% of the inertia force.177 It is
common practice to simplify the above inertia terms into a single
multiplying factor based on surveyed driving patterns.
Significant opportunity exists to reduce the resistive forces in
cars, preventing the dissipation of valuable energy. The amount of
energy which can be saved is practically constrained by the limi-
tations of engineering materials and the available design options.
Practical limits are therefore governed by the possible range of
variables and coefficients in scalar equations. For example, drag
coefficients (CD) range from 1.05 for a cube, to the ideal limit of
0.04 for an elongated ‘teardrop’ shape—the practical limit for a
car is bounded by these two extremes. A value of 0.10 is selected
here for the practical limit, which is marginally higher than 0.09
for a ‘half-teardrop’ shape.
By varying the coefficients in equation 5.22 within realistic lit-
erature values, the practical energy savings available can be quan-
tified. Table 5.17 shows that practical improvements to the car’s
passive system can reduce energy use for cars by 91%.
The current utilisation ratio is calculated using data from Zachari-
adis and Samaras 178 and Hickman 179 which assesses the perfor-
mance of vehicle types across 15 European Union countries in 2000.
Equation 5.22 is used to calculate the mechanical resistance (from
rolling) (FM) and aerodynamic drag (FA) for each design of car.
Scaling factors are taken from Zachariadis and Samaras 178 to ac-
count for the inertia term. These estimate the fraction of power
consumed for acceleration in passenger cars—40% for urban driv-
ing, 30% for rural driving and 20% for highway driving—and are
multiplied by the sum of the other two drag forces, to calculate
(FI). The resulting forces for the 7 different car designs are av-
143
§5.4
eraged to find the ‘current utilisation limit’, using the percentage
distribution of total distance travelled in Europe (fi for i types of
car).
For the practical utilisation limit, the variables and coefficients
are reduced based on the ranges found in literature for real exam-
ples. The most significant gain comes from reducing the car mass
to 300 kg. The ‘world’s most fuel efficient vehicle’, the PAC-Car II
developed by Santin et al. 176 weighs only 29 kg when empty, with
the body being constructed from two layers of carbon reinforce-
ment with a combined mass of 390 g/m2. For safety reasons this
is considered unrealistically low. In contrast, the Rocky Mountain
Institute’s ‘2000 Revolution Hypercar’180 is a sport utility vehi-
cle (SUV) weighing only 857 kg. By reducing the vehicle size it
Table 5.17 Practical savings available in cars
Design1 m ν µ CD A fi FM FA FI F
t m/s m2 % N N N N
Gasoline
<1.4 l 1.0 19 0.015 0.40 1.9 36 147 163 130 440
1.4–2.0 l 1.2 20 0.015 0.40 2.0 28 177 201 155 533
>2.0 l 1.4 21 0.015 0.40 2.1 6 206 232 178 616
LDV2 2.1 17 0.015 0.50 2.2 3 309 199 223 731
Diesel
<2.0 l 1.3 19 0.015 0.40 2.0 11 191 174 156 521
>20.0 l 1.5 19 0.015 0.40 2.1 7 221 193 174 588
LDV2 2.1 16 0.015 0.50 2.2 8 309 165 219 693
Current3 1.3 19 0.015 0.41 2.0 100 188 183 157 528
Practical 0.3 19 0.001 0.10 1.5 100 3 33 13 49
Practical energy savings available 91%
Notes: 1 by fuel type and engine size in litres, 2 LDV = light duty
vehicle, 3 weighted average, by the distribution of total distance trav-
elled (fi). m = mass, ν = average velocity, µ = friction coefficient,
CD = drag coefficient, Af = frontal area, F = force, with subscripts M
mechanical, A aerodynamic and I inertia
144
§5.4
is expected that the empty car mass could be reduced to 200 kg,
with an extra 100 kg for the average passenger loading. Such a
reduction in mass makes energy recovery from the braking sys-
tem marginal, given the inertia forces are much lower, and the
additional equipment mass which would be required to store the
recovered energy.
For the tyre rolling resistance coefficient (µ), Santin et al. 176
quotes the 45-75R16 radial ply tubeless Michelin tyre as having
µ ≈ 0.00082—chosen as the practical limit—which is only frac-
tionally higher than train wheels, which make direct steel-on-steel
contact, for which µ ≈ 0.00073. Some cars on the road today al-
ready have a drag coefficient (CD) as low as 0.2, and the PAC-Car
II was measured in wind tunnels as CD = 0.075, leading to a prac-
tical limit of CD = 0.1 being selected. However, large reductions
in the car’s projected frontal area (Af ) are unlikely, due to the
requirement to carry passengers, resulting in a practical limit of
A =1.5 m2.
Figure 5.8 shows the current and practical limit for the urban,
rural and highway driving cycles. At the average speed, the prac-
tical energy savings are highest for the urban driving, equal to
96%, compared with 90% and 87% for rural and highway driving
respectively. This demonstrates that reduction in the vehicle mass
is particularly beneficial when the driving cycle involves frequent
acceleration and braking. In contrast, for the highway driving cy-
cle the car velocity dominates through aerodynamic drag. Reduc-
ing the car’s maximum speed is an operational rather tha technical
change, and therefore is not considered in this study. However, it
is noted that reducing the car velocity from 120 to 100 km/h re-
sults in a 30% reduction in the propulsive thrust needed from the
engine.
These results do not include any efficiency gains from design
changes such as: improvements to the engine, gearbox or drive
train; recovering energy in the braking system or switching off
the engine during idling; or reducing energy use in electrical ac-
145
§5.4
Figure 5.8 Resistive force versus velocity for the car
cessories, such as the air-conditioner, alternator, power steering or
water pump. The following operational parameters have also been
kept constant: the driving cycles, including the maximum speed,
acceleration and time between stops; and the driving conditions,
such as the number of passengers, the direction of travel relative
to the wind and the gradient of hills. Yet, a ten-fold saving is
available for the passive car system across all driving cycles.
5.4.2 Trucks
The transportation of goods by truck uses 38 EJ of global primary
energy. The loss of energy from the truck passive system is mod-
elled in a similar way to the car, with the results in table 5.18
146
§5.4
Table 5.18 Practical savings available in trucks
Design1 m ν µ cd Af fi FM FA FI F
t m/s m2 % kN kN kN kN
Diesel
3.5–7.5 t 5.5 16 0.012 0.8 7.0 31 0.6 0.9 0.6 2.1
7.6–16 t 11.8 17 0.012 0.9 7.0 20 1.4 1.0 1.3 3.7
16–32 t 24.0 19 0.012 1.2 7.0 34 2.8 1.9 2.7 7.4
>32 t 36.0 19 0.012 1.2 7.0 2 4.2 1.9 3.9 10.0
Coach 10.0 15 0.012 0.6 7.0 12 1.2 0.6 0 .9 2.7
Current2 14.1 18 0.012 0.95 7.0 100 1.7 1.3 1.5 4.5
Practical 12.7 18 0.0052 0.31 7.0 100 0.7 0.4 1.4 2.4
Practical energy savings available 54%
Notes: 1 by gross vehicle tonnage (GVT), 2 weighted average, by the
distribution of total distance travelled (fi)
showing that energy use can be practically reduced by 54%. The
current utilisation ratio is calculated using averaged data from
Hickman 179 and Zachariadis and Samaras 178 covering 5 ‘heavy
duty vehicle’ categories, including trucks, buses and coaches. For
the practical utilisation limit, reductions in the coefficients are
more conservative, than for the car. The light-weighting strategy
used for the car (75% reduction in mass) is limited for trucks be-
cause the transported goods make up a much larger proportion of
the total vehicle mass—typically 60% in trucks versus 5–10% in
cars. Furthermore, the mass of the empty truck performs the es-
sential function of supporting the goods during travel. Therefore,
only a 25% reduction in the unladen mass of the truck is assumed
practical.
Reductions in the projected frontal area (7.0 m2) are not deemed
practical, as most truck payloads are limited by volume rather the
weight, meaning truck designs tend to expand to the maximum
permitted dimensions in their class. The modest gains in rolling
resistance and drag coefficient are based on design values for Class
147
§5.4
8 tractor-trailer vehicles in the United States, provided by Ogburn
et al..181 As trucks are mainly driven on highways, the fraction of
energy use for acceleration is lower, leading to only minor savings
in the intertia force from mass reduction.
5.4.3 Planes
Travel by aircraft was responsible for 10 EJ of primary energy use
in 2005, approximately 10% of direct transport energy use. The
mechanisms for energy loss from the plane passive system are more
complex than for land-based forms of transport for two reasons.
Firstly, once in the air there is no mechanical drag force acting
on the plane, thus µ and FM from equation 5.22 can be set to
zero. Instead, the lift needed to support the weight of the plane
is included in the aerodynamic drag FA term. Secondly, the mass
of fuel forms a large proportion of the aircraft mass, so average
energy use over the entire trip is a function of this changing mass
of fuel, and thus is proportional to the journey distance. Decher 182
explains that over short distances, the energy required for taxiing
and climbing to cruise altitude dominates, leading to high average
energy use per kilometre. For long distances the average energy
also increases, as more energy is needed to carry the extra fuel over
the entire journey. In between is an optimum range for each plane
design, where the energy required per kilometre is minimised.
The Air travel - greener by design technology report by Green
et al. 183 describes in detail the underlying physics of air travel,
which is used for the aircraft model. For steady state level flight
(cruising) the range of an aircraft R ( km) can be calculated using
the Breguet equation:
R = X ln
(
W1
W2
)
(5.25)
where W1 and W2 are the initial and final weight of the aircraft
(mg in Newtons) and X is the key performance parameter in kilo-
148
§5.4
metres, defined as:
X = Hη
L
D
(5.26)
where H is the fuel heating value ( J/N or km), η is the overall
propulsion efficiency, and L/D is the lift/drag ratio (dimension-
less).
The Breguet equation must be divided into those terms which
relate to the conversion device efficiency (excluded from this anal-
ysis) and those which affect the passive system utilisation. The
overall propulsion efficiency (η) is the product of thermal efficiency
(ηE), which together with H are attributed to the conversion de-
vice, and the propulsive efficiency or Froude efficiency (ηP ), which
measures the efficiency of transferring rotational energy in the fan
blades to thrust in the air stream. This is analogous to the slip-
page of car tyres on the road, or the propeller losses in a ship,
so is included in the passive system calculation. Further passive
systems gains result from structural optimisation of the aircraft
frame and engine to reduce the weight of the plane, W1 and W2,
and aerodynamic improvements which affect the L/D ratio.
The propulsive efficiency (ηP ) of today’s ducted fan engines
lies between 80 and 85%183 and is close to the practical design
limit. Further increase to the engine bypass ratio would improve
the propulsive efficiency, but at the expense of an increase in fan
diameter and therefore the weight of the engine and nacelle. How-
ever, using a pair of counter-rotating unducted propellers, which
are not constrained in this way, could deliver significant gains in
propulsive efficiency and acceptable cruising speeds (Mach 0.8) for
long-range journeys.
The lift/drag ratio is optimised by reducing the drag force (FD)
on the aircraft, for a given lift (L), which is the weight of the plane
(mg) at cruising conditions. Two drag components (D) act on the
aircraft:
149
§5.4
Profile drag results from skin friction, pressure drag and wave
drag, and is represented by the coefficient CDO
Vortex drag is caused directly by the generation of lift
Overall drag force is minimised when these components are equal,
which leads to an expression for the maximum lift/drag ratio:(
L
D
)
max
=
√
piA
4kCDO
(5.27)
where A is the wing aspect ratio (= span2/area), k is the vortex
drag factor (= unity for an elliptically loaded wing), and CDO is
the profile drag coefficient (= drag calculated at zero lift). To-
gether, these three dimensionless constants define the maximum
aerodynamic efficiency of the airframe. In practice, most planes
operate at a cruise point where profile and vortex drag are not ex-
actly equal. This is corrected by multiplying the maximum L/D
ratio by a constant typically equal to 0.98.
Rearrangement of the Breguet equation gives equations for the
weight of fuel consumed during the journey (WF ) and the payload
weight (WP ) as functions of range (R), the key performance fac-
tor (X) and two structural constants (c1 and c2) related to the
maximum take-off weight and payload, respectively. These are
combined to give the specific fuel burn (SFB), which measures
the weight of fuel consumed during the entire journey (WF ) di-
vided by the payload weight (WP ) times the range (R), in units
of kg/kgm. The specific fuel burn is used as the utilisation ratio,
and is defined by:
SFB =
WF
WPR
=
( c2
X
)[ (1− 0.978e−Z)
Z (0.978e−Z − c1)
]
(5.28)
150
§5.4
where Z = R/X. The value 0.978 results from the provision of
2.2% additional take-off weight to account for the non-cruising fuel
required for taxiing, acceleration and climb.
Figure 5.9 shows specific fuel burn plotted against range for
a typical modern Swept Winged Aircraft (SWA) and for the ad-
vanced laminar flying wing Laminar Flying Wing (LFW) aircraft,
with Unducted Fan (UDF) propellers, calculated using data from
Green et al..183 For these curves, the aircraft design range is as-
sumed equal to the specific journey range, thus all aircraft fly at
their optimum design range. This is not true in practice, given the
small number of commercial aircraft designs and requirement for
route flexibility, leading to slightly optimistic curves especially for
the swept-wing aircraft. The LFW–UDF aircraft flies at a lower
speed (0.80 versus 0.85 Mach) and altitude (9,000 versus 10,000 m)
Figure 5.9 Specific fuel burn versus range for aircraft
151
§5.4
due to design limitations. The figure also gives the distribution of
fuel burn by journey range, taken from Green et al.,183(fig.70) which
is used to calculate the practical savings available of 46% in the
plane passive system, as shown in table 5.19.
The SWA has remained the dominant airframe design for more
than 50 years, and is therefore used as the base configuration for
the current utilisation ratio. The LFW with UDF is used as the
practical limit. Figure 5.10 shows the radically different shape
of the dominant SWA wing design and the proposed LFW de-
sign. This design concept, employing boundary layer suction to
maintain laminar flow over the entire airframe, and the design pa-
Table 5.19 Practical savings available in planes
Design ηP c1 c2 A k CDO L/D RD SFB
% km kg/t km
SW 81 0.315 2.0 10 1.2 0.0211 17.6 5000 0.176
LFW+UDF 95 0.375 1.9 5 1.1 0.0026 37.1 9000 0.096
Practical energy savings available 46%
Notes: constant values for both ratios: H = 4350 km, η = 0.3, n = 0.98.
RD is the aircraft design range
Figure 5.10 Swept Winged Aircraft versus Laminar Flying Wing
152
§5.4
rameters are based on test flights of the NASA F-94 aircraft, fitted
with a full-chord suction glove on the wing surface. Forty years of
research has shown that a LFW passenger aircraft could be prac-
tically built, despite the design being commercially risky. Using
the UDF engine design lowers the plane velocity from 0.85 to 0.80
Mach, and is likely to incur additional costs for noise reduction
measures and maintenance. However, the increased propulsive
efficiency of the UDF engine from 81% to 95%, in combination
with the LFW, make this option the most ‘economical large civil
aircraft that is feasible’ using kerosene fuel, according to Green
et al..183(p.47) The effects of changing the aircraft velocity, flight
path or propulsion are not included in the analsysis.
5.4.4 Ships
More than 90% of world trade by weight is carried by ships IPCC,151
yet due to their comparatively high efficiency per tonne trans-
ported, they consume less than 10% of primary energy for trans-
port (10 EJ). Ships are supported by the buoyancy force created
by the difference in average density between the ship and the wa-
ter. This allows very large and heavy loads to be transported, but
limits the upper velocity of travel due to rapid increase in resistive
drag at high speeds. The utilisation ratio for ships is influenced
primarily by the design of the propeller and the shape of the hull,
the mechanics of which are discussed below.
Ships are most commonly driven by ‘simple’ screw propellers.
Thrust is provided as the propeller blades turn through the water,
creating localised lift in the same way a wing creates lift in a plane.
An efficient propeller design creates this lift with minimum drag.
However Bertram 184 explains that the relative shortness of pro-
peller blades complicates the modelling of propeller performance.
He describes five different modelling techniques used by today’s
marine engineers.
The simplest of these propeller model, momentum theory, as-
153
§5.4
sumes the propeller is an actuator disk that accelerates the flow
in the axial direction by creating a pressure jump in the propeller
plane. The flow of water into the propeller is given by ρνinAin,
where ρ is the water density, νin is the velocity and Ain is the
cross-sectional area of the propeller plane. The action of the pro-
peller increases the velocity of the water flow (νout), but contracts
the cross-sectional area of the flow ‘tube’ (Aout). The change in
momentum is the thrust delivered by the propeller.
Thus, the ideal efficiency for the propeller can be written as:
η =
2νin
νout + νin
(5.29)
which predicts high efficiencies for propellers that deliver only a
small change in velocity. Practically, this means increasing the
propellor diameter so that thrust is delivered at lower revolutions
per minute (rpm). Further efficiency improvements result from
limiting the frictional drag on the hull downstream of the propeller,
which is caused by the increased velocity and lower pressure of
the flow; reducing interference from the ship’s wake; and avoiding
cavitation, caused by the rapid formation and collapse of vapour
bubbles in regions of low pressure near the blades.
The resistive drag forces acting on the ship’s hull in calm water
can be divided into three main components:
Friction resistance is caused by water particles adhering to the
wetted area of the hull and being dragged along with the ship. The
friction force is related to the shear stresses within the boundary
layer which forms.
Viscous pressure resistance results from variations in localised
flow velocities, vortices and separation of the flow, caused by the
shape of the hull, which increase the average overall shear stress in
the boundary layer. It can be minimised by designing long slender
hulls, allowing for practical limitations such as stability.
154
§5.4
Wave resistance is related to the wave system created by the
ship as it travels through the water (not the ship travelling through
waves). No satisfactory equation has been found to quantify this
effect, instead it is typically found experimentally.
Other less important resistance forces result from: aerody-
namic drag (which is significant in high-speed vessels); drag from
underwater appendages (such as the rudder and keel); wind, waves
and current; shallow water; and the cleanness of the hull.
The total resistive force (F ) can be represented in the single
term:
F =
1
2
ρν2(CT ) (5.30)
where ρ is the water density and CT is the total resistance co-
efficient. Bertram 184 describes several methods for determining
experimentally the forces acting on the hull, each with a separate
definition for CT . However, the components of CT typically in-
clude a frictional coefficient CF , which depends on the Reynolds
number, and residual coefficient CR, which includes the wave resis-
tance amongst other factors, and depends on the Froude number.
These factors depend critically on the velocity of the ship and
favour large ships.
The resulting set of equations are non-linear and can only be
solved using complex fluid dynamic modelling or by testing scale
models experimentally in ‘towing tanks’. This makes the forma-
tion of a simple generic model for ships impossible. Instead, design
parameters from a comprehensive survey of experimentally tested
hull designs are used to determine the practical savings available
of 63%, shown in table 5.20.
The final report to the International Maritime Organization,
entitled Study of greenhouse gas emissions from ships 185 is used
as the principal source for calculating the practical energy savings.
This report finds the technical advances in the efficiency of ships,
155
§5.4
Table 5.20 Practical savings available in ships
Type m ν fi Percentage reductions
Prop Hull Speed Total
kt m/s % % % % %
Oil tanker 275 7 29 5 35 18 58
Bulk carrier 70 7 23 5 28 23 56
Container 36 10 15 10 31 22 63
General cargo 12 8 19 5 43 25 73
Other1 8 14 5 43 25 73
Practical2 7 100 6 35 22 63
Practical energy savings available 63%
Notes: 1estimated based on general cargo, 2weighted average, by the
distribution of carbon emissions (fi)
and then reduces this potential based on the economic barriers to
implementation. For this analysis, the economic constraint is re-
moved and the most aggressive technologies chosen. Improvements
to three components relating to the passive system are examined:
the propeller design, the hull and the ship’s velocity.
For the propeller design, Henningsen compared several tech-
nical advances from literature and estimated their influence on
overall efficiency. Four options in particular are recommended for
consideration when designing new propellers: low RPM large di-
ameter propellers, pre- and post-swirl devices, ducted propellers
(for high thrust low speed vessels such as tankers) and twin counter
rotating propellers (for container vessels). They conclude that the
practical energy savings relating to the choice of propeller range
from 5–10%, depending on the type of ship.
The practical savings for the hull are based on MARINTEK’s
substantial database of model test results, also reported in Hen-
ningsen.185 The model data for each type of ship—oil tanker, bulk
carrier, container and general cargo—is normalised back to the
typical case ship size and plotted on a power versus speed graph
156
§5.4
to find the worst, average and best case curves. The hull designs
show a significant spread, up to plus or minus 30%. The differ-
ence between the average and best case curves is calculated at the
specific ship velocity to find the percentage energy savings.
Henningsen considers the effect of decreasing the ship speed
by 10%. This is accepted for ships, although not for other forms
of transport, because the journey time is not as critical. The
calculation includes an adjustment to increase the entire shipping
fleet by 10% to maintain the same global transport capacity. Yet,
even with this correction, the energy savings from reducing speed
are large and the most easily implemented.
5.4.5 Trains
Trains are used to provide both freight transport and passenger
transport, using 5.6 EJ and 2.2 EJ of global primary energy respec-
tively. Transporting goods and people by rail is typically more
efficient than by road due to ‘convoy effects’, whereby the resis-
tive forces do not increase significantly with length. Raghunathan
et al. 186 comments that although train speeds have reached over
300 km/h, in contrast to airplanes, the flow physics around trains
is not well understood and is complicated by the length of the
train, different wagon shapes and interaction with the fixed track,
structures, tunnels and platforms.
The drag forces acting on a moving train are typically modelled
using a second order polynomial with respect to velocity, of the
form:
F = A+Bν + Cν2 (5.31)
where A, B and C are constants found experimentally. There is
some debate about the underlying physical mechanism for each
coefficient. Term A is normally attributed to mechanical drag and
depends on the axle loading, roller bearing resistance, deflection
157
§5.4
of the track and the wheel to rail friction contact. Term A scales
with the weight of the train (mg), and is thus analogous to µ in
equation 5.22. Term C relates to the aerodynamic drag and, when
found experimentally over drive cycles, includes the inertia force.
According to Raghunathan et al. 186 it can be expressed as:
Cν2 =
1
2
ρAf
(
CD +
λ
d
l
)
ν2 (5.32)
where the additional terms to equation 5.22 are: d hydraulic di-
ameter of train ( m), l train length ( m) and λ hydraulic friction
coefficient (to account for the connecting parts between wagons
and the structures under the train).
Term B is traditionally expressed as a function of mass, for
example in Gawthorpe,187 but a recent study by Lukaszewicz 188
covering modern higher speed trains shows that term B relates
only to the train length. Depending on the exact definition, it is
likely there are contributions from both the mechanical drag and
the energy required to accelerate the intake air for combustion and
ventilation, to the velocity of the train.
Without a definitive model of the physical contributions to
resistance in trains, it is more difficult to accurately predict the
practical utilisation limit. Furthermore, national energy use data
for trains is rarely broken down beyond the high level freight and
passenger categories, and does not correlate with the case stud-
ies for which empirical coefficients are available. Nevertheless,
Lukaszewicz presents general train data (for example, configu-
ration, mass, length) and resistance coefficients (A, B and C)
for 5 configuration of ‘loco-hauled passenger trains’, 3 ‘high-speed
trains’ and 7 ‘freight trains’ operating in Sweden. These were com-
pared with studies by Gawthorpe,187 Hickman,179 Raghunathan
et al. 186 and Kemp 189 to validate the data.
Table 5.21, shows the practical energy savings that can be
achieved in freight trains (62%) and passenger trains (43%). The
158
§5.4
Table 5.21 Practical savings available in trains
Design m ν A B C F Ratio
t m/s N Ns/m Ns2/m2 kN kJ/t km
Freight
Current1 395 28 5,600 160 21 26.0 66
Practical 398 28 2,300 58 8 9.9 25
Practical energy savings available 62%
Passenger
Current2 300 42 3,300 28 11 23.2 77
Practical 398 42 2,300 58 8 18.1 46
Practical energy savings available 41%
High-speed3 398 56 2,300 58 8 29.4 74
Notes: 1 typical freight train has locomotive plus 24 wagons, 2 typical
passenger train has locomotive plus 5 wagons, 3 high-speed passenger
train used as a reference for the practical limit ratios
current utilisation ratios are calculated using poor performing con-
figurations from the Lukaszewicz 188 study, to correct for the tech-
nology differences between Swedish and global average trains. The
selected values are not as high (inefficient) as some configurations
in the literature, which may be due to the test method excluding
inertia forces, but a cautious approach is taken to avoid overesti-
mating the energy savings.
The Swedish X2 high-speed passenger train, with speeds ap-
proaching 250 km/h and a utilisation ratio of 74–80 kJ/t km, is
selected as the basis for the practical utilisation limit for both
freight and passenger trains. Comparison with the French TGV
and the Japanese Shinkansen ( 300 km/h, 67–82 kJ/t km) high-
speed trains show remarkably similar performance. These trains
are highly optimised, making use of light-weight materials and
streamlined body design to reduce aerodynamic and mechanical
drag, and are thought to be approaching practical design limits.
The aim of such advanced designs is to counteract the increase in
159
§5.5
fuel consumption at high speeds.
Two corrections are made to the high-speed train model to al-
low comparison. Firstly, the average speed is reduced from 56 m/s
(200 km/h) to: 28 m/s (100 km/h) for freight train, and 42 m/s
(150 km/h) for the passenger train. This approximates the ef-
fect of using the best available design at the current average train
speed, and reduces the practical utilisation limit. The second, is
to use the mass of the high-speed train (not the freight or passen-
ger train) in the calculation. This avoids a correction to the terms
A, and perhaps B, to account for the change in mass. However,
the difference in mass is cancelled out when the utilisation ratios
are normalised by mass. It is assumed that the use of coasting
and regenerative braking will make the energy to overcome inertia
almost negligible. Improvements to the engine or drive-train are
not considered.
5.5 Results and discussion
The practical energy savings in passive systems can now be calcu-
lated using equation 5.2, in which the primary energy use values
from table 5.1 are multiplied with the percentage gains available
from each model. Using primary energy values has the effect of
compounding the energy savings in the passive system back up
through the entire energy conversion chain, and allows the reduc-
tion in carbon emissions to be inferred.
Table 5.22 summarises the practical energy and carbon savings
for the passive energy systems. It shows that the greatest absolute
energy savings (column 4) are found in buildings, and in particular
heated spaces and appliance systems. As with the ranking of con-
version devices in the previous chapter (see table 4.6), aircraft are
prioritised lowest, demonstrating that the aircraft engine (device)
and the aircraft (system) are comparatively well optimised. In ad-
dition, the low ranking for aircraft confirms that scale of energy
flow through the system is a reasonable indicator of the absolute
160
§5.5
Table 5.22 Practical energy and carbon savings
Passive Practical Energy Energy Carbon Carbon
System savings demand savings emissions savings
% EJ EJ Gt CO2 Gt CO2
Heated space 98 72 71 3.3 3.3
Appliance 67 88 59 4.1 2.8
Furnace 62 67 42 4.0 2.5
Car 91 40 37 2.8 2.6
Driven system 59 56 33 3.3 1.9
Truck 54 38 20 2.6 1.4
Steam system 66 31 20 2.0 1.3
Hot water system 80 23 18 1.1 0.9
Illuminated space 95 18 17 1.1 1.0
Cooled space 100 14 14 0.8 0.8
Ship 63 10 6 0.7 0.4
Train 74 8 6 0.5 0.4
Plane 46 11 5 0.8 0.3
Building 83 215 179 10.5 8.8
Factory 62 154 95 9.3 5.7
Vehicle 70 106 74 7.3 5.1
Total 73 475 348 27.1 19.6
energy savings available.
The analysis presented in this chapter demonstrates that an aver-
age global energy saving of 73% is practically achievable in passive
energy systems. It is the first time that the practical energy sav-
ings in passive systems have been assessed separately from those
in conversion devices. Representative global energy data was of-
ten unavailable across the range of technology options, making
the accurate assessment of current energy use in passive systems
challenging. The allocation of energy use between the conversion
device and the passive system also proved difficult in some cases.
Nevertheless, basing the practical limit on fundamental engi-
neering principles has removed much of the uncertainty from the
analysis. Current energy use is forever changing, but at least the
161
§5.5
practical target by definition will remain stable. Furthermore,
technology options and efficiencies are surprisingly uniform across
the world’s geographic and economic zones. There are clear ex-
ceptions, such as wood fired stoves in the developing world, but in
many cases economic status determines whether or not you own an
energy consuming technology, not the efficiency of the technology.
Therefore, it is hoped that this research proves useful for under-
standing the function and utilisation of energy in passive systems,
and becomes a basis for setting future priorities for action in the
field of energy efficiency.
162
6 Discussion and Conclusions
Having determined the efficiency limits for all energy conversion
devices (theoretical limit) and passive energy systems (practical
limit), it is now possible to identify efficiency options from across
the entire global energy network.
6.1 What new conclusions can now be made?
If all conversion devices and passive systems could be operated
at their efficiency limit, then substantial reductions in primary
energy use and carbon emissions would result. Today’s conversion
devices are inefficient, converting on average only 11% of primary
energy input into a useful energy output. Passive systems use
only 27% of the useful energy input to deliver final services. The
remaining energy is currently lost as low temperature heat to the
environment. Multiplying these efficiency limits together gives an
overall efficiency for the entire network of 3% suggesting more than
a 30-fold improvement in efficiency is technically possible.
Table 6.1 shows the potential energy savings in conversion de-
vices and passive systems. The largest potential saving across all
devices and systems is found in the passive system of the heated
space. This is due to both the scale of energy use for heating build-
ing spaces and the possibility of thermally insulating the buildings
in most parts of the world, such that no artificial heat input is re-
quired to keep the occupants comfortable.
Clearly, the potential savings in energy cannot be all achieved
at the same time. Efficiency gains in the conversion devices cannot
be simply added to the gains in the passive systems (as stressed
163
§6.1
T
ab
le
6.
1
E
ffi
ci
en
cy
li
m
it
s,
en
er
gy
sa
v
in
gs
an
d
ca
rb
on
sa
v
in
gs
C
on
ve
rs
io
n
P
ri
m
ar
y
P
er
ce
n
t
E
n
er
gy
C
ar
b
on
P
as
si
ve
P
ri
m
ar
y
P
er
ce
n
t
E
n
er
gy
C
ar
b
on
d
ev
ic
e
en
er
gy
sa
v
in
gs
sa
v
in
gs
sa
v
in
gs
sy
st
em
en
er
gy
sa
v
in
gs
sa
v
in
gs
sa
v
in
gs
E
J
%
E
J
G
t
C
O
2
E
J
%
E
J
G
t
C
O
2
E
le
ct
ri
c
h
ea
te
r
58
93
54
3.
1
H
ea
te
d
sp
ac
e
72
98
71
3.
3
D
ie
se
l
en
gi
n
e
58
80
47
3.
3
A
p
p
li
an
ce
88
67
59
2.
8
E
le
ct
ri
c
m
ot
or
55
83
46
2.
6
F
u
rn
ac
e
67
62
41
2.
5
B
io
m
as
s
b
u
rn
er
49
94
45
0.
0
C
ar
40
91
37
2.
6
G
as
b
u
rn
er
47
88
41
2.
3
D
ri
ve
n
sy
st
em
56
59
33
1.
9
P
et
ro
l
en
gi
n
e
41
88
36
2.
5
T
ru
ck
38
54
20
1.
4
C
o
ol
er
33
98
33
1.
9
S
te
am
sy
st
em
31
66
20
1.
3
C
oa
l
b
u
rn
er
31
83
26
2.
2
H
ot
w
at
er
sy
st
em
23
80
19
0.
9
O
il
b
u
rn
er
28
86
24
1.
7
Il
lu
m
in
at
ed
sp
ac
e
18
95
17
1.
0
H
ea
t
ex
ch
an
ge
r
20
98
20
1.
2
C
o
ol
ed
sp
ac
e
14
10
0
14
0.
8
L
ig
h
t
d
ev
ic
e
18
96
17
1.
0
T
ra
in
8
74
6
0.
4
E
le
ct
ro
n
ic
16
98
15
0.
9
S
h
ip
10
63
6
0.
4
O
th
er
en
gi
n
e
10
82
8
0.
6
P
la
n
e
10
46
5
0.
3
A
ir
cr
af
t
en
gi
n
e
11
75
8
0.
5
H
ea
t
23
3
90
21
0
10
.4
B
u
il
d
in
g
21
5
83
17
9
8.
8
M
ot
io
n
17
5
83
14
5
9.
6
F
ac
to
ry
15
4
62
95
5.
7
O
th
er
67
98
65
3.
8
V
eh
ic
le
10
6
70
74
5.
1
T
ot
al
47
5
89
42
0
23
.8
T
ot
al
47
5
73
34
8
19
.6
164
§6.1
several times in this thesis) without saving more energy than is
consumed.
6.1.1 Where are efficiency gains most likely?
The analysis has shown that conversion devices on average operate
at only 11% of their theoretical potential. Yet, given the sizeable
effort already in progress to improve device efficiency, it is unlikely
that this ideal—a factor 10 improvement—will be approached in
the near future. Where should action and responses be focused?
Is it better to prioritise efforts on improving coal fired power sta-
tions or diesel engines? This is difficult to answer because the
theoretical saving in both energy and carbon emissions depends
not only on the efficiency of the individual device, but also on the
upstream efficiencies of all devices in the energy chain. A solution
to this question can be found by performing a sensitivity anal-
ysis to assess the energy savings that would be achieved from a
small independent change in efficiency for each type of conversion
device.
Applying an absolute efficiency change (for instance, increasing
each value of  by 1%) to each device might be misleading, as
achieving an equivalent gain in an already efficient device is likely
to be more difficult than for a less efficient device. Instead, the
conversion loss (which equals the theoretical energy saving) for
each device is reduced by 1%, and a modified device efficiency is
calculated using:
 ′ = + (1− )× 1% = 0.99+ 0.01 (6.1)
The efficiency of each device in turn was changed to the modi-
fied value ( ′) and the resulting total global energy input required
to deliver the same useful energy was calculated. This leads to a
sensitivity analysis of energy savings for the same relative level of
improvement in each device, and provides a more equitable way
165
§6.1
to compare and rank individual conversion devices, irrespective of
the location of the device in the energy network. This sensitivity
analysis is performed for individual conversion devices, as opposed
to energy chains, and the results are shown in figure 6.1. The chart
shows the reduction in energy and carbon emissions resulting from
a 1% reduction in the energy loss from each conversion device.
Efforts to improve the efficiency of coal-fired power stations
will deliver the most savings in the upstream fuel conversion and
electricity generation processes, because coal dominates electricity
generation. However, greater energy savings are available from
focusing individually on: biomass burners, coolers, gas burners
and petrol engines. Collectively, prioritising efficiency measures for
end-use conversion devices over fuel transformation and electricity
generation delivers more than five times the potential gain (28 EJ
versus 5 EJ). This is a surprising result, given the emphasis placed
on improving the efficiency of electricity generation, for example in
the International Energy Agency (IEA) report, Energy technology
perspectives 2008.1
Biomass burners emerge as the single most important conver-
sion device and where the largest energy savings can be achieved
from an incremental improvement in efficiency. These burners are
predominantly open fires, which burn wood, dung, crop waste,
coal and charcoal, to meet the energy needs of people living in the
developing world. The reason biomass burners top the sensitivity
list is due to the scale of use—used by half the world’s population
and burning more than 10% of global energy supply—and the
inefficiency of the burners, averaging only 7%. In this analysis,
biomass burners do not contribute to carbon emissions, because it
is assumed that the carbon dioxide (CO2) released during combus-
tion is equivalent to the CO2 absorbed when growing the biomass.
However, if the biomass is not replaced, for example in areas where
deforestation is a problem, then net carbon emissions to the atmo-
sphere result. Improving the efficiency of biomass burning stoves
is technically very easy, and has the added benefit of reducing res-
166
§6.1
Figure 6.1 Sensitivity ranking of individual conversion devices
piratory illness from the inhalation of smoke, which is ‘the single
biggest killer of children under five years of age’.190(p.24) However,
wide-scale dissemination of improved stove technology is held back
167
§6.1
by insufficient international political backing, limited funding and
the enormous number of open fires in use.
6.1.2 Application of this research
The results of this research demonstrate that significant energy
savings are technically possible in conversion devices and passive
systems. Yet such gains are not necessarily achievable, and his-
tory has shown that technical efficiency potentials are typically
under-realised. The use of theoretical and practical efficiency lim-
its presents an ideal, an ideal which may not be realised because
of economic or behavioural constraints. This work makes no at-
tempt to assess the economic costs of developing and deploying
advanced efficiency measures. Neither does it consider the many
socio-economic barriers to the uptake of new technologies. There-
fore, one must be careful to avoid claiming that the calculated
efficiency targets will or even should be attained.
Economists tend to assume that change can be brought about
by the choice of appropriately constructed policies. Such thinking
can neglect questions of the fundamental physical and engineer-
ing laws, which place limits on the energy that can be saved. By
providing an overview of the entire energy network and assess-
ing the potential impact of energy efficiency measures, this thesis
contributes to the field of policy-making by demonstrating both
the potential reach and the limits of energy efficiency. If techni-
cal solutions can be found, and are supported by well designed
policy measures, then large reductions in energy use and carbon
emissions are possible.
Further research is required to evaluate the effect that efficiency
gains have on the embodied energy in the device or system. It is
possible that some of the potential energy savings may be eroded
by the additional embodied energy required to manufacture the
improved conversion devices and passive system. Thus, a fraction
of the saved energy from transport or buildings may reappear as
168
§6.1
an increase in factories for the production of materials and goods.
However, the common assumption that more efficient devices are
always more energy intensive to build, does not always hold. Three
practical examples are given to illustrate this point:
1. It is possible to design a super-insulated and air-tight building,
which does not require artificial external heating or cooling, and
therefore also dispenses with the need for energy intensive capital
equipment such as boilers and air-conditioning units.
2. Light-weight streamlined cars deliver higher fuel efficiency as well
as reductions in the size of the engine, drive train, braking system
and structural components of the vehicle.
3. In many cases it is more cost effective for power utility compa-
nies to give away efficient light-bulbs and appliances than increase
generation capacity by building new power stations.
Such win/win options that reduce both operational energy use and
the energy embodied in capital equipment, should be prioritised.
The scope of this thesis is wide-ranging, covering a large body
of literature and drawing heavily on statistical energy data and
previous efficiency studies. The intricacies of specific energy pro-
cesses are only examined to the level of detail necessary to deter-
mine the limits to efficiency. The accuracy of the analysis could,
with more time, be improved. Despite best efforts to find repre-
sentative global data for current device and system use, in many
cases only regional or country specific data could be found. Allo-
cating energy flows and losses between the conversion devices and
passive systems proved difficult in some cases. It is hoped that
other energy researchers will contribute understanding from their
specific areas of expertise, to improve, correct and validate the re-
search. Nevertheless, even in its current form, the results of this
work are useful for directing future research priorities and setting
energy policy in the field of efficieincy.
169
§6.2
6.2 Outline of future research
Several deserving project ideas have emerged out of this research
into energy efficiency. The general intention is to take the thesis
material, divide and reformat it into more accessible forms, and
present this in a wide range of written and public settings. It is
hoped that this fundamental approach to energy efficiency will not
only gain traction among engineers and technical researchers, but
in addition, will be an influence on energy policy and be integrated
into the wider public debate on climate change. In the immediate
future, it is planned to pursue the following activities to promote
this goal.
6.2.1 Journal papers
In addition to the five published or pending journal papers listed
in the front matter (page iii), three further journal papers are
planned:
The first paper, extends the theoretical analysis of conversion
devices (chapter 4), by asking what are practical limits of energy
efficiency. Simple engineering models are created for each device
to understand how energy is lost during energy conversion. An
estimate of the practical energy savings available can then be cal-
culated by varying the coefficients in the model scalar equations,
within physical limits from literature.
The second paper, examines the efficiency limits for materials.
Allwood, Cullen and Milford 125 have explored the concept of ma-
terial efficiency, identifying several innovative routes for achieving
a 50% cut in industrial carbon emissions. However, their estimate
of potential material efficiency gains in 2050 was based on sur-
veys of known recycling techniques and yield improvements. This
proposed paper would add to this research by completing a fun-
damental study of material efficiency, based on physical limits for
reducing, conserving and recovering material during the product
life-cycle.
170
§6.2
The third paper, will present a summary of the entire body
of work, from the scale of energy flow, to the efficiency limits in
conversion devices and passive systems. This will be published in a
journal with a wide technical readership—such as the Philosophical
Transactions of the Royal Society A, or Environmental Science &
Technology—to provide a single reference point for the research.
6.2.2 Energy efficiency guide book
The wide ranging top-down approach taken to energy efficiency
in this research appears to have broad appeal amongst not just
engineers, but also scientists, economists, business people, policy
makers and the general public. Therefore it is proposed to convert
the thesis material into a guide book for energy efficiency. The
intention is to use simple language and frequent illustrations to
make energy efficiency options understandable and give the non-
academic reader practical efficiency options to reduce carbon emis-
sions. There is much experience to be gained from the successful
launch of MacKay’s book Sustainable energy - without the hot air
and the popular engineering texts published by Ashby, for exam-
ple, Materials and the environment.164 The book is expected to be
well received and will hopefully lead to further opportunities to
speak about energy efficiency to a diverse range of audiences.
6.2.3 Matlab model of global energy efficiency
The aim of this project is to transfer the data model from Microsoft
Office Excel™ to a matrix form in MATLAB® to allow multiple
data sets to be created and maintained. MATLAB® also has the
capability to draw the Sankey diagrams automatically, replacing
the laborious manual drawing process using Adobe® Illustrator®.
This added flexibility will allow:
• the energy maps to be easily translated into equivalent carbon
maps, giving greater accessibility for policy makers and the public
171
§6.2
• future energy scenarios to be explored, such as increased renewable
or nuclear energy, or a switch in transport designs from conven-
tional to hybrid or electrical
• the energy source and device efficiency to be updated for historical
and subsequent years, permitting time-series trend analysis
• the construction of similar regional, country or company based
energy maps
6.2.4 WellMet: steel, aluminium and the carbon targets 2010–2050
As the research into energy efficiency progressed, it was identified
that significant efficiency gains in the production of material goods
would be more difficult to achieve. For buildings and transporta-
tion, technical solutions such as insulation and light-weighting, are
already accepted, understood and likely to deliver significant effi-
ciency gains. In contrast, many of the simplest technologies have
already been exploited in industry, and the long capital cycles and
inertia of dominant technical solutions can prevent new ideas from
reaching the market.
In response Allwood, Cullen and Milford 125 completed an orig-
inal study into industrial carbon emissions from the five most en-
ergy intensive materials: steel, cement, plastic, paper and alu-
minium. This analysis shows that industry efforts to increase en-
ergy efficiency and recycling rates will not be sufficient to meet
a 60% cut in carbon emissions by 2050, against a doubling of
demand for materials. Based on this research work, Dr Julian
Allwood applied for and was awarded a five-year Engineering and
Physical Sciences Research Council (EPSRC) Leadership Fellow-
ship project (RG50904), entitled WellMet: steel, aluminium and
the carbon targets 2010–2050. The author will be part of the
leadership team for this project.
WellMet, aims to identify and validate all means to halve, by
2050, global carbon emissions from the production of steel and
172
§6.3
aluminium goods, against a projected doubling of demand. The
project has received funding of £1.5 million, comprises a team of
seven researchers, and is backed by a consortium of 20 global com-
panies. The research will firstly, evaluate all existing options for
carbon emissions reduction, ranging from low carbon energy sup-
plies (renewables, nuclear and Carbon Capture and Storage (CCS)
to energy and material efficiency. Secondly, radical new options
will be explored—such as non-destructive recycling (the reuse of
materials without remelting), light-weighting, and single-step heat
processing—to deliver much greater carbon reductions. The work
will combine physical and economic modelling, development and
demonstration of new technologies, and ongoing interaction with
industry built around a portfolio of fact sheets, case studies and
workshops on wider themes. Two key publications will be released
over the five year project: a mid-term project report that will be
widely distributed to stakeholders, and a final book to be used as
a reference guide for future work in this area.
6.3 Two promising ideas
Two additional ideas have emerged from the efficiency research,
and given no constraints on time, would be explored further. They
are noted here to enable other researchers to perhaps investigate
them in the future.
6.3.1 Conversion devices: an energy conversion matrix
Imagine that the global energy map (figure 3.2) is printed onto
a A4 rubber sheet. Possible future changes to the energy net-
work could then be visualised by stretching the rubber sheet. For
instance, if in a future scenario renewable energy sources are dou-
bled, then the renewable energy line on the Sankey diagram would
be stretched to be twice the width. If electric vehicles were to re-
place half the existing petrol engines, then the petrol engine line
would halve, a new line for electrical drives in vehicles would ap-
173
§6.3
pear, and some of the upstream oil supply would be displaced by
electricity generation. Thus, the rubber sheet would become a tool
for exploring alternative energy supply options and different tech-
nology pathways—for assessing large-scale changes to the energy
network, rather than just improvements.
During the last century, abundant supplies of inexpensive fos-
sil fuels have proved significant in shaping and driving the global
economy. For this reason, the conversion of chemical energy to
heat using combustion is the dominant conversion process, being
present in 90% of energy conversion pathways. It is not surpris-
ing therefore that much of the current energy efficiency research is
focused on around improving combustion processes. Although al-
ternative conversion pathways are available—for example electro-
magnetic radiation to electricity (in solar panels), nuclear energy
to heat (in fission reactors), and kinetic energy to electricity (in
hydro and wind turbines)—these technologies play only a minor
role in energy conversion.
In contrast to the dominance of fossil fuels, concerns over cli-
mate change and energy use reduction are still relatively new. Yet,
at some point in the future, perhaps in 100 to 200 years, the domi-
nance of fossil fuels energy supplies will likely end (due to excessive
environmental damage or because the fossil fuel supplies dwindle)
making alternative options more cost effective. During this tran-
sition, alternative conversion processes will become increasingly
important.
Energy can be divided into 6 different forms: radiation, chem-
ical, nuclear, thermal, mechanical, and electrical. This gives 36
possible energy transformations for a single conversion process. In
1969, Zwicky 191 proposed creating a matrix of all possible energy
conversions, which he called the ‘Morphological Box of Energy
Transformations’. The idea was to identify and explore alterna-
tive pathways for converting energy, however the work was not
completed. The concept of an energy conversion matrix has been
revisited several times, but only in outline form, for example Sum-
174
§6.3
mers,82(pp.150–1) Smil 40(p.14) and Ashby.164(p.21)
With the current pressures on fossil fuel supplies, it seems
an opportune time to revive Zwicky’s idea, quantify the current
energy conversions using an input-ouput framework and begin a
comprehensive and fundamental search for alternative energy con-
version routes. This would begin by reviewing all known tech-
nologies for converting energy, whether used in practice or still
in the conceptual stage. Then using the morphological approach,
a structured search of potentially new conversion pathways could
be explored. Based on future projections of the fossil fuel avail-
ability and public acceptance, scenarios could be developed and a
roadmap described for large-scale changes to the energy network.
6.3.2 Passive systems: reduce, conserve and recover
The analysis of passive energy systems separate from conversion
devices, has led to a new perspective on saving energy. This new
view originates from the concept of material efficiency, where the
demand for materials can be decreased by: reducing the mate-
rial in the product (reduce), extending the service-life of prod-
ucts (conserve) or recycling and reusing products (recover). It is
proposed to apply these concepts to useful energy—motion, heat,
light, cooling and sound—in passive energy systems.
The ‘reduce’ approach aims to deliver the same amount of final
service, using less useful energy. This includes measures such as:
increasing the passenger loading in cars, turning off light bulbs
when not in use, and ensuring electric motors and drives are
matched with the required load. The underlying aim is to elimi-
nate over-deisgn and use systems to their full capacity. (Energy
savings in passive system, should not be confused with reductions
in energy use through efficiency gains in conversion devices.)
The ‘conserve’ approach involves modifying passive systems to
extend the lifetime over which useful energy is applied. Examples
include: insulating and sealing buildings to conserve heat; reduc-
175
§6.3
ing vehicle aerodynamic drag and tyre-to-road friction to maintain
momentum; designing electronic displays that require no power to
maintain text on the screen.
The ‘recover’ approach requires improving the design of passive
systems to recover useful energy following its application. Motion
in vehicles can be recovered using regenerative braking, a tech-
nique already commonly used in electric motor systems. A possi-
ble alternative is the transfer of momentum to another moveable
body, as is observed in a Newton’s cradle. Interestingly, if a signif-
icant proportion of motion in vehicles could be recovered, it would
negate the need to light-weight vehicles.
The recovery of heat follows the same principle as down-cycling
of materials in products. Currently, most of the thermodynamic
availability of fossil fuels is wasted because it is used for low tem-
perature applications, such as heating air and water. However,
waste heat from an application is not lost (conservation law), in-
stead only its quality is degraded. Thus if the waste heat from
a high-temperature application can be used at a slightly lower
temperature, and so on, then a cascade of reducing heat quality
is formed. This concept is not new and is employed in industry,
using optimisation tools such as the ‘pinch analysis’ methodology
developed by Linnhoff and Hindmarsh.192 However, the theoreti-
cal potential to cascade heat at a national or global level has only
briefly been explored (see Lovins 37 and Nakicenovic and John 193),
and is an area that warrants further investigation.
Table 6.2 gives examples of reduce, conserve and recover for the
passive energy systems. The time scales for useful energy in pas-
sive systems are typically short. Light will be absorbed within less
than a second, sound over a few seconds, kinetic energy perhaps
lasts a few minutes, whereas heat or cooling is available for hours.
In contrast, the long service life time for materials in products—
from days to centuries—makes them an ideal area to pursue the
strategies of reduce, conserve and recover. Allwood, Cullen and
Milford 125 have discussed such ideas under the topic of material
176
§6.3
T
ab
le
6.
2
E
x
am
p
le
s
of
re
d
u
ce
,
co
n
se
rv
e
an
d
re
co
ve
r
in
p
as
si
ve
sy
st
em
s
S
er
v
ic
e
R
ed
u
ce
C
on
se
rv
e
R
ec
ov
er
T
ra
n
sp
or
t
fu
ll
p
as
se
n
ge
r
lo
ad
ae
ro
d
y
n
am
ic
d
es
ig
n
re
ge
n
er
at
iv
e
b
ra
k
in
g
en
gi
n
e
m
at
ch
ed
to
ve
h
ic
le
im
p
ro
ve
d
tr
affi
c
co
n
tr
ol
m
om
en
tu
m
tr
an
sf
er
S
tr
u
ct
u
re
lo
w
en
er
gy
m
at
er
ia
ls
se
rv
ic
e
li
fe
ex
te
n
si
on
m
o
d
u
la
r
d
es
ig
n
li
gh
t-
w
ei
gh
ti
n
g
re
-c
on
fi
gu
ra
b
le
d
es
ig
n
re
u
se
/r
ec
y
cl
e
S
u
st
en
an
ce
le
ss
su
p
p
ly
-c
h
ai
n
w
as
te
ad
va
n
ce
d
p
ac
ka
gi
n
g
n
u
tr
ie
n
t
re
co
ve
ry
?
le
ss
co
n
su
m
er
w
as
te
lo
n
g-
li
fe
p
ro
d
u
ct
s
H
y
gi
en
e
ta
n
k
le
ss
h
ot
w
at
er
in
su
la
te
eq
u
ip
m
en
t
w
as
te
-w
at
er
h
ea
t
re
co
ve
ry
T
h
er
m
al
co
m
fo
rt
h
ea
t/
co
ol
on
ly
p
eo
p
le
in
su
la
te
ai
r-
ti
gh
tn
es
s
ve
n
ti
la
ti
on
h
ea
t
re
co
ve
ry
u
se
so
la
r
h
ea
t
ga
in
re
d
u
ce
ai
r
le
ak
ag
e
th
er
m
al
m
as
s
C
om
m
u
n
ic
at
io
n
p
ri
n
t
le
ss
on
p
ap
er
st
at
ic
im
ag
e
d
is
p
la
y
s
u
n
-p
h
ot
o
co
p
y
in
g
Il
lu
m
in
at
io
n
im
p
ro
ve
d
lu
m
in
ai
re
d
es
ig
n
re
fl
ec
ti
ve
su
rf
ac
es
?
p
h
ot
ov
ol
ta
ic
su
rf
ac
es
?
il
lu
m
in
at
e
on
ly
ta
sk
ar
ea
177
§6.4
efficiency for the five most energy intensive materials: steel, con-
crete, paper, plastics and aluminium. However, the reduce, con-
serve and recover strategies are yet to be examined in detail.
Having ascertained in this research that large opportunities
exist for reducing energy use in passive systems, the next step
required is to identify the specific technical breakthroughs required
to deliver these gains. The field of energy efficiency would benefit
from a structured analysis which explores this potential using the
three different approaches: reduce, conserve and recover.
6.4 Conclusions
This thesis began with the questions: where should engineers focus
their efforts? Are the greatest efficiency gains to be found in light
bulbs or diesel engines, insulating houses or improving coal-fired
power stations? What are the limits to energy efficiency? How
should future research priorities be directed?
Approaching this problem from a technical perspective, based
on physical and engineering laws, has resulted in a consistent
framework for comparing efficiency options. Now, future efficiency
research and energy policy can be directed towards the actions that
will make the most difference.
The following original contributions are documented in this thesis:
1. For the first time the global flow of energy is traced from fuels
through to the final services, focusing on the technical conversion
devices and passive systems in each energy chain. By mapping
the scale and complexity of global energy flow in Sankey diagram
form, the technical areas which are likely to deliver the largest
efficiency gains, can be quickly identified.
2. A novel distinction is made between conversion devices, which up-
grade energy into more useable forms, and passive systems, from
178
§6.4
which energy is lost as low temperature heat in exchange for final
services. Devices and system need to be separated so that poten-
tial efficiency gains can be multiplied, instead of added, avoiding
a common double counting problem.
3. Theoretical efficiency limits are calculated for global conversion
devices using exergy analysis, and show a 89% potential reduction
in energy use. Such an analysis has not been performed since 1990.
The breakdown of conversion loss by engineering loss mechanisms
is the first known attempt to collate and rank global conversion
losses by technical categories.
4. Practical efficiency limits are calculated for global passive systems
based on engineering models, and demonstrate energy savings of
73% are achievable. No previous study has assessed the practical
energy savings in passive systems separately from conversion de-
vices. Significant percentage gains are found in technical solutions
that increase the thermal resistance of building fabrics and reduce
the mass of vehicles.
5. For the first time the relative energy and carbon savings from
fuel transformation, electricity generation, end-use conversion de-
vices are compared on an equal basis using sensitivity analysis. It
is revealed that improvements in end-use conversion devices—for
example, engines, heaters and light bulbs—will collectively deliver
five times more energy savings than the same relative improve-
ments in upstream electricity generation and distribution. For
individual devices, efforts should be focused on improving the effi-
ciency of, in relative order: biomass burners, refrigeration systems,
gas burners and petrol engines. For passive systems the priorities
are insulating buildings and furnaces, and reducing the mass of
vehicles.
179
§6.4
It is certain, that over time the numbers in this thesis will be
updated, the categories will be regrouped in more logical ways,
and further insight will be drawn from the Sankey diagrams. Such
modifications and challenges will be welcomed, because rather
than devalue the research they will serve to validate the energy ef-
ficiency framework presented here. For it is this framework—this
rational basis for assessing all future energy efficiency options—
that is the unique contribution of this thesis.
180
References
1. IEA (2008). Energy technology perspectives 2008 . International Energy
Agency, Paris, France.
2. DTI (2007). The energy challenge: a white paper on energy . Department
of Trade and Industry, The Stationery Office (TSO), London.
3. King, D.A. (2004). Environment - climate change science: Adapt, mit-
igate, or ignore? Science, 303, 176–177.
4. IPCC (2007). Summary for Policymakers. In: Climate Change 2007:
The Physical Science Basis. Contribution of Working Group I to the
Fourth Assessment Report of the Intergovernmental Panel on Climate
Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B.
Averyt, M.Tignor and H.L. Miller (eds.)]. Cambridge University Press,
Cambridge, United Kingdom and New York, NY, USA.
5. G8 (2008). G8 Hokkaido Toyako Summit Leaders Declaration, 8 July
2008 . www.mofa.go.jp (accessed July 2009).
6. UK (2008). Climate Change Act 2008, Chapter 27 . The Stationery Of-
fice, London, UK.
7. Baumert, K.A., Herzog, T. and Pershing, J. (2005). Navigat-
ing the numbers: greenhouse gas data and international climate policy .
World Resources Institute, Washington, USA.
8. IPCC (2001). Climate Change 2001: the scientific basis. Contribution to
the Third Assessment Report of the Intergovernmental Panel on Climate
Change. Cambridge University Press, UK.
9. King, D. (2005). Climate change: the science and the policy. Journal
Of Applied Ecology , 42, 779–783.
10. Kaya, Y. (1990). Impact of carbon dioxide emission control on GNP
Growth: interpretation of proposed scenarios. IPCC Energy and Indus-
try Subgroup. Response Strategies Working Group, Paris.
11. Yamaji, K., Matsuhashi, R., Nagata, Y. and Y., K. (1991). An
integrated systems for CO2/energy/GNP analysis: Case studies on eco-
nomic measures for CO2 reduction in japan. In Workshop on CO2 Re-
duction and Removal: Measures for the Next Century, International In-
stitute for Applied Systems Analysis, Laxenburg, Austria, 19-21 March
1991 .
12. Schipper, L., Unander, F., Murtishaw, S. and Ting, M. (2001).
Indicators of energy use and carbon emissions: explaining the energy
economy link. Annual Review of Energy and the Environment , 26, 49–
81.
181
13. Ramanathan, R. (2006). A multi-factor efficiency perspective to the
relationships among world GDP, energy consumption and carbon dioxide
emissions. Technological Forecasting and Social Change, 73, 483–494.
14. Raupach, M.R., Marland, G., Ciais, P., Le Quere, C.,
Canadell, J.G., Klepper, G. and Field, C.B. (2007). Global and
regional drivers of accelerating CO2 emissions. Proceedings Of The Na-
tional Academy Of Sciences Of The United States Of America, 104,
10288–10293.
15. IPCC (1996). Climate change 1995: The science of climate change..
Cambridge University Press, Cambridge, UK.
16. IPCC (2000). Special report on emissions scenarios. Intergovernmental
Panel on Climate Change, Cambridge University Press, Cambridge, UK.
17. UN (2006). World population prospects: the 2004 revision analytical
report . Department of Economic and Social Affaris, United Nations.
18. Nakicenovic, N. (1996). Decarbonization: Doing more with less. Tech-
nological Forecasting And Social Change, 51, 1–17.
19. Lund, H. (2007). Renewable energy strategies for sustainable develop-
ment. Energy , 32, 912–919.
20. MacKay, D.J. (2009). Sustainable energy - without the hot air . UIT
Cambridge, England.
21. Van der Veer, J. (2007). High hopes and hard truths dictate future,
25 june 2007. Times Online, http://business.timesonline.co.uk (accessed
July 2007).
22. Ion, S. (2007). We need an international and domestic climate change
strategy, warn the country’s leading engineers. The Royal Academy of
Engineering, www.raeng.org.uk (accessed July 2007).
23. Ulaby, F.T. (2006). Climate change and the proper role of scientists
and engineers. Proceedings Of The IEEE , 94, 1471–1472.
24. Gutowski, T.G. (2004). Handbook of Mechanical Engineering , chap.
Design and Manufacturing for the Environment. Springer-Verlag, Berlin,
Germany.
25. Jevons, W.S. (1865). The Coal Question: An Inquiry Concerning the
Progress of the Nation, and the Probable Exhaustion of Our Coal-Mines.
3rd ed. Revised by Flux, A.W., Kelley, A.M., New York.
26. Alcott, B. (2005). Jevons’ paradox. Ecological Economics, 54, 9–21.
27. Polimeni, J.M. and Polimeni, R.I. (2006). Jevons’ paradox and the
myth of technological liberation. Ecological Complexity , 3, 344–353.
28. Princen, T. (2005). The logic of sufficiency . MIT Press, Boston, USA.
29. Ford, K.W., Rochlin, G.I., Socolow, R.H., Hartley, D., Hard-
esty, D.R., Lapp, M., Dooher, J., Dryer, F., Berman, S.M. and
Silverstein, S.D., eds. (1975). Efficient use of energy. Conference on
the technical aspects of the more efficient use of energy, Princeton, NJ,
July 8–August 2, 1974 , American Institute of Physics, New York, USA.
30. Dyer, C.H., Hammond, G.P., Jones, C.I. and McKenna, R.C.
(2008). Enabling technologies for industrial energy demand manage-
ment. Energy Policy , 36, 4434–4443.
31. Smil, V. (2008). Long-range energy forecasts are no more than fairy
tales. Nature, 453, 154–154.
32. Bullard, C. and Herendeen, R.A. (1975). The energy cost of goods
and services. Energy Policy , 3, 268–278.
33. DEFRA (2004). EU emissions trading scheme: calculating the free al-
location for new entrants. DEFRA, UK.
34. Giampietro, M. (2006). Comments on “the energetic metabolism of
the European Union and the United States” by Haber and colleagues—
theoretical and practical considerations on the meaning and usefulness of
traditional energy analysis. Journal of Industrial Ecology , 10, 173–185.
35. Patterson, M.G. (1996). What is energy efficiency? concepts, indica-
tors and methodological issues. Energy Policy , 24, 377–390.
36. Phylipsen, G., Blok, K. and Worrell., E. (1998). Handbook on
international comparisons of energy efficiency in the manufacturing in-
dustry . Utrecht University, Department of Science, Technology and So-
ciety, The Netherlands.
37. Lovins, A.B. (1977). Soft energy paths: towards a durable peace. Har-
mondsworth, Middlesex, UK.
38. Schenk, N.J. and Moll, H.C. (2007). The use of physical indicators
for industrial energy demand scenarios. Ecological Economics, 63, 521–
535.
39. Farla, J.C.M. and Blok, K. (2000). The use of physical indicators
for the monitoring of energy intensity developments in the Netherlands,
1980–1995. Energy , 25, 609–638.
40. Smil, V. (2008). Energy in nature and society: general energetics of
complex systems. MIT Press, Boston, USA.
41. Ahern, J.E. (1980). The exergy method of energy systems analysis.
John Wiley & Sons, New York, USA.
42. de Beer, J., Worrell, E. and Blok, K. (1998). Future technologies
for energy-efficient iron and steel making. Annual Review of Energy and
the Environment , 23, 123–205.
43. C¸engel, Y. and Boles, M. (2006). Thermodynamics: an engineering
approach. McGrawl Hill, UK, 5th edn.
44. Ayres, R.U. (1998). Eco-thermodynamics: economics and the second
law. Ecological Economics, 26, 189–209.
45. Cleveland, C.J., Kaufman, R.K. and Stern, D.I. (2000). Aggre-
gation and the role of energy in the economy. Ecological Economics, 32,
301–317.
46. De Meester, B., Dewulf, J., Janssens, A. and VanLangenhove,
H. (2006). An improved calculation of the exergy of natural resources
for exergetic life cycle assessment (elca). Environmental Science & Tech-
nology , 40, 6844–6851.
47. Ayres, R.U. and Ayres, L.W. (1999). Accounting for Resources, 2:
The life cycle of materials. Edward Elgar, Cheltenham, UK.
48. Ertesvag, I. and Mielnik, M. (2000). Exergy analysis of the Norwe-
gian society. Energy , 25, 957–973.
49. Chapman, P.F. (1974). Energy costs: a review of methods. Energy
Policy , 2, 91–103.
50. Bullard, C.W., Penner, P.S. and Pilati, D.A. (1978). Net energy
analysis : handbook for combining process and input-output analysis.
Resources and Energy , 1, 267–313.
51. Haberl, H. (2006). On the utility of counting joules—reply to com-
ments by Mario Giampietro. Journal Of Industrial Ecology , 10, 187–
192.
52. UN (2006). Energy Statistics Yearbook 2003 . United Nations.
53. IEA (2004). World Energy Outlook 2004 . International Energy Agency,
Paris, France.
54. Council, W.E. (2004). Survey of Energy Resources 2004 . WEC, Lon-
don, UK.
55. BP (2006). Quantifying energy: BP statistical review of world energy
June 2006 . British Petroleum, www.bp.com.
56. Enerdata (2006). Enerdata Yearbook 2006 . Enerdata,
www.enerdata.com (accessed 15 Jan 2007).
57. DTI (2006). Digest of United Kingdom Energy Statistics 2006 . Depart-
ment of Trade and Industry, United Kingdom.
58. EIA (2006). Emissions of greenhouse gases in the United States 2005 .
Energy Information Administration, Washington DC, USA.
59. World Coal Institute, U., London (nd). WCI Website. WCI,
www.worldcoal.org (accessed July 2009).
60. World Steel (nd). World Steel website. World Steel,
www.worldsteel.org (accessed July 2009).
61. UNFCCC (2005). Key GHG data. United Nations Framework Conven-
tion on Climate Change.
62. Carbon Trust (2006). The carbon emission generated in all that we
consume. Carbon Trust, UK.
63. Farla, J.C.M. and Blok, K. (2001). The quality of energy intensity
indicators for international comparison in the iron and steel industry.
Energy Policy , 29, 523–543.
64. Karbuz, S. (1998). Achieving accurate international comparisons of
manufacturing energy use data. Energy Policy , 26, 973–979.
65. Boustead, I. and Hancock, G. (1979). Handbook of industrial energy
analysis. Ellis Horwood, Chichester, England.
66. Leontief, W. (1966). Input-output economics. Oxford University
Press, New York.
67. Vringer, K. (2005). Analysis of the energy requirement for household
consumption. Utrecht University, the Netherlands.
68. Wright, D.J. (1975). The natural resource requirements of commodi-
ties. Applied Economics, 7, 31–39.
69. Mongelli, I., Tassielli, G. and Notarnicola, B. (2006). Global
warming agreements, international trade and energy/carbon embodi-
ments: an input-output approach to the Italian case. Energy Policy , 34,
88–100.
70. Rhee, H.C. and Chung, H.S. (2006). Change in CO2 emission and its
transmissions between korea and japan using international input-output
analysis. Ecological Economics, 58, 788–800.
71. Kok, R., Benders, R.M.J. and Moll, H.C. (2006). Measuring the
environmental load of household consumption using some methods based
on input-output energy analysis: a comparison of methods and a discus-
sion of results. Energy Policy , 34, 2744–2761.
72. Wright, D.J. (1974). Good and services: an input-output analysis.
Energy Policy , 2, 307–315.
73. Hoekstra, R. and van der Bergh, J.J.C.J.M. (2003). Comparing
structural and index decomposition analysis. Energy Economics, 25, 39–
64.
74. Liu, F.L. and Ang, B.W. (2003). Eight methods for decomposing the
aggregate energy-intensity of industry. Applied Energy , 76, 15–23.
75. Ang, B.W. and Zhang, F.Q. (2000). A survey of index decomposition
analysis in energy and environmental studies. Energy , 25, 1149–1176.
76. Lenzen, M. and Dey, C. (2000). Truncation error in embodied energy
analyses of basic iron and steel products. Energy , 25, 577–585.
77. Bates, J., Brand, C., Davison, P. and Hill, N. (2001). Economic
evaluation of emissions reductions in the transport sector of the EU:
bottom-up analysis. European Commission.
78. Michaelis, P., Jackson, T. and Clift, R. (1998). Exergy analysis
of the life cycle of steel. Energy , 23, 213–220.
79. Sakamoto, Y., Tonooka, Y. and Yanagisawa, Y. (1999). Esti-
mation of energy consumption for each process in the Japanese steel
industry: a process analysis. Energy Conversion and Management , 40,
1129–1140.
80. Treloar, G.J. (1997). Extracting embodied energy paths from input-
output tables: towards an input-output based hybrid energy analysis
method. Economic Systems Research, 9, 375–.
81. Schmidt, M. (2008). The sankey diagram in energy and material flow
management. Journal of Industrial Ecology , 12, 82–94.
82. Summers, C.M. (1971). The conversion of energy. Scientific American,
225, 148–160.
83. Sims, R., Schock, R., Adegbululgbe, A., Fenhann, J., Kon-
stantinaviciute, I., Moomaw, W., Nimir, H., Schlamadinger,
B., Torres-Mart´ınez, J., Turner, C., Uchiyama, Y., Vuori, S.,
Wamukonya, N. and Zhang, X. (2007). Energy supply. In Climate
Change 2007: Mitigation. Contribution of Working Group III to the
Fourth Assessment Report of the Intergovernmental Panel on Climate
Change [B. Metz, O.R. Davidson, P.R. Bosch, R. Dave, L.A. Meyer
(eds)] . Cambridge University Press, Cambridge, United Kingdom and
New York, NY, USA.
84. USDOE (nd). Best Practices. Industrial Technologies Program, U.S.
Department of Energy, www.energy.gov (accessed July 2009).
85. EU (nd). Energy efficiency: action plan. EU, http://europa.eu/ (ac-
cessed 20 Dec 2006).
86. Cullen, J.M. and Allwood, J.M. (2009). The role of washing ma-
chines in life cycle assessment studies. Journal of Industrial Ecology , 13,
27–37.
87. Pacala, S. and Socolow, R. (2004). Stabilization wedges: solving
the climate problem for the next 50 years with current technologies.
Science, 305, 968–972.
88. Nakicenovic, N., Grubler, A., Inaba, A., Messner, S., Nils-
son, S., Nishimura, Y., Rogner, H.H., Schafer, A., Schrat-
tenholzer, L., Strubegger, M., Swisher, J., Victor, D. and
Wilson, D. (1993). Long-term strategies for mitigating global warm-
ing. Energy , 18, 401–609.
89. Goldemberg, J., ed. (2000). World Energy Assessment: energy and
the challenge of sustainability . United Nations Development Programme
(UNDP), United Nations Department of Economic and Social Affairs
(UNDESA), World Energy Council (WEC).
90. USDOE (2004). Energy use, loss and opportunities analysis: U.S. man-
ufacturing & mining . Industrial Technologies Program, U.S. Department
of Energy.
91. IEA (2006). Key world energy statistics 2006 . International Energy
Agency, Paris, France.
92. Hawksworth, J. (2006). The world in 2050: implications of global
growth for carbon emissions and climate change policy . PriceWater-
houseCoopers, UK.
93. IEA (2006). World Energy Outlook 2006 . International Energy Agency,
Paris, France.
94. Blok, K. (2004). Improving energy efficiency by five percent and more
per year? Journal of Industrial Ecology , 8, 87–99.
95. Craig, P.P., Gadgil, A. and Koomey, J.G. (2002). Can history
teach us? a retrospective examination of long-term energy forecasts for
the united states. Annual Review of Environment and Resources, 27,
83–118.
96. Focacci, A. (2005). Empirical analysis of the environmental and energy
policies in some developing countries using widely employed macroeco-
nomic indicators: the cases of Brazil, China and India. Energy Policy ,
33, 543–554.
97. von Weizacker, E., Lovins, A.B. and Lovins, L.H. (1998). Factor
four: doubling wealth, halving resource use. Earthscan, London, UK.
98. Monbiot, G. (2006). Heat: how to stop the planet burning . Allen Lane,
Penguin Press, UK.
99. Worrell, E., Price, L. and Galitsky, C. (2004). Emerging energy-
efficient technologies in industry case studies of selected technolgies. En-
vironmental Energy Technolgies Division, University of California, USA.
100. EU (2006). Action plan for enrgy efficiency: realising the potential .
Commision of the European Communities.
101. IEA (2005). Energy technologies at the cutting edge. International En-
ergy Agency, Paris, France.
102. HM Treasury (2005). Energy efficiency innovation review . HM Trea-
sury, UK.
103. Vattenfall (2007). Global climate abatement map. Vattenfall,
www.vattenfall.com (accessed 20 Feb 2007).
104. Bressand, F., Farrell, D., Hass, P., Morin, F., Nyquist, S.,
Remes, J., Roemer, S., Rogers, M., Rosenfeld, J. and Woet-
zel, J. (2007). Curbing global energy demand growth: the energy pro-
ductivity opportunity. Tech. rep., McKinsey Global Institute.
105. Energetics (2004). Steel Industry: Energy Bandwidth Study . US De-
partment of Energy.
106. Stubbles, J. (2000). Energy Use in the U.S. Steel Industry: A Histor-
ical Perspective and Future Opportunities. US Department of Energy.
107. Energetics (2000). Energy and Environmental Profile of the U.S. Iron
and Steel Industry . US Department of Energy.
108. Fruehan, R., Fortini, O., Paxton, H. and Brindle, R. (2000).
Theoretical minimum energies to produce steel for selected conditions.
US Department of Energy, Office of Industrial Technologies, Washing-
ton, DC, USA.
109. Reistad, G.M. (1975). Available energy-conversion and utilization in
united-states. Journal of Engineering for Power-Transactions of the
ASME , 97, 429–434.
110. Ertesvag, I. (2001). Society exergy analysis: a comparison of different
societies. Energy , 26, 253–270.
111. Nakicenovic, N., Gilli, P.V. and Kurz, R. (1996). Regional and
global exergy and energy efficiencies. Energy , 21, 223–237.
112. Rosen, M., Dincer, I. and Kanoglu, M. (2008). Role of exergy
in increasing efficiency and sustainability and reducing environmental
impact. Energy Policy , 36, 128–137.
113. Lightfoot, H.D. (2007). Understand the three different scales for mea-
suring primary energy and avoid errors. Energy , 32, 1478–1483.
114. IEA (2008). 2005 Balance table for the world . International Energy
Agency, Paris, France, www.iea.org, (accessed July 2008).
115. IEA (2007). Key world energy statistics 2007 . International Energy
Agency, Paris, France.
116. Gantz, J., Chute, C., Manfrediz, A., Minton, S., Reinsel, D.,
Schlichting, W. and Toncheca, A. (2008). The diverse and explod-
ing digital universe. IDC/EMC, www.emc.com (accessed Jan 2009).
117. IEA (2006). Light’s labour’s lost: policies for energy-efficient lighting .
International Energy Agency, Paris, France.
118. Ashby, M. (1999). Material selection in mechanical design.
Butterworth-Heinemann, Oxford, UK, 2nd edn.
119. IEA (2006). Energy Technolgy Perspectives: scenarios and strategies to
2050 . International Energy Agency, Paris, France.
120. EUROFER (2008). Economic and steel market outlook 2008–2009 . Eu-
ropean Confederation of Iron and Steel Industries, Brussels, Belgium.
121. IEA (2007). Tracking industrial energy efficiency and CO2 emissions.
International Energy Agency, Paris, France.
122. BCA (2008). Private communication. British Cement Association, UK.
123. FAOSTAT (2008). Food and Agriculture Organization Corporate Sta-
tistical Database. United Nations Food and Agriculture Organisation,
http://faostat.fao.org/ (accessed July 2009).
124. IAI (2006). Global aluminium recycling: a cornerstone of sustainable de-
velopment . Global Aluminium Recycling Committee, International Alu-
minium Institute, London, UK.
125. Allwood, J., Cullen, J. and Milford, R. (2010). Options for
achieving a 50% cut in industrial carbon emissions by 2050. Environ-
mental Science & Technology , (in press).
126. IEA (2008). Worldwide trends in energy efficiency . International Energy
Agency, Paris, France.
127. IEA (2005). Energy statistics manual . International Energy Agency,
Paris, France.
128. O’Callaghan, P. (1981). Design and management for energy conser-
vation. Pergamon Press, Oxford, UK.
129. Culp, A. (1991). Principles of energy conversion. McGrawl Hill, UK,
2nd edn.
130. Gilli, P.V., Nakicenovic, N. and Kurz, R. (1996). First- and
second-law efficieincies of the global and regional energy system. Inter-
national Institute for Applied Systems Analysis (IIASA), Austria.
131. Hammond, G. and Stapleton, A. (2001). Exergy analysis of the
United Kingdom energy system. Proceedings of the Institution of Me-
chanical Engineers Part A-Power and Energy , 215, 141–162.
132. Warr, B., Schandl, H. and Ayres, R. (2007). Long term trends in
resource exergy consumption and useful work supplies in the UK, 1900–
2000 . CSIRO Sustainable Ecosystems, Canberra, Australia.
133. Ayres, R.U., Ayres, L.W. and Pokrovsky, V. (2005). On the effi-
ciency of us electricity usage since 1900. Energy , 30, 1092–1145.
134. IEA (2007). Energy use in the new millenium: trends in IEA countries.
International Energy Agency, Paris, France.
135. Dunbar, W. and Lior, N. (1994). Sources of combustion irreversibil-
ity. Combustion Science and Technology , 103, 41–61.
136. Prins, M. and Ptasinski, K. (2005). Energy and exergy analyses of
the oxidation and gasification of carbon. Energy , 30, 982–1002.
137. Dunbar, W., Moody, S. and Lior, N. (1995). Exergy analysis of an
operating boiling-water-reactor nuclear-power station. Energy Conver-
sion and Management , 36, 149–159.
138. Durmayaz, A. and Yavuz, H. (2001). Exergy analysis of a pressurized-
water reactor nuclear-power plant. Applied Energy , 69, 39–57.
139. Rakopoulos, C.D. and Giakoumis, E.G. (2006). Comparative first-
and second-law parametric study of transient diesel engine operation.
Energy , 31, 1927–1942.
140. Turgut, E.T., Karakoc, T.H. and Hepbasli, A. (2007). Exergetic
analysis of an aircraft turbofan engine. International Journal of Energy
Research, 31, 1383–1397.
141. Mecrow, B. and Jack, A. (2006). Efficiency trends in electric ma-
chines and drives. Foresight programme, Office of Science and Innova-
tion, UK.
142. Kotas, T. (1985). The exergy method of thermal plant analysis. But-
terworths, London, UK.
143. EN15251 (2007). Indoor environmental input parameters for design and
assessment of energy performance of buildings addressing indoor air
quality, thermal environment, lighting and acoustics. Brussels: Comite
Europe en de Normalisation; 2007.ating and Air-Conditioning Engi-
neers (ASHRAE); 2004 Brussels: Comite Europe en de Normalisation;
2007.ating and Air-Conditioning Engineers (ASHRAE); 2004 Brussels:
Comite Europe en de Normalisation; 2007.ating and Air-Conditioning
Engineers (ASHRAE); 2004 The European Committee for Standardiza-
tion (CEN), Brussels.
144. Nicol, F. and Humphreys, M. (2009). Derivation of the adaptive
equations for thermal comfort in free-running buildings in european stan-
dard en15251. Building and Environment , In Press, Corrected Proof,
–.
145. Olesen, B.W. (2007). The philosophy behind en15251: Indoor envi-
ronmental criteria for design and calculation of energy performance of
buildings. Energy and Buildings, 39, 740–749.
146. BRE (2009). PassivHausUK: towards sustainable design. BRE, Watford,
UK, www.passivhaus.org.uk (accessed Aug 2009).
147. CIESIN (2009). Centre for International Earth Science Infor-
mation Network (CIESIN). Columbia University, New York,
www.ciesin.columbia.edu/ (accessed July 2009).
148. PVGIS (2009). Photovoltaic Geographical Information System
(PVGIS). Institute for Energy, European Commission, Italy,
http://re.jrc.ec.europa.eu/pvgis/index.htm (accessed July 2009).
149. Angel, S., Mayo, S.K. and Stephens, W.L., Jr (1993). The hous-
ing indicators program: a report on progress and plans for the future.
Netherlands Journal of Housing and the Built Environment , 8, 13–48.
150. Hoare, R. (2009). World Climate database. www.worldclimate.com
(accessed July 2009).
151. IPCC (2007). Climate Change 2007: Mitigation. Contribution of Work-
ing Group III to the Fourth Assessment Report of the Intergovernmen-
tal Panel on Climate Change [B. Metz, O.R. Davidson, P.R. Bosch,
R. Dave, L.A. Meyer (eds)] . Cambridge University Press, Cambridge,
United Kingdom and New York, NY, USA.
152. USDOE (2009). Industrial Technologies Program website. Energy
Efficieincy and Renewable Energy, U.S. Department of Energy,
www.energy.gov (accessed July 2009).
153. Kreith, F., Goswami, D.Y. and Sandor, B.I., eds. (2004). The
CRC Handbook of Mechanical Engineering . CRC, 2nd edn.
154. Ja¨ger-Waldau, A., ed. (2007). Scientific techinical reference system
on renewable energy and energy end-use efficiency: status report 2006 .
Institute for Environment and Sustainability, European Commission,
Luxembourg.
155. Warwick, H. and Doig, A. (2004). Smoke - the killer in the kitchen:
indoor pollution in developing countries. ITDG Publishing, London, UK.
156. Brundrett, G.W. and Poultney, G. (1979). Saucepan lids: the key
to low energy cooking. International Journal of Consumer Studies, 3,
195–204.
157. Goorskey, S., Smith, A. and Wang, K. (2004). Home energy briefs:
#8 kitchen appliances. Rocky Mountain Institute, Snowmass, Colorado,
USA.
158. LBNL (1998). Technical support document: energy conservation stan-
dards for consumer products: cooking products. Lawrence Berkely Na-
tional Laboratory, Berkley, California, for U.S. Department of Energy.
159. Rosenfeld, A.H. (1999). The art of energy efficiency: Protecting the
environment with better technology. Annual Review Of Energy And The
Environment , 24, 33–82.
160. Strandbakken, P. (2005). Social constraints to eco efficiency: refrig-
erators and freezers. In European Sociological Association (ESA) confer-
ence, Torun, Poland .
161. Sunfrost (2009). Sun Frost website. Sun Frost, www.sunfrost.com (ac-
cess July 2009).
162. Goorskey, S., Smith, A. and Wang, K. (2004). Home energy briefs:
#6 cleaning appliances. Rocky Mountain Institute, Snowmass, Col-
orado, USA.
163. Palandre, L. and Clodic, D. (August 2003). Comparison of heat
pump dryer and mechanical steam compression dryer. International
Congress of Refrigeration, Washington, D.C..
164. Ashby, M. (2009). Material and the environment . Elsevier:
Butterworth-Heinemann, Oxford, UK.
165. Hemingway, B.S. (1990). Thermodynamic properties for bunsenite,
nio, magnetite, Fe3O4, and hematite, Fe2O3, with comments on selected
oxygen buffer reactions. American Mineralogist , 75, 781–790.
166. Hemingway, B.S., Robie, R.A. and Kittrick, J.A. (1978). Re-
vised values for gibbs free-energy of formation of Al(OH)4aq, diaspore,
boehmite and bayerite at 298.15k and 1 bar, thermodynamic properties
of kaolinite to 800 k and 1 bar, and heats of solution of several gibbsite
samples. Geochimica Et Cosmochimica Acta, 42, 1533–1543.
167. Neelis, M., Patel, M., Blok, K., Haije, W. and Bach, P. (2007).
Approximation of theoretical energy-saving potentials for the petro-
chemical industry using energy balances for 68 key processes. Energy ,
32, 1104–1123.
168. Taylor, M., Tam, C. and Gielen, D. (2006). Energy efficiency and
CO2 emissions from the global cement industry . Energy Technology Pol-
icy Division, International Energy Agency, Paris, France.
169. Kinstrey, R.B. and White, D. (2006). Pulp and paper indus-
try: energy bandwidth study . American Institute of Chemical Engineers
(AIChE), USA.
170. de Beer, J., Worrell, E. and Blok, K. (1998). Long-term energy-
efficiency improvements in the paper and board industry. Energy , 23,
21–42.
171. Chen, W.H., Chung, Y.C. and Liu, J.L. (2005). Analysis on energy
consumption and performance of reheating furnaces in a hot strip mill.
International Communications In Heat And Mass Transfer , 32, 695–
706.
172. Zircoa (2009). Zircoa insulating structures. Zircoa, www.zircoa.com
(accessed July 2009).
173. Lovins, A. (2007). Stanford energy lectures. Rocky Mountain Institute,
www.rmi.org (accessed July 2009).
174. Gabrielli, G. and von Ka´rma´n, T. (1950). What price speed? spe-
cific power required for propulsion of vehicles. Mechanical Engineering
ASME , 72, 775–81.
175. Yong, J., Smith, R., Hatano, L. and Hillmansen, S. (2005). What
price speed - revisited. Ingenia, 22, 46–51.
176. Santin, J.J., Santin, Onder, C., Bernard, J., Isler, D., Kobler,
P., Kolb, F., Weidmann, N. and Guzzella, L. (2007). The world’s
most fuel efficient vehicle: design and development of PAC-Car II . vdf
Hochschulverlag AG an der ETH Zu¨rich, Zu¨rich.
177. Sorvan, G. and Blaser, S. (2003). A contribution to understanding
automotive fuel economy and its limits. S.A.E. Transactions, 112, 1715–
1740.
178. Zachariadis, T. and Samaras, Z. (2001). Validation of road trans-
port statistics through energy efficiency calculations. Energy , 26, 467–
491.
179. Hickman, A., ed. (1999). MEET: methodology for calculating transport
emissions and energy consumption. Office for Official Publications of the
European Communities, Luxemborg.
180. Lovins, A.B. and Cramer, D.R. (2004). Hypercars®, hydrogen, and
the automotive transition. International Journal of Vehicle Design, 35,
50–85.
181. Ogburn, M., Ramroth, L. and Lovins, A. (2008). Transformational
trucks: determining the efficiency limits of a Class-8 tractor-trailer .
Rocky Mountain Institute, USA.
182. Decher, R. (1994). Energy conversion. Oxford University Press, Ox-
ford, UK.
183. Green, J.E., Cottington, R.V., Davies, M., Dawes, W.N.,
Fielding, J.P., Hume, C.J., Lee, D.J., McClarty, J., Mans,
K.D.R., Mitchell, K. and Newton, P.J. (2003). Air Travel -
Greener by design: the technology challenge. Report of the Technnol-
ogy Subgroup, www.greenerbydesign.org.uk (accessed July 2009).
184. Bertram, V. (2000). Practical ship hydrodynamics. Butterworth-
Heinemann, Oxford, UK.
185. Henningsen, R.F. (2000). Study of greenhouse gas emissions from
ships. Norwgian Marine Technology Research Institute (MARINTEK),
Trondheim, and International Maritime Organisation (IMO), London,
UK.
186. Raghunathan, R.S., Kim, H.D. and Setoguchi, T. (2002). Aero-
dynamics of high-speed railway train. Progress in Aerospace Sciences,
38, 469–514.
187. Gawthorpe, R. (1982). Train drag reduction from simple design
changes. International Journal of Vehicle Design, 3, 263–273.
188. Lukaszewicz, P. (2007). Running resistance -results and analysis of
full-scale tests with passenger and freight trains in sweden. Proceedings
of the Institution of Mechanical Engineers, Part F: Journal of Rail and
Rapid Transit , 221, 183–192.
189. Kemp, R. (2007). Traction energy metrics (T618). Rail Safety and Stan-
dard Board, UK.
190. Gordon, B., MacKay, R. and Rehfuess, E. (2004). Inheriting
the World: The Atlas of Children’s Health and the environment . World
Health Organization, Geneva.
191. Zwicky, F. (1969). Discovery, invention, resarch: through the morpho-
logical approach. The Macmillan Company, New York, USA.
192. Linnhoff, B. and Hindmarsh, E. (1983). The pinch design method
for heat-exchanger networks. Chemical Engineering Science, 38, 745–
763.
193. Nakicenovic, N. and John, A. (1991). CO2 reduction and removal
and removal measures for the next century. Energy , 16, 1347–1377.
List of Figures
2.1 Diagram of the potential gains from energy efficiency 10
2.2 Outline of chapter 11
2.3 Levels in process analysis 29
3.1 The flow-path of energy 49
3.2 Tracing the global flow of energy from fuel to service 58
3.3 Delivering passenger transport using alternative energy chains 62
4.1 The global map of energy conversion efficiency 79
5.1 Balancing heat loads in a heated building space 94
5.2 Heat requirement in temperate and cold climatic zones 101
5.3 Balancing heat and cooling loads in a cooled building space 103
5.4 On-demand hot water and heat recovery from waste water 107
5.5 Reduction in luminous flux from using task lighting 111
5.6 Heat balance for cooking vessel (pot) 116
5.7 Forces on a car 140
5.8 Resistive force versus velocity for the car 146
5.9 Specific fuel burn versus range for aircraft 151
5.10 Swept Winged Aircraft versus Laminar Flying Wing 152
6.1 Sensitivity ranking of individual conversion devices 167
195
List of Tables
2.1 Units for measuring energy use 12
2.2 Methods for aggregating energy use 20
2.3 Input-output matrix example 24
2.4 Second order consumption for a unit of steel 24
2.5 Available energy savings from end-use sectors 39
2.6 Energy consumption for steel 1990 41
2.7 Current approaches for prioritising energy use reduction 44
3.1 Energy sources and transfer mediums 51
3.2 End-use conversion devices 53
3.3 Passive energy systems 54
3.4 Materials and products 54
3.5 Final services 55
3.6 Technical components ranked by the scale of energy use 60
4.1 Energy and exergy efficiencies for conversion devices 70
4.2 Survey of conversion device efficiencies 72
4.3 Energy and exergy efficiencies of end-use conversion devices 74
4.4 Energy loss mechanisms 76
4.4 Energy loss mechanisms (continued) 77
4.5 Comparing the efficiency of conversion devices 81
4.6 Theoretical energy and carbon savings 83
4.7 Categorisation of energy loss by mechanism 85
5.1 Global energy supply allocated to passive systems 90
5.2 Current utilisation ratio for final services 92
5.3 Typical and practical limit constants for heated spaces 99
5.4 Climatic data for temperate and cold geographical zones 100
5.5 Climatic data for tropical and desert zones 104
5.6 Practical energy savings in hot water systems 109
5.7 Practical energy savings in passive lighting systems 112
5.8 Practical energy savings in appliance systems 114
5.9 Enthalpy change for heating and boiling a litre of water 116
5.10 Practical energy savings in passive cooking systems 120
5.11 Practical energy savings in refrigerators and freezers 122
5.12 Enthalpy change for cooling and freezing a litre of water 123
5.13 Energy use in industry 130
196
5.14 Practical energy savings in the furnace 133
5.15 Energy use in driven systems 138
5.16 Practical energy savings in steam systems 139
5.17 Practical savings available in cars 144
5.18 Practical savings available in trucks 147
5.19 Practical savings available in planes 152
5.20 Practical savings available in ships 156
5.21 Practical savings available in trains 159
5.22 Practical energy and carbon savings 161
6.1 Efficiency limits, energy savings and carbon savings 164
6.2 Examples of reduce, conserve and recover in passive systems 177
Nomenclature
CH4 methane
CO2 carbon dioxide
N2O nitrous oxide
BAT Best Available Technologies
BTU British Thermal Unit
CCS Carbon Capture and Storage
CIESIN Center for International Earth Science Information Network
CFL Compact Fluorescent Light-bulb
CHP Combined Heat and Power
COP Coefficient of Performance
DTI Department of Trade and Industry
EDP Electronic Data Processing
EIA Energy Information Administration
EPSRC Engineering and Physical Sciences Research Council
EU European Union
GDP Gross Domestic Product
GHG greenhouse gas
IEA International Energy Agency
IESNA Illuminating Engineers Society of North America
IFS International Financial Statistics
IMF International Monetary Fund
IPCC Intergovernmental Panel on Climate Change
ITP Industrial Technologies Program
LCA Life Cycle Analysis
LED Light Emitting Diodes
LFW Laminar Flying Wing
NASA National Aeronautics and Space Administration
OECD Organisation for Economic Co-operation and Development
PVGIS Photovoltaic Geographical Information System
PPP Purchasing Power Parities
198
SEC Specific Energy Consumption
SWA Swept Winged Aircraft
UDF Unducted Fan
UK United Kingdom
UN United Nations
UNDP Untied Nations Development Programme
UNFCCC United Nations Framework Convention on Climate Change
USA United States of America
USDOE United States Department of Energy
WEC World Energy Council
WRI World Resources Institute