Energy security under de-carbonization scenarios: An assessment
framework and evaluation under different technology and
policy choices
Jessica Jewell a,b,n
, Aleh Cherp b,c
, Keywan Riahi a,d
a
International Institute for Applied Systems Analysis, Laxenburg, Austria
b
Central European University, Budapest, Hungary
c
Lund University, Lund, Sweden
d
Graz University of Technology, Graz, Austria
H I G H L I G H T S
 We explore energy security implications of long-term energy decarbonization scenarios.
 We define energy security as low vulnerability of vital energy systems.
 The trade-related risks are considerably lower in decarbonization scenarios.
 Diversity of energy systems is generally higher in the first half of the century.
 Vulnerability is lowest in scenarios with both high efficiency and renewable energy constraints.
a r t i c l e i n f o
Article history:
Received 17 October 2012
Received in revised form
15 October 2013
Accepted 17 October 2013
Available online 23 November 2013
Keywords:
Energy security
Climate change
Indicators
a b s t r a c t
How would a low-carbon energy transformation affect energy security? This paper proposes a framework
to evaluate energy security under long-term energy scenarios generated by integrated assessment
models. Energy security is defined as low vulnerability of vital energy systems, delineated along
geographic and sectoral boundaries. The proposed framework considers vulnerability as a combination
of risks associated with inter-regional energy trade and resilience reflected in energy intensity and
diversity of energy sources and technologies. We apply this framework to 43 scenarios generated by the
MESSAGE model as part of the Global Energy Assessment, including one baseline scenario and 42 ‘lowcarbon’
scenarios where the global mean temperature increase is limited to 21C over the pre-industrial
level. By and large, low-carbon scenarios are associated with lower energy trade and higher diversity of
energy options, especially in the transport sector. A few risks do emerge under low-carbon scenarios in
the latter half of the century. They include potentially high trade in natural gas and hydrogen and low
diversity of electricity sources. Trade is typically lower in scenarios which emphasize demand-side
policies as well as non-tradable energy sources (nuclear and renewables) while diversity is higher in
scenarios which limit the penetration of intermittent renewables.
& 2013 Elsevier Ltd. All rights reserved.
1. Introduction
A radical transformation of energy systems is required to
reduce greenhouse gas emissions and avoid long-term consequences
from global climate change. However, policy makers are
typically more concerned with immediate (rather than long-term)
and national (rather than global) effects of energy policies. One
such immediate national issue is energy security. Thus, understanding
energy security implications of climate mitigation
policies is critically important for anticipating the degree of
political support they are likely to command.
There are three main challenges to characterizing the energy
security of low-carbon energy futures. First, there are scholarly
disagreements on the meaning of and the ways to measure energy
security. For example, there are debates on whether energy
security includes economic, environmental and social considera-
tions.1
Other disagreements are over the most appropriate scale
(national, regional, local, etc.) of analyzing energy security, the
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/enpol
Energy Policy
0301-4215/$ - see front matter & 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.enpol.2013.10.051
n
Corresponding author at: Energy Group, International Institute for Applied
Systems Analysis, Schlossplatz 1, Laxenburg, Austria. Tel.: þ43 2236 807 445.
E-mail address: jewell@iiasa.ac.at (J. Jewell).
1
For those scholars who consider environmental impacts of energy systems a
“dimension“ of energy security (Sovacool and Brown, 2010) the very question of
the relationship between climate and energy security goals does not make sense,
since in their view these goals are identical.
Energy Policy 65 (2014) 743–760
extent to which energy security is a generic or context-dependent
concept, the relative importance of various risks (geopolitical,
technological, natural, economic) and the most appropriate methods
of assessing energy security.
Second, even the existing academic and policy consensus on
what energy security is and how it can be evaluated is not always
possible to extend into long-term future scenarios. Most existing
studies project present energy security concerns such as oil and
gas trade and resource scarcity into the future (e.g. Turton and
Barreto, 2006, Rozenberg et al., 2010; Costantini et al., 2007;
Bollen et al., 2010). While these studies provide useful insights,
they do not account for the fact that if energy systems undergo
radical transformations (for example, if oil is no longer the
dominant fuel in the transport sector), new energy security
concerns (such as trade in biofuels) may replace current ones.
Other studies provide a more generic approach to evaluating
future energy security based on overall net import dependency
(McCollum et al., 2011) or import dependency and diversity
combined into a single indicator (McCollum et al., 2013). However,
there is little evidence that real-life energy security policies are
guided by such highly aggregated and generic indicators. Thus, an
appropriate method to assess energy security implications of longterm
climate policies should be both reflective of policy concerns
and suitable for future energy systems that may be radically
different from present ones.
Third, assessing long-term energy security requires a concrete,
preferably quantitative, representation of a future, or a range of
potential futures. Over the past several decades the development
of Integrated Assessment Models (IAMs) (Leimbach et al., 2009;
Rao and Riahi, 2006; van Vuuren et al., 2011; Manne and Richels,
2004; Bosetti et al., 2011) which present detailed quantitative
descriptions of low-carbon futures has made this possible.
The purpose of this paper is to develop and apply a method for
assessing energy security implications of low-carbon energy
futures under different policy and technology choices. It overcomes
the three limitations of present energy security studies by:
(a) formulating a coherent concept of energy security which both
accurately reflects historic and current energy security policy
concerns and yet is sufficiently generic to be applicable to
energy systems which are radically different from present
ones (Sections 2.1 and 2.2);
(b) translating this concept into a framework for assessing energy
security under radical transformations of energy systems
(Section 2.3);
(c) applying this assessment framework to the energy decarbonization
pathways (described in Section 3) developed within
the Global Energy Assessment (GEA, 2012) to assess energy
security under various decarbonization scenarios (Section 4
presents the results Section 5 the discussion and Section 6
concludes with the policy implications).
2. Framework and indicators for evaluating future energy
security
For the purposes of this analysis we define energy security as
‘low vulnerability of vital energy systems’. In line with the Global
Energy Assessment (GEA) (Cherp et al., 2012) and other mainstream
definitions of energy security (for an overview see Winzer
(2012)), this definition is sufficiently flexible to be applicable in
diverse situations, including in future energy systems which may
be very different from present ones. Evaluating energy security in
accordance with this definition involves (1) identifying vital
energy systems including those which may emerge under future
scenarios; (2) identifying vulnerabilities of such systems; and
(3) developing, applying and interpreting indicators to characterize
these vulnerabilities.
2.1. Vital energy systems
Energy security is about protecting energy systems whose
failure may disrupt the functioning and stability of a society. Such
vital energy systems can be defined in terms of their geographic
boundaries (national, sub-national, regional or the world as a
whole) or in terms of their sectoral boundaries (a primary energy
source such as crude oil, an energy carrier such as electricity or an
energy end-use such as transportation). Different combinations of
geographic and sectoral boundaries yield a potentially large
number of vital energy systems (e.g. “the global oil market“, “the
European electricity network” or “transportation in China“) each
of which can be the subject of an energy security assessment.
With respect to geographic boundaries, the current and historic
focus of energy security policies has been national. This is logical,
because historically nation states have been responsible for
security in all areas and most energy policies are developed and
implemented at the national level. At the same time, many
contemporary energy security policies focus on regional or global
energy systems rather than merely national ones. For example, the
European Union's (EU) energy security policies address electricity
systems in the EU and their integration with neighboring countries
(European Parliament, 2006) as well as the Eurasian and global
natural gas markets (European Union Council, 2004). Regional and
global energy markets are also considered in energy security
policies and policy-driven assessments in the UK (Wicks, 2009),
Japan (Pant, 2006; Atsumi, 2007) and Australia (Australian
Government Department of Resources Energy and Tourism, 2011,
2009). Concerns about the global oil market are clear from the
presence and policies of international organizations such as the
IEA and OPEC.
National, regional and global energy systems are likely to remain
relevant to energy security in the future although their relative
importance may change depending on the dynamics of energy
trade and dependence. As explained in the next section, Integrated
Assessment Models (IAMs) typically provide regional and global
rather than national level resolution which restricts our energy
security analysis to these two levels. However, the proposed framework
can also be used for the analysis of national energy security if
relevant data are available (for example, Jewell et al. (forthcoming)
apply this framework to China, the EU, India and the US–major
economies the size of global regions).
With respect to energy sectors, energy security studies typically
focus on ‘security of supply’ comprised of primary energy
sources. In particular, there is extensive literature on measuring
security of oil supplies (see for example Gupta, 2008 and Greene,
2010). The IEA's Model of Short-term Energy Security (MOSES)
evaluates oil, natural gas, coal, biomass, nuclear and hydropower
supply as well as four energy carriers (biofuels and three types of
oil products) (IEA, 2011; Jewell, 2011). There are a number of
energy security studies which focus on electricity (Stirling, 1994;
Grubb et al., 2006). Finally, a few studies focus on security of
energy end-uses, sometimes called ‘energy services security’
(Jansen and Seebregts, 2009).
Projecting energy sectors into the future is less straightforward
than projecting geographic boundaries of vital energy systems.
In particular, key primary energy sources and energy carriers can
change under radical energy transitions. For example, while oil lies
at the heart of today's energy security concerns, over the longterm
natural gas, electricity or biomass production could become
central to ensuring energy security. Liquid energy carriers which
today are mostly oil products could be replaced by biofuels,
J. Jewell et al. / Energy Policy 65 (2014) 743–760744
synthetic fuels2
, hydrogen or electricity. However, end-use sectors
– transportation, industrial and residential & commercial – are
unlikely to change in nature although their relative size and
importance could. Thus, to evaluate future energy security we
use generic categories of energy sources, energy carriers and
energy end-uses. Table 1 compares these systems to the vital
energy systems used in the analysis of present-day energy security
in GEA (Cherp et al., 2012).
2.2. Vulnerabilities
The second step in constructing an energy security assessment
framework is defining vulnerabilities of vital energy systems.
As in the case of vital energy systems, vulnerabilities should
be defined specifically enough to echo current and historical
energy security concerns, yet generically enough to be applicable
to future energy systems potentially very different from present
ones.
Vulnerabilities of an energy system are a combination of its
exposure to risks and resilience, i.e. its capacity to respond to
disruptions. Some energy security assessments focus primarily on
risks (e.g. APERC, 2007; Winzer, 2012), others focus primarily on
resilience (e.g. Stirling, 1994, 2010) and some look at both risks and
resilience (Kendell, 1998; Gupta, 2008; Jewell, 2011). Vulnerabilities
may be physical (disruptions of energy flows) or economic (disruptive
variations in energy prices and costs (see Keppler, 2007; Greene,
2010; Helm, 2002)) and may come in the form of shocks (rapidly
unfolding short-term disruptions) or stresses (slowly approaching and
longer-lasting phenomena) (Stirling, 2010; Cherp and Jewell, 2013).
The distinction between risks and resilience, physical and
economic disruptions and shocks and stresses is relatively common
in the literature, however, there is less agreement concerning
more specific classification of vulnerabilities. The most widely
proposed ‘dimensions’ of energy security cannot be used for the
analysis of future energy security because of at least one of the
following reasons:
1. They lack solid theoretic foundations and thus their use is limited
to illustrating the current rather than conceptualizing potential
future vulnerabilities. For example, the widely cited “four A's of
energy security” – “Affordability, Availability, Accessibility, and
Acceptability” – which are similar to the “five A's of health care”
(Penchansky and Thomas, 1981), cannot be meaningfully used for
assessing the vulnerability of energy carriers, energy end-uses or
primary energy sources other than fossil fuels whose role in future
energy systems may not be as important as today.
2. They are open to a wide variety of interpretations which makes
quantification difficult. For example, the concept of ‘affordability’
(as well as a variety of related terms such as ‘fair’ or
‘reasonable’ prices) is for the most part used rhetorically and
can be interpreted to mean stability of prices, competitiveness,
low prices or even protection from energy poverty (Cherp and
Jewell, 2013).3
3. They are too narrow and/or too data-intensive to be used for
generic quantitative evaluations either in present-day or in
future energy systems. This relates, for example to many of the
over 300 indicators (ranging from “energy literacy of users” to
“annual volume of sales from woodlots”) for 20 dimensions of
energy security proposed by Sovacool and Mukherjee (2011).
In this article we use a more universal way of structuring
vulnerabilities which is based on generic ‘perspectives’ of energy
security that have emerged over the last century (Table 2).
All three perspectives are likely to be relevant for analyzing
energy security in the long-term although their nature may change
as a result of low-carbon energy systems transformations. The
sovereignty perspective has been around for over 100 years and is
likely to persist in the future unless all types of polities and
conflicting interests dissolve. At the same time, shifts in trade
affecting national interdependencies and energy power balances
would affect this perspective. Robustness concerns are not likely to
subside but rather intensify as energy systems become more
advanced, dynamic and integrated. Resilience is the most generic
perspective since it does not depend on specific configurations of
energy systems but rather reflects generic concerns arising from their
exposure to complex and uncertain factors.
Shifting trade patterns, along with a growing reliance on
renewable energy sources, would profoundly impact the geography
of energy systems (network density, number and nature of
connections and size of installations), which is at the heart of the
robustness and resilience perspective. In some ways this transition
would alleviate current concerns by, for example, avoiding choke
points present in oil trade (Lehman Brothers, 2008). While a shift
to renewables would mean increased deployment flexibility, both
Table 1
Vital energy systems used for evaluating energy security at present and for future energy scenarios.
Geographic boundaries Sectoral boundaries
Energy sources Energy carriers Energy end-uses
Presentn
Sub-national, national, regional, global Oil, natural gas, hydropower,
nuclear, biomass, renewable
energy sources (RES)
Oil products, biofuels,
electricity
Transportation, industry,
buildings, exports
Future Nationalnn
, regional, global Oil, natural gas,
hydropowernn
, nuclearnn
,
biomassnn
, RESnn
Oil products, synthetic
fuels, hydrogen, electricity,
biofuels
Transportation, industry,
residential &
commercial, exportsnn
Notes:
n
As used in GEA (Cherp et al., 2012).
nn
Show energy systems which can potentially be evaluated but are not evaluated in this paper.
2
Refers to liquefied coal and natural gas.
3
The IEA remarks that “Energy insecurity stems from the welfare impact of
either the physical unavailability of energy, or prices that are not competitive or
overly volatile” (Lefèvre, 2007, 12). Extremely low energy prices are in many ways
just as dangerous as high prices since they can lead to under-investment in
resource extraction or infrastructure (Alhajji, 2008) as most recently evidenced by
an electricity shortage in China during the summer of 2011 following caps on
electricity prices.
J. Jewell et al. / Energy Policy 65 (2014) 743–760 745
in terms of location and size of generation, the hyper-scalability of
renewable energy sources (Walker and Cass, 2007) also has a
downside, potentially resulting in the emergence of trade patterns
similar to those of oil or natural gas today.
Although the energy security assessment framework we propose
is in principle sufficiently flexible to address all these evolving
concerns, we could not quantitatively evaluate some of them in this
article. This is due to limitations of the integrated assessment model
(MESSAGE) that generated the scenario data for our analysis. Similar
to most other IAMs, MESSAGE models energy mixes in and energy
trade between 11 world regions but not the specific topography of
energy infrastructure or the specific geography of energy trade. In
particular, exact spatial locations of energy infrastructure are not
modeled and neither are bilateral trade flows. Instead there is a
global pool from or to which regions can buy or sell energy.4
Thus,
many interesting and important energy security concerns associated
with spatial characteristics of energy flows and infrastructure
could not be addressed by this analysis.
2.3. Indicators
Quantitative evaluations of energy security are used to compare
energy security of different countries (Gupta, 2008;
Gnansounou, 2008; International Energy Agency, 2011; Le Coq
and Paltseva, 2009), or plot the evolution of energy security over
time (Lefèvre, 2007; Löschel et al., 2010; Sovacool and Brown,
2010) including analyzing future energy security (Turton and
Barreto, 2006; Costantini et al., 2007). All such evaluations use
indicators: quantitative proxies of vulnerabilities of energy systems.
Hundreds of energy security indicators have been proposed
in dozens of scholarly articles and policy papers, but only a small
number of them can be used for evaluating energy security under
long-term energy scenarios. Indicators for evaluating future
energy security should meet the following criteria:
1. They should be policy relevant to current and/or historical
energy security concerns;
2. They should be sufficiently generic to be applicable to energy
systems which are radically different from present ones;
3. They should be possible to calculate from available and meaningful
data in the model or scenario which is being used to
represent the future;
4. They should provide information which is additional to that
provided by other indicators;
5. They should reflect key vulnerabilities of vital energy systems
and clarify policy trade-offs.
Criteria (1) and (2) were already discussed in Section 2.2 and
can be illustrated by our choice of indicators for the sovereignty
perspective. Import dependency is commonly cited as a major
driver of energy security policies and rhetoric (Kuzemko, 2011;
Greene, 2010). The global proxy of this measure is interregional
energy trade5
(expressed as absolute volume and relative to the
total primary energy supply (TPES)). The other sovereignty indicator
we use is the geographic concentration of exports of a
particular fuel or carrier as measured by the diversity of exporting
regions contributing to the tradable share of an energy commodity
(Lefèvre, 2010; Costantini et al., 2007). This reflects the current
energy security concern associated with oil that it is only produced
in a small number of countries and regions. To meet criteria
(2) and (5), we apply these two indicators not only to oil and gas,
which dominate energy trade today, but also to coal and “newfuels”
(hydrogen, biofuels and synthetic fuels) which could
become important in the future.
The resilience perspective is particularly relevant to criterion
(2) since the future is associated with many complexities and
uncertainties. The most commonly used indicator for resilience is
the Shannon-Wiener diversity index which was first applied to
electricity systems (Stirling, 1994) and has been subsequently used
in Jansen et al. (2004), O'Leary et al. (2007) and many other
studies. Finally, energy intensity is a commonly used indicator of
an economy's ability to deal with both physical and economic
shocks and stresses (Gnansounou, 2008; Cherp et al., 2012).
A note should also be made about the indicators which we
excluded because they did not meet one or more of the five
criteria. Of all robustness indicators, very few can be meaningfully
estimated in IAMs (i.e. they do not meet criterion 3). This is
because most infrastructural attributes of energy systems are
either not represented or are endogenously optimized (for example
the replacement of power plants follows planned retirement
ages which renders meaningless a key robustness concern such as
ageing). Resource extraction compared with known reserves and
resources is a robustness indicator which can in principle be
analyzed in IAMs (Turton and Barreto, 2006; Kruyt et al., 2009).
Due to space limitations we do not include this indicator in the
present analysis but instead refer the reader to several related
Table 2
Three perspectives on energy security.
Source: Summarized from Cherp and Jewell (2011).
Perspective Sovereignty Robustness Resilience
Historic roots War-time oil supplies and the 1970s
oil crises
Electricity blackouts and concerns
about resource scarcity
Liberalization of energy systems
Key risks for energy systems Intentional actions by malevolent
actors
Predictable natural and technical
factors
Diverse and partially unpredictable
factors
Primary protection mechanisms Control over energy systems and
institutional arrangements to prevent
disruptive actions
Upgrading infrastructure and
switching to more abundant
resources
Increasing the ability to withstand
and recover from various
disruptions
Parent discipline Security studies, international
relations, political science
Engineering, natural science Economics, complex system science
4
Except with natural gas trade for which bilateral trade is depicted for certain
regions.
5
In many instances interregional energy trade represents realistic energy
security concerns (e.g. EU and China energy imports). In some other cases more
granular representation of energy trade between nations rather than simply
between world regions would be preferable. However, energy trade between
individual nations so far cannot be modeled in long-term energy scenarios.
J. Jewell et al. / Energy Policy 65 (2014) 743–760746
studies exploring similar sets of de-carbonization scenarios (Jewell
et al., forthcoming; McCollum et al., in press). Finally, some
indicators were excluded because they did not meet criteria 4;
for example costs of imports and import dependence of specific
end-use sectors did not provide information additional to the
more generic import dependence indicator.
3. Methods and scenarios
We use the indicators developed in Section 2 and summarized in
Table 3 to evaluate energy security under a set of de-carbonization
scenarios generated by the MESSAGE IAM (Rao and Riahi, 2006;
Messner and Strubegger, 1995; Riahi et al., 2007) in the framework of
the GEA (Riahi et al., 2012). We compare the energy security under
these transformational scenarios – further referred to as “low-carbon
scenarios” –to the energy security under a Baseline (counterfactual)
scenario also from the GEA. Since these scenarios provide detailed
quantification of future developments of the energy system for various
energy supply and demand-side configurations, we are able to apply
each indicator to either the energy system as a whole (for example we
measure the energy intensity for the economy as a whole) or a set of
energy subsystems (for example, regional diversity of fuel or carrier
exports for each globally traded fuel or carrier—seven in all: oil, gas,
coal, hydrogen, electricity, synthetic fuels and biofuels). This application
of the framework illustrates how the objective of measuring
energy security can be combined with the concept of vital energy
systems in order to provide a rigorous and robust evaluation of energy
security under de-carbonization scenarios.
In all low-carbon scenarios the increase in the global mean
temperature is stabilized with 50% probability to 21C above preindustrial
levels by 2100 under medium GDP and population
growth projections (in other words GHG concentration is stabilized
at 450 ppme). This requires massive changes in both supplyand
demand-side energy technologies so that the greenhouse gas
(GHG) emissions from energy systems decline over the 21st
century in stark contrast to the Baseline scenario as shown in
Fig. 1. The exact nature of these changes varies among the lowcarbon
scenarios depending on the supply- and demand-side
configurations as described below.6
There are three dimensions of technological and policy choices
which potentially affect energy security in the low-carbon scenarios.
The first dimension concerns energy demand, where the lowcarbon
scenarios fall into three groups:
 Efficiency scenarios where the focus of policy and investment is
on energy efficiency improvements resulting in significantly
suppressed overall energy demand due to lower energy
intensity;
 Supply scenarios where policy and investments are focused on
low-carbon energy supply technologies resulting in more rapid
transformation of the energy mix and relatively fast growth in
energy demand;
 Mix where equal focus is given to supply- and demand-side
policies and investments.
Fig. 2 shows energy intensity in the three groups of scenarios.
Under a given GDP assumption, higher energy intensity translates
into higher demand while lower intensity translates into lower
demand.
The second dimension of choices affecting energy security
is constraints imposed on supply-side technologies in selected
scenarios, namely:
 Limited renewable energy sources (limitRES) scenarios where
intermittent solar and wind energies make up no more than
20% of final energy consumption;
 Limited bioenergy (limitBE) scenarios with bioenergy limited to
no more than 50% of the estimated global potential;
Table 3
Indicators of long-term energy security.
Energy systems Perspectives
Sovereignty Resilience
Primary energy sources Total Primary Energy Supply (TPES) Global energy trade (absolute and relative
to the TPES) and Net import dependence*
Diversity of TPES and Energy intensity
Oil, gas, coal Global fuel trade
Regional diversity of fuel exports
Carriers Hydrogen, synthetic fuels,
electricity, biofuels
Global trade in carrier Diversity of primary energy sources
used in carrier production
Regional diversity of carrier exports
End-use sectors Transport, industry, residential
& commercial
Diversity of primary energy sources
used in end-use sector
Notes: All indicator formulas are presented in the Appendix along with additional indicators which could be used in another study. All resilience indicators can be applied at
both the global and regional level. Most sovereignty indicators can be applied at the global level except for net import dependence (marked with *
) which is applied at the
regional level.
Fig. 1. Annual GHG emissions in the Baseline and low-carbon scenarios.
6
This article highlights the main characteristics of low-carbon scenarios with a
focus on energy-system changes which are particularly relevant to energy security.
More extensive documentation can be found in the GEA report (Riahi et al., 2012)
and the GEA web-database (http://www.iiasa.ac.at/web-apps/ene/geadb).
J. Jewell et al. / Energy Policy 65 (2014) 743–760 747
 Nuclear phaseout (noNUC) scenarios where no additional
nuclear capacity is built after 2020 and all nuclear power is
phased out by 2060;7
 No carbon capture and storage (noCCS) scenario with no
development of carbon capture and storage (CCS);
 No bioenergy CCS (noBECCS) scenarios where CCS technologies
are not applied in conjunction with biomass
combustion;
 No carbon sinks beyond the baseline scenarios where additional
(non-energy) carbon sinks are not created.
The third dimension of choices within the low-carbon
scenarios concerns the configuration of transport systems,
namely:
 Conventional transportation (CTR) scenarios with transport
systems relying primarily on liquid fuels;
 Advanced transportation (ATR) scenarios with transport systems
increasingly relying on electric and hydrogen propulsion
of vehicles.
Not all combinations of demand, supply and transport constraints
are present among low-carbon scenarios. “Efficiency”
scenarios allow for climate goals to be reached with a broader
range of supply-side constraints (e.g. a combination of limitRES
and limitBE or noNUC and noCCS). “Supply” scenarios allow for
relatively few supply-side constraints (e.g. nuclear phaseout combined
with no development of CCS is not possible). Similarly,
“ATR” allows for more flexibility in the energy system and can
meet 21C stabilization under a wider range of energy supply
restrictions. The full list of scenarios is presented in Table 4.
Different levels of energy demand and alternative assumptions
about possible restrictions for supply-side technologies have
major implications for the future portfolio of energy options.
The GEA scenarios depict many possible evolutions of the energy
system, exploring alternative routes of low-carbon energy
transitions. Some scenarios are for example characterized by a
relatively high contribution of renewables while others emphasize
carbon capture and storage or nuclear energy. Energy technologies
in the transport sector are also varied ranging from advanced
electrification to continuous reliance on liquid fuels. Primary
energy portfolios of the low-carbon scenarios for which we
conduct our energy security analysis are shown in Fig. 3. For a
more detailed discussion of the GEA scenarios, see Riahi et al.,
(2012).
Table 4
Low-carbon scenarios of energy transitions analyzed in this article.
Supply Mix Efficiency
Advanced
transport
Conventional
transport
Advanced
transport
Conventional
transport
Advanced transport Conventional transport
Full portfolio of supply options SupplyATR
Full
SupplyCTR Full Mix ATR Full MixCTR Full EfficiencyATR Full EfficiencyCTR Full
Limited renewable energy sources SupplyATR
limitRES
– MixATR
limitRES
MixCTR limitRES EfficiencyATR limitRES EfficiencyCTR limitRES
Limited bioenergy Supply ATR
limitBE
– MixATR
limitBE
MixCTR limitBE EfficiencyATR limitBE EfficiencyCTR limitBE
Limited RES & Limited bioenergy – – – – EfficiencyATR limitRES &
limitBE
EfficiencyCTR limitRES &
limitBE
Nuclear phaseout (noNUC) SupplyATR
noNUC
SupplyCTR
noNUC
MixATR
noNUC
MixCTR noNUS EfficiencyATR noNUC EfficiencyCTR noNUC
No carbon capture and storage – – MixATR
noCCS
MixCTR noCCS EfficiencyATR noCCS EfficiencyATR noCCS
Nuclear phaseout & No carbon
capture and storage
– – – – EfficiencyATR noNUC & noCCS EfficiencyCTR noNUC &
noCCS
No bioenergy CCSn
SupplyATR
noBECCS
– MixATR
noBECCS
– EfficiencyATR noBECCS EfficiencyCTR noBECCS
No additional carbon sinks beyond
the baselinen
SupplyATR
noSinks
– MixATR
noSinks
MixCTR noSinks EfficiencyATR noSinks EfficiencyCTR noSinks
No bioCCS & No sinks & Limited
BEn
– – – – EfficiencyATR noBECCS &
noSinks & limitBE
EfficiencyCTR noBECCS &
noSinks & limitBE
Note:
n
These type of constraints had only a small effect on energy security and while included in the analysis are not specifically mentioned in the paper. Cells marked
with “-” denote scenarios where the low-carbon energy transformation was found infeasible under the given combination of energy demand and supply-side
restrictions.
Fig. 2. Energy intensity in the Baseline and low-carbon scenarios.
7
This assumes a 40-year life-span for nuclear power plants.
J. Jewell et al. / Energy Policy 65 (2014) 743–760748
Fig. 3. Composition of the global TPES in 2005 and 2050 for low-carbon GEA scenarios. Note: figure modified from Riahi et al., (2012). ‘X’s indicate infeasible pathways.
J. Jewell et al. / Energy Policy 65 (2014) 743–760 749
4. Results
This section presents the results of assessing energy security in
the GEA energy scenarios listed in Table 4 using the indicators
listed in Table 3.
4.1. Sovereignty
4.1.1. Global energy trade
In the Baseline scenario, with a higher level of demand and a
high reliance on fossil fuels, global energy trade rises dramatically
from the current 100 EJ/year to over 400 EJ/year by 2100. The
levels of trade in the low-carbon scenarios are much lower,
ranging from 40 EJ/year to 240 EJ/year by 2100. Trade initially
rises in all low-carbon scenarios and declines in the second half of
the century in certain Efficiency and Mix scenarios (Fig. 4). The
lower level of trade in low-carbon scenarios is explained by
(a) generally lower energy supply and use (especially in the
Efficiency scenarios) and (b) a higher share of non-tradable
energies (renewables and nuclear)8
in the energy mix.
The impact of technology and policy choices on the levels of
global energy trade is illustrated in Fig. 5. In general, the volume of
trade correlates with the overall level of energy demand: under
other equal assumptions, trade in Supply scenarios is higher than
in Mix scenarios which is in turn higher than in Efficiency
scenarios since higher overall demand increases the demand for
tradable fuels.
In all Supply scenarios, energy trade increases to about one and
a half times current levels by 2030. In Supply scenarios with no
nuclear or with limited renewables trade continues to rise for the
rest of the century. This is because when domestic sources
(nuclear energy and renewables) are limited, the energy system
is forced to use more globally-traded fuels. In all other supply
scenarios, where the share of non-tradable energies (nuclear and
renewables) in the energy mix is much higher, energy trade
plateus at $100–140 EJ/year by 2100.
In all Mix scenarios, global energy trade rises to $120–130 EJ/
year by 2030. Subsequently the highest trade is in scenarios in
which conventional transport is combined with limitations on
renewables; these constraints require more tradable fuels in the
energy system since the transport system continues to be dependent
on liquid fuels and there are limitations on domestic sources.
In contrast, the lowest trade is under advanced transport with no
limitations on nuclear or renewables, especially combined with
limitations on CCS. All of these supply choices lead to higher
shares of non-tradable sources and electricity as a carrier in the
energy mix.
In the majority of Efficiency scenarios, the initial moderate rise
in energy trade is followed by a decline below the current levels by
the end of the century. In Efficiency scenarios with limited
renewables, energy trade does not decline and when limited
renewables are combined with conventional transport and limited
bioenergy the trade actually rises by the end of the century to
about two and a half times the current level because of continued
dependence on traded energy.
In summary, the higher the demand, the more easily the rise of
energy trade is triggered by additional constraints:
 In Supply scenarios, higher trade is triggered by limitations on
renewables or nuclear energy;
 In Mix scenarios, higher trade is triggered by limitations on
renewables combined with conventional transport;
 In Efficiency scenarios, higher trade is triggered by limitations
on RES and bioenergy combined with conventional transport.
Energy trade intensity (shown in Fig. 6) rises in the Baseline
scenario from the current 20% to 25% by 2030 before leveling off at
$20%. In contrast, in all low-carbon scenarios, trade intensity
peaks at a lower level and declines after 2030. Unlike trade
volumes, trade intensity does not notably vary across Supply,
Mix and Efficiency scenarios: though Efficiency scenarios are
generally associated with lower trade volumes the overall energy
demand is also lower which results in similar trade intensity to
Supply and Mix scenarios.
At the same time, trade intensity is affected by supply-side
constraints. In scenarios with no limitations on renewables, trade
intensity declines to 1–10% by the end of the century. When
renewables are limited, this decline is less pronounced (11–15% by
the end of the century) since the world is pushed to using more
tradable fuels. If limiting renewables is combined with conventional
transport and limited bioenergy, the trade intensity of the
low-carbon scenario is only marginally lower than that observed
in the Baseline since the transport system continues to be
dominated by liquids but is unable to take full advantage of
domestic biofuels.
4.1.2. Regional energy balances
A detailed analysis of regional energy security in low-carbon
scenarios is beyond the scope of this article. Instead we present
selected data to illustrate how the global picture may be reflected
at the regional level.
Though in general regional import dependencies follow global
trends they are also influenced by a region's resource availability
and pace of economic development. Fig. 7 shows the import
dependency of Western Europe and South Asia and the exports
of the Middle East and North Africa region. Import dependency of
both importing regions is lower than in the Baseline and is
generally higher in scenarios with limited renewables since for
both of these regions, their main domestic energy source is
renewables. At the same time, the import dependency of Western
Europe either declines or stays similar to the current level whereas
in South Asia it initially peaks and in some scenarios stays above
the current levels. While net energy exports from the Middle East
and North Africa dramatically fall in all low-carbon scenarios and
in the Baseline, the annual export volumes from this region
initially rise before leveling off at current levels in the latter half
of the century.
4.1.3. Trade in individual fuels
Fig. 8 illustrates the trade in fossil fuels, which currently
makes up the bulk of the global energy trade. The most striking
difference between the Baseline and low-carbon scenarios is in
relation to oil trade. Whereas in the Baseline scenario, oil trade
steadily rises and more than doubles by the end of the century, in
low-carbon scenarios it peaks around 2030 and then rapidly
declines as oil is phased out of the energy system in order to de-
carbonize.
Natural gas trade rises in both the Baseline and low-carbon
scenarios in the first half of the century. In the second half of the
century, the trade in the Baseline continues to rise reaching over
8
Nuclear energy is generated from uranium resources and enriched fuel, both
of which are traded, however these were excluded for both theoretical and
practical reasons. Since refueling a nuclear power plant typically provides fuel for
two to three years (Nelson and Sprecher, 2008) and nuclear fuel can even be
stockpiled for up to ten years (IAEA, 2007), most countries use nuclear energy as a
way to curtail risks from imported fossil fuels. Practically, MESSAGE does not depict
trade in either raw uranium and to the best of our knowledge no global IAM depicts
trade in either enriched nuclear fuel or nuclear power plant components, which are
the biggest energy security issues for nuclear power (Cherp et al., 2012).
J. Jewell et al. / Energy Policy 65 (2014) 743–760750
100 EJ/year (more than oil trade at present). At the same time the
low-carbon scenarios diverge falling roughly into three groups:
(a) In one Supply and one Mix scenario with limitations on
renewables, gas trade increases to levels comparable to the
Baseline and exceeding present-day oil trade volumes. In these
scenarios, with limited renewables, natural gas continues to be
a critical part of the energy system until the end of the century
(marked with dark gray lines in Fig. 8).
(b) In several scenarios, gas trade plateaus (with some gradual
growth or decline) at levels below the present volumes of oil
trade and below the Baseline. These scenarios are: Supply
combined with nuclear phase-out (leading to gas being used in
electricity); Supply or Mix combined with conventional transport
(leading to the use of gas-to-liquids in transport); Mix or
Efficiency combined with limited renewables (leading to a lack
of alternatives to gas); and Efficiency combined with conventional
transport and limited bioenergy (leading to gas liquids
being used in transportation instead of biofuels).
(c) In other scenarios gas trade significantly declines in the latter
half of the century. These include: the most advanced transport
scenarios (where there is not a limitation on renewables or
nuclear energy); and Efficiency scenarios with conventional
transport and no limitations on bioenergy. In these scenarios,
gas serves the role of a bridge fuel, being gradually replaced by
other energy sources towards the end of the century.
In the Baseline scenario global coal trade rises from its current
10 EJ/year to over 90 EJ/year by 2100. Coal trade in low-carbon
scenarios varies depending on supply and demand constraints. In
scenarios with limited CCS the use of coal is not compatible with
GHG limitations so coal trade virtually disappears. Coal trade is
higher in scenarios with limited renewables and nuclear (when
combined with Mix or Supply) where it is used in combination
with CCS to provide an alternative to electricity generation.
In addition to traditionally traded fossil fuels, some scenarios
include significant trade in “new” fuels and carriers: biofuels,
synthetic fossil fuels, and hydrogen (Fig. 9). In the Baseline
scenario the trade in biofuels rises after 2040 to ca 20 EJ/year by
the end of the century. In low-carbon scenarios, trade in biofuels
increases to comparable levels (earlier in the century), but less so
in scenarios where the production of bioenergy is limited since
this in turn limits the extent of biofuel use. In all scenarios the
levels of trade in biofuels are two to ten times lower than the
volumes of oil trade at present. The trade in synthetic fuels (liquids
produced from coal or gas) in the Baseline scenario rises over 40
EJ/year but stays below 12 EJ/year in all low-carbon scenarios.
In contrast to synthetic fuels, hydrogen trade is present in some
low-carbon scenarios, but not in the Baseline scenario. Towards the
end of the century, trade in hydrogen rises to levels comparable to oil
trade today in Supply scenarios with advanced transport or a
phaseout of nuclear energy. For the advanced transport scenarios
this is because these scenarios assume a higher potential for fuel cell
technologies which when combined with high demand drives up
hydrogen trade. For the nuclear phaseout scenarios, this is because
limitations on nuclear energy limit the number of regions where it is
economically feasible to produce hydrogen, for which there is high
demand around the world.
Fig. 6. Global trade intensity.
Fig. 5. Global energy trade and choices within the low-carbon scenarios.
Fig. 4. Global energy trade in the Baseline and low-carbon scenarios.
J. Jewell et al. / Energy Policy 65 (2014) 743–760 751
High volumes of trade may be especially risky if the fuel in
question is primarily produced in a limited number of regions. Fig. 10
illustrates the geographic concentration of exports of tradable gas,
coal and hydrogen (i.e. the fuels which are highly traded in some
low-carbon scenarios).
In the case of limited renewables and other scenarios associated
with higher gas trade (groups (a) and (b) in the explanation
to Fig. 8 described on page 7) natural gas is indeed produced in
fewer and fewer regions. This is because natural gas resources are
unevenly distributed and large volumes of extraction inevitably
lead to increasing geographic concentration of exports. In fact, in
these scenarios (as well as in the Baseline) gas exports may
become far more geographically concentrated than oil
exports today.
Fig. 10 also illustrates that the geographic diversity of coal
exports remains high even in scenarios with higher coal trade. The
same is true with respect to biofuels (not shown on the Figure).
This is because coal and bioenergy resources as well as the
capacity to produce hydrogen are more evenly distributed around
the world than natural gas or oil resources. The geographic
diversity of hydrogen exports remains high under most scenarios
but not all. Under supply scenarios with no nuclear development,
the geographic diversity of hydrogen exports dips to that of oil's
today. This is because the limitations on nuclear limit where it is
economically-feasible to produce hydrogen.
4.2. Resilience
4.2.1. Energy intensity
Fig. 2 illustrates that energy intensity in low-carbon declines at
a much faster rate than in the Baseline. This decline in energy
intensity means gains in energy security as economies become less
sensitive to energy price fluctuations.9
4.2.2. Diversity of primary energy supply
Fig. 11 illustrates the diversity of energy sources in the total
primary energy supply (TPES), electricity generation and the
transport sector. The diversity of TPES and electricity show largely
similar trends: in the Baseline scenario it slowly but steadily rises
whereas in low-carbon scenarios it rapidly rises between now and
2030 or 2040 and then either declines (to levels below the
Fig. 7. Net energy balance of selected regions.
Fig. 8. Global trade in fossil fuels.
9
Energy intensity is an endogenous variable in these low-carbon scenarios and
therefore its decline is essentially programmed in the model rather than being an
independent outcome of pursuing climate protection targets.
J. Jewell et al. / Energy Policy 65 (2014) 743–760752
Baseline and the present value) or stays at an elevated level
depending upon supply options. The mid-century peak in diversity
in low-carbon scenarios occurs when “old“ and “new“ energy
technologies coexist, it starts declining as the low-carbon energy
sources replace carbon-intensive ones.
In scenarios with limited penetration of renewables, the
diversity of TPES and electricity generation is comparable to
the baseline development and generally higher than today's
diversity by the end of the century. This is because with limitations
on renewables, no energy source is able to dominate the
energy mix. In contrast, in scenarios with a phaseout of nuclear
energy the diversity of electricity production declines to
significantly lower levels than both the Baseline and the
present value.
Fig. 9. Global trade in “new” fuels and carriers.
Fig. 10. Geographic diversity of supply of fuels with the highest global trade (depicted in Figs. 8 and 9).
Fig. 11. Diversity of TPES, Electricity and Transport sector.
J. Jewell et al. / Energy Policy 65 (2014) 743–760 753
The low diversity of energy sources used in transport is one of the
main energy security concerns at present (Cherp et al., 2012).
Diversity in this sector rises much more rapidly and stays higher
for most of the century in low-carbon scenarios than in the Baseline.
In other end-use sectors (industry as well as residential and
commercial) the diversity of energy sources does not change
significantly in either the Baseline or low-carbon scenarios.
In summary, the diversity of energy options in the mediumterm
(20–50 years) is higher in low-carbon scenarios than in the
Baseline. In the longer term, the diversity of some low-carbon
scenarios drops below the diversity of the Baseline because of the
dominance of renewables, particularly in scenarios where renewables
are not limited or where nuclear energy is.
4.2.3. Diversity of energy supply at the regional level
Fig. 12 shows TPES diversity in Western Europe, South Asia
(dominated by India) and Centrally-planned Asia (dominated by
China). It illustrates that whereas the three representative regions
generally repeat the global pattern (compare with Fig. 11, first
graph), there are certain regional differences. For example the risedecline
pattern is more profound in the region which is dominated
by rapidly developing China (Centrally-planned Asia). Diversity
also reaches lower levels in populous South Asia with limited
energy options.
5. Discussion
This section discusses how the long-term evolution of energy
security in low-carbon scenarios depends on policy and technology
choices.
5.1. Energy security in low-carbon scenarios
Table 5 lists selected energy security indicators in 2010, 2050
and 2100 in low-carbon scenarios and the Baseline.
By 2050, low-carbon scenarios perform better than the Baseline
with respect to all energy security indicators except natural gas trade,
which is higher in some scenarios. Especially notable are the
decrease in oil trade and the increase in diversity of energy used
for transport and electricity production.
By 2100, the picture becomes more nuanced. The overall
energy trade and oil trade are both lower in low-carbon scenarios.
However, in some low-carbon scenarios the levels of gas trade
reach the level of oil trade today and its production becomes even
more geographically concentrated than for oil today. Thus, while
oil ceases to be a major energy security issue, natural gas trade
could acquire insecure patterns resembling those of the global oil
market today. It should be noted, however, that natural gas does
not dominate any of the end-use sectors to the extent that oil
dominates the transport sector today, therefore even relatively
high trade and concentration of natural gas production would be a
lower energy security risk compared to the risks associated
with present oil trade, particularly since the overall energy use
in these high-trade scenarios would be much higher than at
present. The diversity of energy sources in electricity generation
and the overall TPES is also lower in some low-carbon scenarios
than in the Baseline and at present. Both higher trade in natural
gas and lower diversity of energy systems is associated with
certain policy and technology choices assumed in some lowcarbon
scenarios as further explored in the next sub-section.
5.2. Impact of policy and technology choices on energy security in
low-carbon scenarios
Potential long-term energy security concerns within lowcarbon
scenarios highlighted in the previous section are triggered
by different combinations of demand and supply choices
(Fig. 13):
 Higher gas and/or hydrogen trade is observed in Supply
scenarios with limited renewables or nuclear energy phase-
out;
 Lower diversity of electricity and TPES production is observed
in scenarios with unlimited renewables, particularly combined
with advanced transport and limitations on nuclear
energy.
Both effects are more pronounced in Supply and Mix scenarios,
particularly when nuclear energy is phased out. Fig. 13 shows that
only a limited number of scenarios are not located in “dangerous”
corners where either trade is very high or diversity is very low.
The relatively “secure” scenarios are Efficiency with limitations on
renewables where both high diversity and lower energy trade can
be assured simultaneously.
5.3. Compound energy security indices
This paper uses relatively straightforward methods of presenting
energy security indicators for vital energy systems including
trade-offs between different dimensions of energy security (e.g. on
Fig. 13). In many cases, however, such a direct presentation of
Fig. 12. Regional diversity of TPES.
J. Jewell et al. / Energy Policy 65 (2014) 743–760754
multiple indicators may be insufficient to communicate the results
of an energy security assessment. While it is necessary to use
multiple indicators to portray an integrated picture of energy
security, too much data can also lead to confusion, especially if
indicators tell different stories. Thus, energy security studies often
use compound indices - calculated from several different indicators
- to reduce the amount of data and increase the accessibility of
the results of an energy security assessment. However, most of the
compound indices in the literature (e.g. Gupta, 2008; Scheepers
et al., 2007) cannot be used to analyze energy security in longterm
scenarios because of the data and assumptions they use.
One notable exception is the compound index based on the
modified Shannon-Wiener diversity index which ‘penalizes’ energy
sources for being imported (see the formula in the Appendix). This
indicator was used in connection with long-term future energy
security studies, including in Riahi et al. (2012) and McCollum et al.
(2013). It was based on a more complex index originally proposed
by Jansen et al. (2004) (who also suggested penalizing sources
coming from unstable countries or scarce resources).
While this index shows the potential for aggregation, it has several
limitations. It is useful when diversity and import dependency are
within a moderate range of values and correlate with each other. In
such situations the compound diversity index reduces the number
of variables that need to be considered in the assessment and may be
useful for monetization or other such calculations. However, this
index can obscure the policy trade-offs where diversity and import
dependency tell different stories. In particular, this index fails to
account for the fuel diversity of imports. For example if a country's
energy system imports all of its energy, this compound diversity
index will always be zero regardless of the how many sources it
depends on.10
Thus, any aggregation must strike a very delicate balance between
on the one hand reducing the amount of data and on the other hand
staying true to the systems, vulnerabilities, and priorities of policy
makers. Cherp and Jewell (2013) consider situations and methods for
appropriately aggregating indicators as well as suggest alternative
approaches to making sense of multiple indicators. For the purposes
of this paper other approaches for presenting several indicators
proved to be more suitable than aggregated indices.
6. Conclusions and policy implications
Mitigating dangerous climate change requires unprecedented
policy commitment to transforming energy systems. The purpose of
Fig. 13. Diversity of electricity production, trade in gas and hydrogen in low-carbon scenarios. The lower right corner represents the most ‘secure’ situations with low trade
and high diversity, whereas the upper left corner shows the ‘danger zone’ with high trade and low diversity.
Table 5
Selected energy security indicators in 2010, 2050 and 2100 in GEA pathways and the Baseline scenario.
Indicator 2010 2050 2100
Low-carbon Baseline Low-carbon Baseline
total interregional trade 106 EJ 89–175 EJ 243 EJ 42–227 EJ 420 EJ
total interregional trade intensity 20% 10%–19% 21% 3%–18% 19%
Oil trade 82 EJ 21–71 EJ 135 EJ 0–8 EJ 179 EJ
Gas trade 9.4 EJ 28–68 EJ 47 EJ 7–98 EJ 94 EJ
Coal trade 1.1 0.8–1.1n
1.0 0.1–0.7n
0.8
Hydrogen trade 10 EJ 3–34 EJ 40 EJ 0–75 EJ 96 EJ
Geographic diversity of gas exports 1.6 1.4–1.5n
1.7 1.4–1.5n
1.6
Geographic diversity of coal exports – 0–6 EJ – 5–86 EJ 4
Geographic diversity of hydrogen exports – 1.5–1.9n
– 1.1–1.6n
0
Electricity diversity 1.5 1.6–1.9 1.5 0.9–1.8 1.6
TPES diversity 1.7 1.8–2.1 1.7 1.1–1.9 1.8
Transport diversity 0.2 1.3–2.0 0.7 1.1–1.8 1.2
n
Only reports geographic diversity of exports for scenarios with high trade in that fuel or carrier (see Fig. 10).
10
A country may rely only on imported natural gas or it may rely on imported
natural gas, coal, oil, and bioenergy. While these two situations are by common
sense drastically different, the compound diversity index does not
distinguish them.
J. Jewell et al. / Energy Policy 65 (2014) 743–760 755
this article is to evaluate one major factor that influences such
commitment: the energy security implications of de-carbonization.
To achieve this aim, we propose a systematic energy security assessment
framework that is sufficiently specific to be policy relevant and
yet sufficiently generic to be applicable to energy systems radically
different from the present ones, given all the uncertainties associated
with possible configurations of such systems.
Our energy security assessment framework is based on the fact
that policy-relevant energy security concerns focus on distinct and
concrete vital energy systems rather than vague and abstract
‘energy’ as a whole. Under different energy scenarios some of
today's vital energy systems would persist and evolve (e.g. generation
of electricity and transport energy use), some could
disappear (e.g. oil and its products) and some might emerge or
radically expand (e.g. global markets for natural gas, biofuels and
hydrogen). By evaluating vulnerabilities of both existing and
potentially new energy systems this article portrays energy futures
in the familiar light of policy-relevant energy security concerns.
We consider vulnerabilities of each of the vital energy systems
as a combination of risks associated with energy trade and
resilience represented by the diversity of energy options and
energy intensity. This approach reflects two policy perspectives
of energy security which have existed for most of the history of
modern energy systems: ‘sovereignty’—focused on the degree of
domestic control over energy systems and ‘resilience’—focused on
the capacity of energy systems to respond to disruptions. We also
note the third, ‘robustness’, perspective on energy security, but
due to data and space limitations do not evaluate the robustness of
future energy systems in this paper. We select indicators of
vulnerabilities of future vital energy systems according to their
policy relevance and data availability and apply these indicators to
over 40 energy transformation scenarios developed within the
Global Energy Assessment.
All in all, our results indicate that energy systems in low-carbon
scenarios have lower trade and higher diversity than in the
Baseline scenario. These gains are most profound in the midterm
perspective (by 2050) although most of them persist through
2100. In particular, present-day energy security risks associated
with global oil trade rapidly subside and eventually disappear
under low-carbon scenarios.
Both these results and the proposed energy security assessment
framework have several novel policy implications. First, the
finding that long-term de-carbonization has considerable energy
security benefits should bring the global energy security and
climate mitigation agendas closer together. While several previous
studies have explored the energy security implications of climate
mitigation policies for certain regions (Costantini et al., 2007;
Criqui and Mima, 2012; Shukla and Dhar, 2011) or sectors (Grubb
et al., 2006; Rozenberg et al., 2010), our paper explores and reaffirms
these earlier findings more systematically and comprehensively.
In particular, we contrast the very rapid rise of global
energy trade and import dependence in the Baseline with the
decline or stagnation of trade and imports in climate stabilization
scenarios. While previous studies almost exclusively focus on fossil
fuels (particularly oil) trade we show that future trade in all other
fuels is unlikely to cause as serious energy security concerns as oil
causes today.
Second, we identify several potential vulnerabilities of certain
energy systems in some scenarios, particularly by the end of the
century. With respect to sovereignty, this is a potentially large
trade in gas and hydrogen11
and with respect to resilience this is
low diversity of electricity generation options. We show that
policy focus on energy efficiency and on limiting potential domination
of solar energy make it possible to avoid both of these
concerns simultaneously. These findings can inform long-term
technology choices and policies.
Third, we indicate that energy security benefits of de-carbonization,
apparent at the global scale, may not be equal for all countries
and regions, a factor potentially defining national support for the
global climate regime. This is consistent with other findings that
climate policies lead to divergence between regional energy systems
(Cherp et al., in Press). Although regional and national energy security
implications of climate mitigation policies need to be studied separately,
we do provide initial insights into this topic. In particular, we
show that while exports of oil from Middle East and North Africa
(MENA), decline in the climate stabilization scenarios, this decline is
generally similar to the one expected in the Baseline. Because this
initial observation contrasts the conventional wisdom that MENA
would be a major loser of the global climate regime, it should be
investigated in more details using a range of models and assumptions
on fossil fuel resource availability (see a meta-review of literature on
this topic in Jewell, (2013)).
Our study can also be used in national energy security assessments,
which are increasingly being used to guide national energy
policies, to characterize the global context within which national
energy futures may develop (Australian Government Department of
Resources Energy and Tourism, 2011; Wicks, 2009). In addition, the
systematic, rigorous and yet flexible framework proposed here may
be used by national policy makers to identify and assess energy
security implications of different policy options and scenarios.
The analysis presented in this article also opens an extensive
research agenda. First of all, our results come from one modeling
framework and one set of assumptions. Evaluating energy security
in other models and under different assumptions would help to
validate both the framework and the findings of our research.
Second, the results should be obtained for the national rather than
the global level, possibly starting with major economies (China,
EU, India and the US) which are more easily represented in IAMs.
Third, it is important to explore the ‘robustness’ perspective on
energy security for example by studying ‘buffers’ such as remaining
resources and spare electricity generation capacity. Finally,
detailed understanding of energy security concerns may allow
introducing relevant constraints in energy models and thus
producing more policy-realistic depictions of possible futures.
Acknowledgements
This work was developed as part of the Seventh Framework
Programme Low climate IMpact Scenarios and the Implications of
Tight emission control Strategies (LIMITS) project Grant agreement
no. 28246. The research also benefited from discussions during the
Global Energy Assessment (www.globalenergyassessment.org). Jessica
Jewell did exploratory work for this project during the Young
Scientists Summer Program (YSSP) at the International Institute
for Applied Systems Analysis (IIASA), with support from the
National Academy of Sciences, thanks to NSF Grant Number
OISE-738129. The authors would also like to thank Volker Krey,
Vadim Vinichenko, Nikolai Denisov, and Peter Kolp.
Appendix. Formulas for Indicators
This appendix contains a description of all the indicators
discussed in the paper and listed in Table 6.
11
As discussed above, although significant, this trade will be less of an energy
security concern than oil trade today because no end-use sector will depend upon
gas to the extent the transport sector depends on oil today.
J. Jewell et al. / Energy Policy 65 (2014) 743–760756
Table 6
Indicators of energy security in long-term energy transformation scenarios.
Indicator Energy security concern(s) Unit Definition (formula)n
Sector Geography
Sovereignty indicators
Global energy trade (absolute) Disruption of trade flows by various
factors
Ej/year Total flows of trade between regions in
a given year
TPES, oil, gas, coal, hydrogen, biomass,
synfuels, electricity, uranium, oil
products, other fuels and carriers
Global
Global energy trade (intensity) Same as above Share (0–1) Global energy trade divided by global
energy supply
As above (only TPES used in this
article)
Global
Geographic diversity of exports Same as above Non-dimensional SWDI or HHI As above Global
Net import dependency Regional vulnerability to trade
disruptions by various factors
share (0–1) Net energy imports divided by total PES
or total primary energy of a given
source
As above (only TPES used in this
article)
Regional
Cost of energy imports in relation to
GDP
Regional vulnerability to trade
disruptions by various factors
Share (0–1) Energy import value divided by GDP TPES or a particularly vulnerable fuel Regional
Cost of energy exports in relation to
GDP
Regional vulnerability to disruptions of
energy exports
Share (0–1) Energy export value divided by GDP TPES Regional
Carriers dependence on imported
fuels
Vulnerability of carriers to trade
disruptions
Share (0–1) Share of energy carriers produced from
imported sources divided by the total
energy carrier
Electricity, hydrogen, and other carriers Regional
End-use sectors dependence on
imported fuels
Vulnerability of end-use sectors to
trade disruptions
Share (0–1) Share of end-use sectors produced from
imported fuels
Transportation, industry, residential
and commercial
Regional
Resilience indicators
Energy intensity Overall vulnerability to energy supply
and price shocks
MJ/$ GDP TPES divided by GDP TPES Global or Regional
Diversity of energy sources in
primary energy supply (PES)
Overall vulnerability to various primary
energy source disruptions
Non-dimensional SWDI or HHI TPES Global or Regional
Diversity of primary energy sources
in carriers
Carrier vulnerability to various primary
energy source disruptions
Non-dimensional SWDI or HHI Electricity, hydrogen, liquid fuels, and
other carriers
Global or Regional
Diversity of primary energy sources
in end-use sectors
End-use vulnerability to various
primary energy source disruptions
Non-dimensional SWDI or HHI Transportation, industrial, residential
and commercial
Global or Regional
End-use sector diversity of carriers as above Non-dimensional SWDI or HHI Transportation, industrial, residential
and commercial
Global or Regional
Robustness indicators
Reserves or Resource to production
ratios
Vulnerability to energy shocks Years Reserves or resources divided by
production rates
Oil, gas, and coal Global or Regional
Average age of infrastructure Reliability of energy conversion and
transmission
Years in relation to
projected life-time
The age of all infrastructural facilities Electricity transmission and
generation; potentially other carriers or
fuels
Global or regional
Spare capacities for electricity
generation
Reliability of electricity generation % Installed capacity divided by the critical
or average load
Electricity Regional
Rate of energy sector growth Burden on energy systems associated
with fast growth
%/year The growth in energy supply (or use) in
fuel, carrier or end-use
End-uses, carriers, sectors Global or regional
Rate of energy export revenue
decline
Instability associated with fast decline
of energy export revenues
%/year The change in energy export revenues
year on year
Energy exports Regional
Compound indicators
Compound diversity index Combined diversity and sovereignty
concerns
Non-dimensional Modified SWDIn
TPES Regional
Bold text represents the indicators used or energy systems addressed in this article.
n
See the formulas and the explanation in the main text.
J.Jewelletal./EnergyPolicy65(2014)743–760757
Energy trade
Global energy trade is the sum of all net exports for each
globally-traded fuel or carrier. This paper analyses trade in oil, gas,
coal, electricity, hydrogen, biofuels, electricity and fossil synfuels.
Other fuels and carriers, e.g. uranium or specific types of bioliquids
could also be analyzed. Global energy trade is the sum of all
global fuel trade. This value only accounts for interregional trade
which is likely to be a large part of country-to-country trade.12
Trade intensity is calculated by dividing the total volume of
energy trade (or the volume of trade for a carrier or a fuel) by the
total primary energy supply (or the total supply for a carrier or a
fuel). In this study, we use the substitution equivalent but it could
be done using another primary energy accounting method as well.
The important thing is to use the same primary energy accounting
method to compare different scenarios.
On the regional level, net-import dependence is the difference
between exports and imports divided by the regional TPES.
Similarly, fuel import dependency or carrier import dependency
is the net-imports divided by the primary energy supply of that
source. A symmetrical indicator can be calculated for energy
exports. Import and export indicators can be expressed (a) as
absolute volumes of traded energy (aggregated or by individual
fuel or carrier); (b) as shares of a traded energy/fuel/carrier in its
overall regional supply; (c) as costs of a traded energy/fuel/carrier
expressed in absolute terms or as a share of regional GDP.
Another regional trade indicator is the reliance on imported
fuels in carrier production or end-use sectors. This can be
calculated by decomposing the end-use sectors into their
globally-traded fuels and carriers (similar to the primary energy
source decomposition in Fig. 14 below) and sum the net-imports
for each globally-traded fuel or carrier. This indicator could
capture concerns such as the high import dependence of the
transportation sector in most countries which dominates the
energy security landscape today (Cherp et al., 2012).
The final energy trade indicator is the geographic diversity of
exports for each globally-traded fuel or carrier. The regional
exports for each globally-traded fuel or carrier is calculated by
dividing a region's net-exports of a fuel or carrier by the total
volume of trade for the respective fuel or carrier. Then the SWDI
index (see next section) is calculated for the distribution between
energy exporting regions.
Diversity
For diversity, we use the Shannon-Wiener diversity index
(SWDI) which is calculated as
SWDI¼Σi pi ln(pi)
where pi is the share of the primary energy source i in the TPES.
The Herfindahl-Hirschmann index13
(HHI) has also been used in
the literature as a measure of diversity (Neff, 1997; Jewell, 2011;
International Energy Agency, 2011; Grubb et al., 2006). Stirling
(1998) argues that the SWDI is better than the HHI because the
ordering of results are not influenced by the base of the logarithm
which is used and Grubb et al. (2006) found no difference in their
conclusions when comparing results from the SWDI and the HHI.
Much more important than the question of which diversity
index is the issue of the diversity of what. The most meaningful
analysis of diversity is one that measures the diversity of energy
options within a vital energy system. The term “system” means
that it consists of resources, infrastructure, technologies, markets
and other elements connected to each other stronger than they are
connected to the outside world. From an energy security angle, it
means that in the case of a disruption, the elements within a
Fig. 14. Proportional allocation of primary energy sources for transportation.
12
Country-to-country oil trade in 2005 was about 110EJ (British Petroleum,
2009) compared to 83EJ of the interregional trade in the MESSAGE model for this
year. Thus, interregional trade currently accounts for about 75% of all oil trade. 13
Herfindahl Hirschmann index¼Σipi.
J. Jewell et al. / Energy Policy 65 (2014) 743–760758
system can replace each other but the elements outside the
system cannot (Cherp and Jewell, 2013). Indeed diversity indices
were first proposed to measure the diversity of sources in an
electricity system (Stirling, 1994) which are readily-substitutable;
that is, there is no difference between electricity produced from a
gas-fired power plant or a concentrated solar power system.
In this article, we present the diversity of TPES as well as the
diversity of energy sources used for electricity generation and
transport. The TPES diversity was calculated based on the proportion
each primary energy source contributed to the TPES (using
the substitution equivalent primary energy accounting method).
The SWDI for electricity reflects the diversity of fuel sources used
for electricity generation. Such an index can also be used for other
carriers such as synfuels, liquid fuels in general or hydrogen. The
SWDI is calculated for end-uses based on the diversity of primary
energy sources by proportionally allocating different energy carriers
to their respective sources (see Fig. 13). This proportional
allocation needs to be tailored for the specific configuration of the
energy system, actual or modeled. The way we apply the end-use
diversity index accounts for disruptions which would occur at the
primary energy level. It's also possible to measure the diversity of
carriers (e.g. electricity vs. liquid fuels) used in an end-use sector.
We also apply the SWDI to the geographic diversity of exports.
Measuring the geographic diversity of supply (or geographic
concentration of supply) has been done in Jewell, (2011),
Costantini et al. (2007) and Lefèvre (2007).14
The geographic
diversity of supply uses the same formula as the diversity of TPES,
electricity and transportation, but pi is the share of a given energy
source from region i in the total interregional trade of that energy
source or carrier.
As discussed in Section 5.3, some studies have used a compound
diversity index (CDI) on the regional level to combine the
import dependency and diversity on the regional level according
to the following formula:
CDI ¼ Σi 1Àmið1À
Sm
i
Sm;max
i
Þ
!
ðpilnðpiÞÞ
( )
where, pi is the share of the primary energy source i in the TPES;
mi is the share of imports of net imports in primary energy supply
of resource i; and Sm
i is the Shannon diversity index of import
flows of resource i and Sm;max
i is the maximum possible value of the
Shannon index if all regions exported an equal amount.
Energy intensity
Energy intensity is the amount of energy used per dollar of GDP
or value-added (in this paper MJ/US2005$). In this study it is a
unique indicator because it is partially-exogenous to the GEAmodeling
framework. Thus while we calculated the other energy
security indicators ex-post, we tested the effect energy intensity
has on other aspects of energy security.
Robustness indicators
Robustness indicators include resource scarcity, the rate of
demand growth, aging and reliability of infrastructure, the presence
of spare capacity, strategic stocks and resource buffers. As we
explain in the main text it is often difficult to meaningfully
represent these variables in Integrated Assessment Models because
they are not depicted in a model, are exogenous or are endogenously
optimized. Nevertheless they can be used for exploring the
relationship between different aspects of energy security as well as
between energy security and climate mitigation measures as is
done for example in Turton and Barreto (2006). Other robustness
indicators which have been suggested in the literature on current
energy security concerns include: rates of demand growth, reliability
of electricity and heating supply, energy infrastructure age,
spare storage capacities and number of import entry points (Cherp
et al., 2012; Jewell, 2011; Winzer, 2012).
References
APERC (Asia Pacific Energy Research Centre), 2007. A Quest for Energy Security in
the 21st Century: Resources and Constraints. Institute of Energy Economics,
Japan, available at: http://aperc.ieej.or.jp/file/2010/9/26/APERC_2007_A_Quest_
for_Energy_Security.pdf.
Alhajji, A.F., 2008. What is energy security? Definitions and concepts. Oil, Gas &
Energy Law Intell. 6 (3).
Atsumi, Masahiro, 2007. Japanese energy security revisited. Asia-Pac. Rev. 14 (1),
28–43, http://dx.doi.org/10.1080/13439000701330353.
Australian Government Department of Resources Energy and Tourism, 2009.
National Energy Security Assessment 2009. Commonwealth of Australia,
available at: http://www.aip.com.au/pdf/National-Energy-Security-Assess
ment-2009.pdf.
Australian Government Department of Resources Energy and Tourism, 2011.
National Energy Security Assessment 2011. Canberra, AU, Commonwealth of
Australia, available at: http://www.aip.com.au/pdf/National-Energy-Security-As
sessment-2011.pdf.
Bollen, Johannes, Hers, Sebastiaan, Bob van der, zwaan, 2010. An integrated
assessment of climate change, air pollution, and energy security policy. Energy
Policy 38 (8), 4021–4030, http://dx.doi.org/10.1016/j.enpol.2010.03.026.
Bosetti, Valentina, Carraro, Carlo, Duval, Romain, Tavoni, Massimo, 2011. What
should we expect from innovation? A model-based assessment of the environmental
and mitigation cost implications of climate-related R&D. Energy Econ.
33 (6), 1313–1320, http://dx.doi.org/10.1016/j.eneco.2011.02.010.
British Petroleum, 2009. BP Statistical Review of World Energy June 2009. London:
British Petroleum.
Cherp, Aleh, Jewell, Jessica, 2011. The three perspectives on energy security:
intellectual history, disciplinary roots and the potential for integration. Curr.
Opin. Environ. Sustainability 3 (4), 202–212, http://dx.doi.org/10.1016/j.
cosust.2011.07.001.
Cherp, Aleh, Jewell, Jessica, 2013.Energy security assessment framework and three
case-studies, in: Dyer, H., Trombetta, M.J. (Eds.), International Handbook of
Energy Security. Edward Elgar Publish- ing, Cheltenham,UK and Northampton,
MA, pp.231–277.
Cherp, Aleh, Adenikinju, A, Goldthau, Andreas, Hughes, Larry, Jansen, Jaap C., Jewell,
Jessica, Olshanskaya, Marina, et al., 2012. Chapter-5 Energy and Security, in
Global Energy Assessment: Toward a More Sustainable Future, Cambridge
University Press, Cambridge, UK and New York, NY, USA and the International
Institute for Applied Systems Analysis, Laxenburg, 325–384.
Cherp, Aleh, Jessica Jewell, Vadim Vinichenko, Nico Bauer, and Enrica De Cian. in
press. “Global Energy Security Under Different Climate Policies, GDP Growth
Rates and Fossil Resource Availabilities.” Climatic Change. http://dx.doi.org/10.
1007/s10584-013-0950-x.
Costantini, Valeria, Gracceva, Francesco, Markandya, Anil, Vicini, Giorgio, 2007.
Security of energy supply: comparing scenarios from a European perspective.
Energy Policy 35 (1), 210–226, http://dx.doi.org/10.1016/j.enpol.2005.11.002.
Criqui, Patrick, Mima, Silvana, 2012. European climate—energy security nexus a
model based scenario analysis. Energy Policy 41, 827–842, http://dx.doi.org/
10.1016/j.enpol.2011.11.061.
European Parliament, 2006. Directive 2005/89/EC of the European Parliament and
of the Council of 18 January 2006 concerning measures to safeguard security of
electricity supply and infrastructure investment. Off. J. Eur. Union, 22–27.
(February 2).
European Union Council, 2004. Council directive 2004/67/EC of 26 April 2004
concerning measures to safeguard security of natural gas supply. Off. J. Eur.
Union, 92–96. (April 29).
GEA, 2012: Global Energy Assessment - Toward a Sustainable Future, Cambridge
University Press, Cambridge, UK and New York, NY, USA and the International
Institute for Applied Systems Analysis, Laxenburg, Austria.
Gnansounou, Edgard, 2008. Assessing the energy vulnerability: case of industrialised
countries. Energy Policy 36 (10), 3734–3744, http://dx.doi.org/10.1016/j.
enpol.2008.07.004.
Greene, David L., 2010. Measuring energy security: can the United States achieve oil
independence? Energy Policy 38, 1614–1621, http://dx.doi.org/10.1016/j.
enpol.2009.01.041.
Grubb, Michael, Butler, Lucy, Twomey, Paul, 2006. Diversity and security in UK
electricity generation: the influence of low-carbon objectives. Energy Policy 34
(18), 4050–4062, http://dx.doi.org/10.1016/j.enpol.2005.09.004.
Gupta, Eshita, 2008. Oil vulnerability index of oil-importing countries. Energy
Policy 36 (3), 1195–1211, http://dx.doi.org/10.1016/j.enpol.2007.11.011.
Helm, D., 2002. Energy policy: security of supply, sustainability and competition.
Energy Policy 30, 173–184.
14
In these publications, the authors use the HHI index, however, as discussed
at the beginning of the section, which diversity index to use does not generally
change the conclusions and some authors prefer the SWDI on theoretical grounds.
J. Jewell et al. / Energy Policy 65 (2014) 743–760 759
IAEA (International Atomic Energy Agency), 2007. Considerations to Launch a
Nuclear Power Programme, International Atomic Energy Agency. Vienna.
IEA (International Energy Agency), 2011. Measuring Short-Term Energy Security.
OECD/IEA, Paris.
Jansen, Jaap C., Seebregts, Ad J., 2009. Long-term energy services security: what is it
and how can it be measured and valued? Energy Policy 38 (4), 1654–1664,
http://dx.doi.org/10.1016/j.enpol.2009.02.047.
Jansen, J.C,vanArkel,W.G., Boots, M.G., 2004. Designing indicators of long-term
energy supply security. Energy Research Centre of the Netherlands (ECN),
available at: http://www.ecn.nl/docs/library/report/2004/c04007.pdf.
Jewell, Jessica, 2011. The IEA Model of Short-Term Energy Security (MOSES). :
primary energy sources and secondary fuels. OECD/ IEA, Paris.
Jewell, Jessica, 2013. Energy Security and Climate Change Mitigation. the interaction
in long-term global scenarios. Doctoral dissertation, Department of
Environmental Sciences and Policy, Central European University, Budapest.
Jewell, Jessica, Cherp, Aleh, Vinichenko, Vadim, Bauer, Nico, Kober, Tom, McCollum,
Davivd, van Vuuren, Detlef P, van der Zwaan, Bob, forthcoming. Energy security
of China, India, the E.U. and the U.S. under long-term scenarios: results from six
IAMs. accepted to Climate Change Economics.
Kendell, James M., 1998. Measures of Oil Import Dependence. U.S. Energy
Information Agency, Washington, DC, available at: http://www.eia.doe.gov/
oiaf/archive/issues98/oimport.html.
Keppler, Jan Horst, 2007. Energy supply security and nuclear energy: concepts,
indicators, policies, background study in the context of the NDD security of
supply project.
Kruyt, B, van Vuuren, Detlef P, De Vries, H.J.M., Groenenberg, H., 2009. Indicators
for energy security. Energy Policy 37 (6), 2166–2181, http://dx.doi.org/10.1016/j.
enpol.2009.02.006.
Kuzemko, Caroline, 2011. UK Energy governance in the twenty-first century:
unravelling the ties that bind. Doctoral dissertation, Politics and International
Studies, University of Warwick, Coventry.
Le Coq, Chloé, Paltseva, Elena, 2009. Measuring the security of external energy
supply in the European Union. Energy Policy 37 (11), 4474–4481, http://dx.doi.
org/10.1016/j.enpol.2009.05.069.
Lefèvre, Nicolas. 2007. Energy security and climate policy: assessing interactions.
OECD/IEA, Paris.
Lefèvre, Nicolas, 2010. Measuring the energy security implications of fossil fuel
resource concentration. Energy Policy 38 (4), 1635–1644, http://dx.doi.org/
10.1016/j.enpol.2009.02.003.
Lehman Brothers, 2008. Global Oil Choke Points. New York, January 18.
Leimbach, Marian, Nico, Bauer, Baumstark, Lavinia, Edenhofer, Ottmar, 2009.
Mitigation costs in a globalized world: climate policy analysis with REMINDR.
Environ. Model. Assess. 15 (3), 155–173, http://dx.doi.org/10.1007/s10666-
009-9204-8.
Löschel, Andreas, Ulf, Moslener, Rübbelke, Dirk T.G., 2010. Indicators of energy
security in industrialised countries. Energy Policy 38 (4), 1665–1671, http://dx.
doi.org/10.1016/j.enpol.2009.03.061.
Manne, Alan S., Richard G. Richels. 2004. MERGE: an Integrated Assessment Model
for Global Climate Change, available at: http://www.cepe.ethz.ch/education/
EnergyPolicy/GERAD1.pdf.
McCollum, David, Nico Bauer, Katherine Calvin, Alban Kitous, and Keywan Riahi. In
press. “Fossil Resource and Energy Security Dynamics in Conventional and
Carbon-Constrained Worlds.” Climatic Change. http://dx.doi.org/10.1007/
s10584-013-0939-5.
McCollum, David L., Krey, Volker, Riahi, Keywan, 2011. An integrated approach to
energy sustainability. Nat. Climate Change 1 (9), 428–429, http://dx.doi.org/
10.1038/nclimate1297.
McCollum, David L, Krey, Volker, Riahi, Keywan, Kolp, Peter, Grubler, Arnulf,
Makowski, Marek, Nakicenovic, Nebojsa, 2013. Climate policies can help resolve
energy security and air pollution challenges. Clim. Change , http://dx.doi.org/
10.1007/s10584-013-0710-y. (February 19).
Messner, S., Strubegger, M., 1995. User's Guide for MESSAGE III, Working Pater WP-
95-069. International Institute for Applied Systems Analysis (IIASA), Laxenburg,
Austria.
Neff, T., 1997. Improving energy security in Pacific Asia: diversification and risk
reduction for fossil and nuclear fuels. Project Commissioned by the Pacific Asia
Regional Energy Security (PARES) Massachusetts Institute of Technology Center
Int. Stud. (December).
Nelson, Paul, Christopher M. Sprecher. 2008. What Determines the Extent of
National Reliance on Civil Nuclear Power? Nuclear Security Science & Policy
Institute, College Station, TX, NSSPI-08-014. available at: http://nsspi.tamu.edu/
media/36556/p21_pub1.pdf.
O'Leary, Fergal, Morgan Bazilian, Martin Howley, Brian O Gallachoir. 2007. Security
of Supply in Ireland. Energy Policy Statistical Support Unit. available at: http://
www.seai.ie/Publications/Statistics_Publications/EPSSU_Publications/Energy_
Security_in_Ireland/SEI_EPSSU_Security_of_Supply_Third_Report.pdf.
Pant, Girijesh, 2006. Energy security in Asia: the necessity of interdependence.
Strategic Anal. 31 (3), 523–542, http://dx.doi.org/10.1080/09700160701432037.
Penchansky, R.R., Thomas, J.W., 1981. The concept of access: definition and
relationship to consumer satisfaction. Med. Care 19 (2), 127–140.
Rao, S., Riahi, Keywan, 2006. The role of non-CO2 greenhouse gases in climate change
mitigation: long-term scenarios for the 21st century. Energy J. 27 (3), 177–200.
Riahi, Keywan, Grubler, Arnulf, Nakicenovic, Nebojsa, 2007. Scenarios of long-term
socio-economic and environmental development under climate stabilization.
Technol. Forecasting Soc. Change 74 (7), 887–935, http://dx.doi.org/10.1016/j.
techfore.2006.05.026.
Riahi, Keywan, Dententener, Frank, Gielen, Dolf, Grubler, Arnulf, Jewell, Jessica,
Klimot, Zbigniew, Krey,Volker, et al., 2012. Chapter 17 – Energy pathways for
sustainable development, in Global Energy Assessment: Toward a More
Sustainable Future. Cambridge University Press, Cambridge, UK and New York,
NY, USA and the International Institute for Applied Systems Analysis, Laxenburg,
1203–1306.
Rozenberg, Julie, Hallegatte, Stéphane, Vogt-Schilb, Adrien, Sassi, Olivier, Guivarch,
Céline, Waisman, Henri, Hourcade, Jean-Charles, 2010. Climate policies as a
Hedge against the uncertainty on future oil supply. Clim. Change 101 (3-4),
663–668, http://dx.doi.org/10.1007/s10584-010-9868-8.
Scheepers, M., Seebregts, Ad J., de Jong, J., Maters, H., 2007, EU Standards for Energy
Security of Supply: updates on the crisis capability index and the supply/
demand index quantification for EU-27. ECN/Clingendael International Energy
Programme, The Hague and Petten, NL. available at: ftp://ftp.ecn.nl/pub/www/
library/report/2007/e07004.pdf.
Shukla, P R, Dhar, Subash, 2011. Climate agreements and India: aligning options and
opportunities on a new track. Int. Environ. Agreements: Polit., Law Econ. 11 (3),
229–243, http://dx.doi.org/10.1007/s10784-011-9158-6.
Sovacool, Benjamin K, and Marilyn A Brown. 2010. “Competing Dimensions of
Energy Security: an International Perspective.” Annual Review of Environment
and Resources 35 (1) (November 21): 77–108. http://dx.doi.org/10.1146/
annurev-environ-042509-143035.
Sovacool, Benjamin K, Mukherjee, Ishani, 2011. Conceptualizing and measuring
energy security: a synthesized approach. Energy 36 (8), 5343–5355, http://dx.
doi.org/10.1016/j.energy.2011.06.043.
Stirling, Andy, 1994. Diversity and ignorance in electricity supply investment:
addressing the solution rather than the problem. Energy Policy 22 (3), 195–216,
http://dx.doi.org/10.1016/0301-4215(94)90159-7.
Stirling, Andy. 1998. On the Economics and Analysis of Diversity. Science Policy
Research Unit (SPRU), University of Sussex, Brighton. Electronic Working Papers
Series, Paper 28(December 18):1–156.
Stirling, Andy, 2010. Multicriteria diversity analysis a novel heuristic framework for
appraising energy portfolios. Energy Policy 38 (4), 1622–1634, http://dx.doi.
org/10.1016/j.enpol.2009.02.023.
Turton, Hal, Barreto, Leonardo, 2006. Long-term security of energy supply and
climate change. Energy Policy 34 (15), 2232–2250, http://dx.doi.org/10.1016/j.
enpol.2005.03.016.
van Vuuren, Detlef P, Stehfest, Elke, Elzen, Michel G.J., Kram, Tom, Vliet, Jasper,
Deetman, Sebastiaan, Isaac, Morna, et al., 2011. RCP2.6: exploring the possibility
to keep global mean temperature increase below 2 1C. Clim. Change 109 (1-
2), 95–116, http://dx.doi.org/10.1007/s10584-011-0152-3.
Walker, Gordon, Cass, Noel, 2007. Carbon reduction, ‘the public’ and renewable
energy: engaging with socio-technical configurations. Area 39 (4), 458–469.
Wicks, Malcom.2009. Energy Security: a National Challenge in a Changing World.
London, available at: http://www.igd.com/Documents/Industry%20Reports/
Energy/Energy%20Security%20a%20national%20challenge%20in%20a%20chan
ging%20world%202009.pdf.
Winzer, Christian, 2012. Conceptualizing energy security. Energy Policy 46 (C),
36–48, http://dx.doi.org/10.1016/j.enpol.2012.02.067.
J. Jewell et al. / Energy Policy 65 (2014) 743–760760