Energy Policy 108 (2017) 312321

Contents lists available at ScienceDirect

Energy Policy
journal homepage: www.elsevier.com/locate/enpol

Energy transitions and emerging economies: A multi-criteria analysis of MARK

policy options for hydropower surplus utilization in Paraguay

G. Blanco , R. Amarilla, A. Martinez, C. Llamosas, V. Oxilia
Polytechnic Faculty, National University of Asuncin, Paraguay

A R T I C L E I N F O A BS T RAC T

Keywords: In the context of increasing global eorts to migrate towards more sustainable energy systems, Paraguay is
Analytic Hierarchy Process (AHP) notable for boosting a power system based entirely on renewable sources. Paradoxically it has one of the lowest
Emerging economies rates of electricity consumption in Latin America. Hence, a carefully crafted energy transition policy aimed at
Energy transitions maximizing its energy resources could be the cornerstone of the country's long-term development strategy. The
Public policy under uncertainty
process of crafting public policy, particularly for emerging countries, entails an intricate decision problem
Multi-Criteria Decision Analysis (MCDA)
involving several policy options with the potential to inuence the country's welfare and all dimensions of socio-
economic development. Thus, we present a policy-making tool applied to the multi-criteria decision energy
policy problem in Paraguay, based on an Analytic Hierarchy Process model, which can aid in the crafting of
more nuanced decisions. Here, four policy options are considered based on economic, technical, social,
environmental and political criteria. Our ndings are signicant for the on-going policy debate on the surplus
question in Paraguay and suggest that, considering the model employed, a policy oriented at directing the
surplus to leverage the industrial cluster development is the best option. Additionally, we believe the proposed
model could benet other emerging economies for addressing similar policy dilemmas.

1. Introduction
Because of the need to migrate to increasingly renewable and
cleaner energy systems on the face of the widely acknowledged non-
sustainability of the global energy systems, countries all over the world
have been increasing their eorts to secure long-run energy security
based on sustainable energy sources. Among the tasks which must be
addressed to attain these goals, are the energy mixes diversication to
reduce the systems vulnerability, and the improvement of the employ-
ment of existing energetic resources to maximize their utility. To eect
these tasks, governance and decision-making processes at government
level play a prominent role (Markard et al., 2012). While the call for an
Energy Transition is a global issue, the task of re-steering national
policies to achieve the necessary transformations rests predominantly
in the hands of domestic governments.
Paraguay presents an interesting case of study, highlighting a
possible transition from an unsustainable biomass-burning running
economy towards an increased use of the already available clean and
sustainable electricity. The country's Energy Sector is a very particular
case amongst other emerging economies. For instance, given the sharp
divergence between its hydropower installed capacity and the average
electric consumption Paraguay is one of the largest per capita hydro-
electric power generators in the world, getting its energy from hydro-
electric plants built in association with Brazil (Itaipu) and Argentina
(Yacyreta) holding a 50% share in each project. Itaipu has an installed
capacity of 14000 MW, with an annual average production of
98,287 GWh and Yacyreta has an installed capacity of 3200 MW, with
an annual average production of 20,867 GWh (Sauer et al., 2015);
however, the country's consumption levels are very low. The World
Bank estimates Paraguay's domestic consumption per capita at
1400 kWh/year, equivalent to one third of the consumption levels of
Brazil and Chile (World Bank, 2014).
Furthermore, the country's indigenous primary energy production
is based wholly on renewable energy sources: hydropower and biomass.
In contrast, its consumption mix is made up primarily of fossil fuels
(38%) -the transport sector being the largest user- and unsustainable
biomass (44%) -used mainly by households for cooking and heating,
and by the industry (Ministry of Public Works, 2015). Although it is
one of the largest exporters of hydropower in the world, electricity
makes barely 18% of its nal consumption mix. The signicant
participation of unsustainable biomass -largely extracted from unma-
naged forests, contributing to the countrys already high deforestation
rates - and of imported fossil fuels, could be reverted by taking full
advantage of its hydroelectricity surplus.

Correspondence to: Facultad Politcnica, Universidad Nacional de Asuncin. Campus Universitario San Lorenzo, San Lorenzo 2111, Paraguay.
E-mail address: gblanco@pol.una.py (G. Blanco).

http://dx.doi.org/10.1016/j.enpol.2017.06.003
Received 17 October 2016; Received in revised form 29 May 2017; Accepted 1 June 2017
0301-4215/ 2017 Elsevier Ltd. All rights reserved.
G. Blanco et al.
The country's already existing large hydro power surplus opens up a
real possibility to enable a transition towards a more sustainable
energy system, provided that the solutions proposed by the policy
makers are aligned with the goal of making the most out of the
country's surplus hydropower for the benet of the society. Hence,
taking Paraguay as a case study, the goal of this paper is to contribute
with the use of an AHP-based tool to support decision making in
emerging and developing economies in the context of energy transi-
tions.
In this work, we consider Paraguay's decision making problem by
looking at four hypothetical policy options: Continue with the current
situation by continuing to cede hydropower surplus to Brazil (continu-
ing with the normal course of operation) (A1); Selling hydropower
surplus at the Brazilian wholesale power market (A2); Installing an
electro-intensive aluminum factory (A3), and; Incentivizing the devel-
opment of small industrial parks with an accumulated electricity
demand of 1100 MW (A4). To carry out the assessment, we present a
multi-criteria decision analysis (MCDA), based on an Analytic
Hierarchy Process (AHP) model under ve criteria: economic, techni-
cal, social, environmental, and political feasibility. Even more, we posit
that the use of MCDA/AHP in emerging economies can help by
providing a more holistic and multilateral view to decision making
processes and allow a more balanced approach to energy policy
making.
This paper is organized as follows. After a short background on the
country's energy issues, a brief account of state-of-the-art literature on
Energy Transitions and decision making in public policy is made in the
next subsection. In Section 2, we present the methods MCDA and AHP
in general and, later, in the specic case of energy planning. In this
vein, we propose a specic methodology for tting the energy problem
space in Paraguay. The scenarios are introduced and we describe the
criteria chosen and their indicators. In Section 3, we show the results
and discussion of our analysis, including the sensitivity analysis of the
model. In Section 4, the conclusions and policy implications of our
work are presented.
1.1. Country background and energy issues
Unique features make Paraguay worthy of closer attention as a
subject of study for the relatively straightforward and almost immedi-
ately visible benets it could derive from an energy transition. Macro-
economically, Paraguay has made signicant progress in scal and
monetary aspects, and with the onset of major social reforms (World
Bank, 2014) with a GDP which has been growing steadily, having
reached its peak in 2013, when the GDP expanded to 13.6% highest
rate of growth in Latin America in that year. On a similar note, the
American investor service Moody's upgraded Paraguay's government
bond rating to B(A1) from B(A2), and changed the Paraguayan outlook
to stable from positive (Moodys Investor Services, 2015).
As already described above, Paraguay's Energy Sector is character-
ized by the high disparity between its hydropower capacity and electric
consumption. The country has a very large hydroelectric power
potential per capita. With a population of around seven million people,
it has been estimated to have the resources to produce 68 TWh/year of
hydroelectric power, of which 64 TWh/year are thought to be econom-
ically exploitable (Toledano and Maennling, 2013). For instance, the
generated electricity in 2011 was ~55 TWh and the domestic electricity
demand ~7.5 TWh barely 13% of the generated electricity (Toledano
and Maennling, 2013).
The main source of surplus is the Itaipu Dam, built as a binational
project since 1973 on the basis of a treaty between Paraguay and Brazil
for the development and exploitation of Parana River's hydroelectric
potential. The Itaipu Dam alone produces about 90,000 GWh/year, of
which 50% correspond to Paraguay. Since Paraguay only consumes
about 10% of its total share (Amarilla et al., 2015), the hydropower
surplus of its share is ceded to Brazil in exchange for a xed monetary
313
Energy Policy 108 (2017) 312321
compensation in accordance with the Treaty which stipulates that the
surplus not consumed by one of the parties, in this case Paraguay, shall
be transferred to the other contracting party, Brazil, limiting the chance
to sell the energy to third-parties. Annex C of the treaty containing this
provision is due to be revised in 2023.
Despite of being one of the largest exporters of hydroelectric power
in the world, almost half of its domestic energy consumption is made
up of biomass with a 44% participation of this energy carrier.
Additionally, more than a third of its energy consumption basket
corresponds to oil fuels at 38% (MOPC, 2015; Amarilla et al., 2015)
while electricity, its second top export (Center for International
Development at Harvard University, 2016), constitutes barely around
18% of its nal energy demand. Not taking full advantage of its
electricity surplus, plus the signicant participation of non-sustainable
biomass and oil-based energy sources is a clear imbalance detrimental
to the country's energy security, the environment and the population's
quality of life therefore touching upon all fronts: social, economic, and
environmental.
On a positive note, the country's already existing large hydropower
surplus opens up a real possibility to make feasible an energy transition
with relative ease through the redirection of the country's Energy
Policy guided by a willful decision to encourage the spread of hydro-
power in its energy consumption basket. Fostering increased consump-
tion of indigenous hydroelectricity entails encouraging a transition
from other less sustainable energy sources and an increased use of
modern energy sources, such as electricity as opposed to unsustainable
biomass and imported fossil fuels. Moreover, it could benet the
population and mitigate negative environmental impacts while redu-
cing the biomass and oil dependency of the energy sector and therefore
improving its balance of payment and energy security. In this sense, a
hydropower based Energy Transition must constitute the cornerstone
of Paraguay's development by supporting and accompanying the
consolidation of the important economic growth the country has been
displaying in the past years.
1.1.1. Policy debate surrounding the surplus question
As awareness on the existence of large hydropower surplus and the
hydroelectricity sovereignty discourse became widespread among the
general public after 2009 (Folch, 2015), discussions regarding the re-
negotiation of the bi-national treaties as well as on how best to employ
the power surplus for society's overall welfare have been going on since
then. In 2011, former president Fernando Lugo, during his adminis-
tration, started pre-negotiations for the potential installation of an
energy-intensive aluminum smelter by Rio Tinto Alcan (RTA)which
proposed the production of 670000 MT of aluminum a year, with a
yearly projected consumption of 1100 MW.1
Following these events, two reports were written at the
Government's request on the assessment of the installation of RTA in
Paraguay and other alternatives for the utilization of the hydropower
surplus. The rst report was written by London based CRU Strategies
Ltd, and recommended either the installation of RTA starting in 2016
or to continue ceding hydroelectricity to Brazil (MIC, 2011; Congreso
Nacional-Sala Bicameral, 2011). Two years later, Jerey Sachs lead
Columbia Center on Sustainable Development formerly Vale
Columbia Center on Sustainable Development issued a second
report, hereafter referred to as the Sachs Report recommending to
sell the electricity surplus to the regional market and to use the
nancial resources in order to invest in the country's infrastructure
and health system, advising against the installation of RTA's aluminum
smelter (Toledano and Maennling, 2013).
Neither of the reports provided a complete analysis taking full
account of all relevant aspects. For instance, during the Public
Hearings, it was established that CRUs Analysis was on a Cost-
1
Compare to a peak demand of 3.026MW registered in February 2017.
G. Blanco et al.
Benet Analysis (CBA) which, although it mentioned them, did not
include environmental and social issues in the calculation of net
present value of the considered project. Whereby, the CBA considered
a discount rate of 8% for developing countries when dealing with
environmental and social issues, while it has been established that a
discount rate of 1012% is recommended (Markandya et al., 2001).
Further, as it was mentioned that the CRU Analysis is based on a CBA,
it is important to mention some weaknesses inherent to CBA assess-
ments. Such assessments generally present severe limitations when
assessing mega-projects that can compromise the natural endowments
of a country and create serious conicts among dierent political,
social, economic, technical and environmental objectives and may lead
to many inaccuracies that prevent an improvement in Pareto and
Kaldor-Hicks eciency. This is so mainly because of an overreliance on
data from past projects; the use of subjective impressions by assess-
ment team members, the inappropriate use of heuristics to derive
money cost of intangible elements, and, a conrmation bias among
project supporters, here it is to be added, that it is not uncommon for
interest groups to attempt to include or exclude signicant costs from
an analysis to inuence the outcome (Huesemann and Huesemann,
2011).
Neither analysis oered a systematic formal technique of assessing
the dierent ways of dealing with the issue of decision making under
uncertainty and neither of them tackles the imbalance of the energy
matrix and how to encourage an energy transition that could simulta-
neously generate opportunities for Paraguay's predominantly young
population through job creationsocial criterion, and ease the envir-
onmental pressure by cutting down the deforestation and biomass
burning emission rates,the environmental criterion. Regarding the
considered options, neither of the reports considered the possibilities
of selling the electricity surplus at market prices in the region or
allowing the installation of other industries. Such options were widely
discussed among experts and debated publicly.
1.2. Energy transitions and decision-making in public policy
1.2.1. Energy transitions
The goal to secure long-run energy supply and demand based on
sustainable energy systems can only be achieved through a system-
wide energy transition, which should be an inherent component of the
public agenda. We believe that in order to improve energy pathways
within the country, the political momentum can be used as a leverage,
as fuel price uctuations, environmental concerns, aspects of techno-
logical changes, and the democratization of energy access are currently
part of the political agenda (Arajo, 2014). Energy Transitions
generally involve a change in an energy system, either from a particular
source, a technology or a "prime mover (Miller et al., 2015). Arajo
(2014) employed a cross-cutting denition to avoid the confusion that
entails relying on dierent denitions, dening Energy Transitions as
being a shift in the nature or pattern of how energy is utilized within a
system. Given the fundamental uncertainties that transitions entail,
the denition of policy problems, the decision for a specic approach or
strategy and the formulation are ambiguous and contested and
fundamentally value-laden (Kern and Howlett, 2009; Markard et al.,
2016). Even more so, since energy systems are embedded in a myriad
of intricate patterns of social, economic, and political life and organiza-
tion. In order to allow positive and signicant changes in energy
systems, these should be accompanied by social, economic, and
political shifts. Therefore, energy policy is a problem of socio-energy
system design and requires a multi-faceted approach to energy related
problems (Miller et al., 2015).
On account of this high complexity, decision making for energy
policy should be backed with participative, systematic, and holistic
methods that take account of the relationships between dierent
stakeholders and the cross-cutting nature of energy issues that are
ubiquitous within any society, and refrain from adopting an excessively
314
Energy Policy 108 (2017) 312321
technical approach to transitions by focusing on the technological
aspects of transitions and their costs (Miller et al., 2015). To contribute
to these goals, we propose the use of MCDA methods to the case of
Paraguay and consider four hypothetical -but already discussed- policy
options for allowing a sustainable energy transition in the country.
1.2.2. Decision making in energy policy making
In the past decade or two, various decision-making approaches
have been used to scientically inform policy on the feasibility of
specic strategic transition paths, among which MCDA has gained
considerable relevance (Ren and Sovacool, 2015; Stein, 2013). When
many dierent objectives that conict with each other are at stake,
MCDA seems to be a suitable tool for decision making under
uncertainty. In the case of how best to incentivize energy transitions,
the requirement of simultaneously meeting multiple criteria econom-
ic, political, technical, social and environmentalis a pressing chal-
lenge for energy policy makers who need to articulate plans according
to multiple objective functions that incorporate the most important
possible tradeos and conicts among multiple viewpoints. Within this
context, MCDA has been widely used in order to support policy
decision making to complex energy planning problems and has the
advantage of oering tools for the better understanding of intrinsic
characteristics of the decision problem, encourage the role of partici-
pants in the decision making process, enable compromise and collec-
tive decisions, and provide transparency to the insights of the model
and analysis. In words of Pohekar and Ramachandran (2004), MCDA
can support policy makers in order to enhance the decision quality by
allowing them to be more explicit, rational, and ecient. In turn, AHP
is probably the most widely-used of the MCDA methods. This paper
presents a MCDA, based on an AHP model, of the sustainable use of
hydropower surplus for society's overall welfare in Paraguay that would
make possible such a transition. AHP is a structured technique for
organizing and analyzing complex decisions. Developed by Thomas L.
Saaty in the decade of 1970, it has been widely employed and
developed since then. AHP addresses the issue on how to determine
the relative importance of a set of activities in a multi-criteria decision
problem. This method is based on three principles: (1) structure of the
model; (2), comparative judgment of the alternatives and the criteria;
and (3) synthesis of the priorities (Saaty, 1994). Thus, the method
allows combining complex information for decomposing alternatives
and relevant information into a hierarchy structure of criteria. Rather
than prescribing an "optimal" decision, AHP supports decision makers
to nd an alternative that best ts an overall goal and their under-
standing of the problem. It provides a comprehensive and rational
framework for arranging a decision problem, for representing and
quantifying its elements, for relating those elements to overall goals,
and for evaluating alternative solutions. A detailed description on how
to structure and pursue an AHP for decision making is found in Saaty
(1994). We posit that the use of MCDA/AHP in emerging and
developing economies, such as in Paraguay, can help democratize
and allow a more inclusive approach to energy policy making.
Especially in countries like Paraguay with a long story of exclusivist
as opposed to participative practices encouraged by the author-
itative tradition of its political structure. In our view, the proposed
methodology can render more adequate results when dealing with
complex decisions for policy making, than for instance CBA-based
approaches.
2. Methodology
2.1. Applications of MCDA/AHP in energy policy and planning
MCDA methods have been extendedly applied in energy policy and
planning. For instance ENE-MCA is an interactive policy analysis tool
developed at the International Institute for Applied System Analysis
(IIASA), which purpose is to aid decision makers in their assessments
G. Blanco et al.
of future policy choices and how those choices impact the multiple
dimensions of energy sustainability. Browne et al. (2010) describe the
use of MCDA to explore alternative domestic and electricity policy
scenarios in Ireland. Tsoutsos et al. (2009), apply MCDA methods for
analyzing sustainable energy planning in Crete and Klein and Whalley
(2015), compare the sustainability of U.S electricity options through
MCDA. There are a number of specic methods of MCDA used in
energy planning: PROMETHEE, (Goumas and Lygerou, 2000),
ELECTRE (Beccali et al., 2003), MAUT (Voropai and Ivanova, 2002),
and fuzzy methods and decision support systems or DSS (Skikos and
Machias, 1992). However, for our purposes, AHP seems to best t our
problem space given its capacity of converting a complex problem into
a simple hierarchy, its exibility, intuitive appeal, and its ability to mix
qualitative as well as quantitative attributes in the same decision
making exercise (Pohekar and Ramachandran, 2004). Likewise,
MCDA/AHP have been extendedly used in relevant research elds
such as renewable energy (Ramanathan and Ganesh, 1995; Ren and
Sovacool, 2015), energy resource allocation (Hobbs and Horn, 1997);
transport energy (Yedla and Shrestha, 2003);and energy utility plan-
ning (Stein, 2013).
2.2. Proposed methodology for assessing alternatives of the
hydropower surplus utilization based on AHP
We present a Multi-Criteria Decision Analysis (MCDA) based on an
Analytic Hierarchy Process (AHP) applied to energy policy options and
consider multiple criteria under uncertainty. Under the lenses of
MCDA/AHP, we analyze dierent energy policy options for the use of
hydropower surplus. In the context of two workshops held in the years
2015 and 2016 with local stakeholders, the goal society's overall
welfare was chosen and the relevant criteria were elucidated. Each of
the presented options are meant to be considered as hypothetical
courses of action that decision makers might or might not choose to
take. The options were generated on the basis of results of a consulta-
tion among stakeholders, at the same time they are realistic and
relevant given that some of them are already included in the political
agenda, with options A1, A2 and A3 already being considered in the
CRU and Sachs Analyses. On the other hand, A4 is a policy option
which has not yet been seriously considered in the policy agenda until
now. Furthermore, the ve criteria (economic, technical, social, en-
vironmental, and political feasibility) were categorized according to
their degree of importance based on the results of a survey conducted
among 80 stakeholders to collect information on their cardinal
preferences of the criteria in a scale from 1 to 9.
The policy options considered are:
(A1): Continue with normal course of operations: Paraguay re-
mains ceding its electricity surplus to Brazil (i.e. Business as Usual
or BaU).
(A2): High hydropower export level: Electricity is sold in the
Brazilian wholesale power market at market prices, generating a
net rent of around 50USD/MWh (CCEE, 2015; Toledano and
Maennling, 2013).
(A3): High electro-intensive industry penetration level: a factory
resembling Aluminum Smelter RTA of 1100 MW is installed in the
country.
(A4): High development of small industry: Many small industries
are installed. Industrial clusters of 180 MW are established at the
baseline scenario and reach up an accumulated demand of
1100 MW.
Policy options A1 and A2 could potentially exhort and keep
perpetuating rent-seeking behavior, as some powerful stakeholders in
Paraguay can expend resources on political activity to increase their
share of the existing wealth arising from the sale or cession of
hydropower surplus. Regarding option A3 we particularly chose to
315
Energy Policy 108 (2017) 312321
consider the aluminum smelter as a policy option, as it has been in the
public agenda and considering that the government has shown interest
in it. Additionally, it was considered based on the fact that there are
updated and well backed studies on its potential with updated
information on the project. Some caveats were already pointed out in
the Sachs Report, warning that worldwide only about one third of the
large aluminum smelters process 50% or more domestically (Toledano
and Maennling, 2013). Where this is the case, smelters benet from
strong government intervention and the host countries are relatively
rich like Saudi Arabia, Oman, Brazil, and Canada (Toledano and
Maennling, 2013). In our point of view, option A4 has the potential
to unleash the technological and associated institutional/organizational
transformations in energy end-use that are recognized as fundamental
drivers of historical energy transitions. A transition to electricity wide-
use could expand the demand for electricity and give rise to a transition
in energy services so new technological combinations can enable new
and/or improve services at greater eciency and decreasing costs
(Grubler, 2012). Even more, option A4 is a realistic option in the case
of Paraguay, specially bearing the fact that developing countries
generally experience energy-related transitions at lower levels of
income, with faster rates of change in conditions over time, and tend
to use less energy per capita than developed countries at lower
systematic environmental impact per capita (Marcotullio and Schulz,
2007).
2.3. Criteria and indicators of evaluation
The policy options, or alternatives, were evaluated according to
economic, technical, environmental, political feasibility, and social
criteria. In our view, these decision criteria belong to the fundamental
aspects that are central to this particular policy decision making
process, and are crucial when elucidating and justifying preferences
of a society. Adequate criteria selection constitutes an essential stage in
this process. An inadequate approach can lead to poor results or can
even invalidate the entire decision-making process. The criteria set can
always change according to the point of view of the policy maker and
the problem to be analyzed so they are not xed. However, the criteria
should be understandable and measurable so they can actually express
the fulllment of the objective. Correspondingly, a known scale has to
be associated to each criterion either through obtaining a probability
distribution of the dierent levels of the attributes for every alternative
or through the assignation of an order of preferences. The attributes
need to embody properties like completeness, (the attributes are
adequate), decomposability (attributes allow the simplication of the
evaluation process in a piecemeal approach), no redundancy (no
double entries), and minimalist (small is smart and beautiful). On
the basis of this, once the criteria were dened, we proceeded to
construct their hierarchy. The hierarchy allows analyzing the problem
in-depth so it can be represented in the most complete and global way,
considering the environment of the problem and identifying the
attributes that contribute to the solution while taking the stakeholders
into consideration.
Keeping all the above exposed in mind, Fig. 1 below shows the
hierarchy of the energy policy decision we have congured. In the next
paragraphs, we link the criteria with their indicators.
2.3.1. Economic criterion
this criterion takes account of the general dimensions that con-
tribute to services and goods creation in a country. The indicator for
this criterion is the average GDP growth rate over the analyzed horizon.
Firstly, we prove the causality of the industrial electric consumption on
the GDP by a Granger causality test, where historic time series (period
19912011) are analyzed .2 This allows us to accurately articulate the
2
Dataset of the economic criteria can be found at: https://goo.gl/SXSVKd
G. Blanco et al.
Societys Overall
Economic Social
Continuing to cede Selling hydropower
hydropower surplus surplus at the
to Brazil Brazilian wholesale
power market
Fig. 1. Hierarchy of the energy decision.
evolution of GDP in terms of the industrial electric consumption. Thus,
we dene a linear function in order to forecast the temporal evolution
of the GDP as function of the industrial electricity demand in Paraguay
(Eq. (3)) based on the Paraguayan GDP (Banco Central del Paraguay,
2011) and electric consumption (Amarilla et al., 2015) historical time
series:
GDP (t ) = 558 + 13. 5 ED (t )
where GDP(t) is GDP in thousand million of PYG at the year t and
ED(t) is the industrial Paraguayan electricity consumption in GWh at
the year t (Eq. (2)). Under this context, each alternative has a particular
treatment regarding its marginal contribution to the GDP. For A1, the
trend of these variables is estimated based on these functions in order
to estimate the average GDP growth. In the case of A2, we added to the
GDP the income generated by the sale of electricity at a rent equal to
50 USD/MWh, assuming that Paraguay sells about 9504 GWh each
year starting in 2017. Likewise, for A3 the contribution of the RTA's
Aluminum Smelter is added to the GDP. These values are obtained
from the CRU analysis (Ministerio de Industria y Comercio, 2011,
Congreso Nacional Sala Bicameral, 2011). Finally, in the case of A4,
the average GDP growth is estimated by adding to the electricity
consumption corresponding to the installation of the new industries
(1037 GWh per year in an arithmetic progression).
2.3.2. Social criterion
this criterion takes account of a fundamental aspect that contri-
butes to the social welfare: job creation (Wang et al., 2009). We have
chosen job creation as an indicator, among other potential indicators,
mainly given the relevance of the creation of long-term jobs not
associated with transfers subject to royalties which are transitory,
dependency-producing and of a handout-approach nature that will
always depend on the existence of a surplus on the short term. In the
specic case of Paraguay, the importance of job creation as a social
criterion is paramount given the high levels of sub-employment and
informal jobs (ILO, 2014) and the fact that it would allow to empower
sectors of the population which would otherwise rely on handout
approach programs, given that the welfare system (unemployment
services) is not very strong in Paraguay, as is the case of advanced
welfare states (such as in Scandinavia or central European countries).
We estimate the number of jobs either by means of projections based
on historical data3 or specic additional studies depending on the
option which is being analyzed. Whereas for the alternatives A1 and A2
they do not entail new direct job creation, employment generation
values for A3 were obtained from data provided during the Public
3
Dataset used in the social criteria can be found at: https://goo.gl/9Sf2ad
Energy Policy 108 (2017) 312321
Welfare
Political
Environmental Opportunity
feasibility
Installing an Incentivizing the
electro-intensive development of small
aluminum factory industrial parks
Hearings organized in 2011 to discuss the potential installation of Rio
Tinto Alcan in Paraguay. It is estimated that the establishment of the
aluminum smelter would translate in the creation of 7000 direct job
opportunities (Ministerio de Industria y Comercio, 2011; Congreso
Nacional - Sala Bicameral, 2011). For option A4, the employment
creation value is obtained by a function which links the number of jobs
generated and the electricity consumption. This function is tted based
(1)
on data of employment generation by industries in Paraguay during the
period 20072011 and the corresponding electricity consumption of
this sector (Eq. (1)). The regression equation is:
EG (t ) = 45945 + 327 ED (t ) 21155 t (2)
where EG(t) is the total number of generated jobs and ED (t) the
industrial electricity demand (GWh) in year t. The coecient of
determination, R2 (R-square) of the proposed regression is 92.1%.
Additionally, we estimate a lineal regression of the industrial
electricity consumption of Paraguay (Eq. (2)) based on historical data
and dened in the following way:
ED (t ) = 1216. 36 + 69.7 t (3)
2
In this case, the R is 95.1%, so this entails the signicant goodness
of t of the model.
2.3.3. Political feasibility criterion
Within this criterion, we propose to summarize the feasibility of the
policy options given the political and institutional landscape through a
ranking based on the likelihood of implementation of the analyzed
policy options, by trying to measure the risk of eective implementa-
tion, based on an expert panel consultation. We assume that A1 has the
lowest risk since this scenario is already running. Under A3, the risk
level is relatively low, since negotiations are done solely with a single
agent, which in turn reduces the process complexity. Conversely,
under A2, Paraguay would have to rstly negotiate with Brazil in order
to be allowed to enter the Brazilian wholesale market. For this option,
an additional negotiation stage is necessary as others agents of the
market have to be approached in order to obtain a Power Purchase
Agreement (PPA) under convenient conditions. This process can be
quite challenging; therefore, the implementation of A2 is relatively
risky. Finally, the riskiest alternative under this criterion is A4. This is
so mainly because it entails a complete change of the paradigm in the
Paraguayan energy and productive sector, which could be quite
cumbersome given the existing political and institutional constraints.
Moreover, under A4, a massive installation of industries must be
incentivized and negotiations with several parties are required. In
addition, signicant infrastructure investments, including in the power
system and other basic infrastructure such as internal roads and water
provision infrastructure, telecommunication networks, among others,
316
G. Blanco et al.
are needed in order to follow the industrial development adequately.
Hence, when all these factors related to the negotiation costs of the
policy option implementation such as those mentioned above are
considered, plus the eorts needed to boost the coordination and
promotion of local and foreign direct investment, A4 is the riskiest of
all policy options.
2.3.4. Technical criterion
We choose as indicator of this criterion the Expected Cost of Energy
Not Supplied (CENS). This indicator attempts to measure the
impact of a given alternative on the electric energy security of the
country.
We represent the Paraguayan power system in a Direct Current
(DC) model of 94 buses. The parameters of transmission lines,
generation units, and loads are carefully selected based on the database
of the Paraguayan electric utility (ANDE) 4 for the existing system as
well as its expansion in the future. In this sense, Optimal Power Flow
(OPF) calculations in the Paraguayan power system are conducted in
order to determine the minimum (CENS) under the execution of each
alternative. Thus, the optimization problem is formulated as following:
min (Cg (Pgi ) + (CENS))
i q
P i P i F i = 0
g g d d l l
s. t . Pgi, min Pgi Pgi, max
Flmin Fl Flmax
where Cg is the supplier bid curve and andPgi Pdi
are the power
generated and demanded by unit g and customer d respectively at node
i. The ow in all lines connected to node i is denoted by Fl . The
operation limits of each generator unit are stated by Pgi, min, max and the
network constraints are set by Flmin, max .
Since the future energy security relies mainly on the electric
demand growth, which in Paraguay is quite volatile, we consider the
demand growth as an uncertain variable in order to cover all the
possible future scenarios. Hence, the stochastic evolution of the power
market model taken as a basis in this work can be characterized as a
bottom-up model, where the annual generation costs are directly
inuenced by the long-term stochastic movements of the electric
demand and the system topology. Then, the stochastic model of the
demand growth rate dRi , along an interval dt, can be represented by a
generalized Brownian Motion according to the following expression:
dR (t ) = Li dt + Li dz
where Li is the estimated unconditional mean load growth rate for the
year t, L2i the estimated unconditional variance for this time interval
and dz the Wiener process.
Under these conditions, an N-1 contingency analysis is performed
in each realization. The DC-OPF is calculated using the MATLAB-based
power system simulation toolkit Matpower 3.2. In the case of the
scenarios A1 and A2, we consider the baseline ANDE system at the year
2017, in which we model demand growth with uncertainty according to
the proposed stochastic process. Scenario A3 considers the additional
installation of a unique factory of 1100 MW. Likewise, within scenario
A4, it is considered an additional increment of demand of 180 MW per
year until 1100 MW. In all cases, under energy decit scenarios, the
marginal price is set equals to the value of VOLL (Value of Lost Load)
at the bus with load curtailment, which has been assumed to be
483 USD/MWh. 1000 Monte Carlo simulations are performed under
each scenario, year, and contingency respectively in order to estimate
the Expected Cost of Energy Not Supplied- (CENS).
4
Dataset of the Power system model, demand and generation patterns can be found
at: https://goo.gl/RmCKVM
Energy Policy 108 (2017) 312321
2.3.5. Environmental criterion
The indicator of this criterion is the average growth rate of the
emission of greenhouse gases (GHG), which is estimated by a projec-
tion including the energy consumption as an exogenous variable. The
time series analysis is performed based on historical data of CO2
emissions .5 For scenarios A1 and A2, we obtain the relationship
between of GHG emission (metric tons per capita) and consumption of
energy in Paraguay (all energy sources). The following regression
(R2=81.5%) is obtained based on historical data:
GHG (t ) = 0.92 + 0. 0190 t + 0. 000106 TED (t ) (6)
where GHG(t) are the greenhouse gases emitted in Paraguay at year t,
and TED(t) is the Total Energy Demand (including all the energy
sources) at year t. The evolution of the TED is estimated based on its
historical growth rate. A3 adds to the baseline the GHG emitted by the
Aluminum Smelter's emissions of greenhouse gases.. Finally, A4 adds
to the baseline scenario the amount of greenhouse gases caused by the
installation of blocks of industries of 180 MW. The corresponding TED
for this scenario is estimated by considering that the electric demand is
equal to about 23% of the total energy in Paraguay.
3. Results and discussion
The relative importance of each decision criterion was established.
Pair-wise comparisons between the dierent criteria were conducted
according to their importance in achieving the goal. In this sense,
surveys were conducted in order to gather the opinions of 80 signicant
(4) stakeholders of the economic, technical, environmental, social and
political sectors of Paraguay. Based on these ndings, a comparison
matrix of criteria was constructed. In this case, the cardinal preferences
of the surveyed agents are characterized from an absolute scale, where
several comparison were taken into account, and each opinion was
aggregated by using the geometric mean in order to derive a group
function (see Table 1). In this sense, the absolute cardinal preference
relation related to a pair wise judgment relation was the principal
eigenvector of the matrix of pair wise comparison of the stakeholders
opinions (Saaty and Vargas, 2013). Thus, this eigenvector helped to
establish the relative weight between each criterion obtained by
calculating the arithmetic mean of all criteria. The result was that the
sum of all values in the vector is always equal to one (see Table 1).
It is important to remark that the group function derived by using
the geometric mean satises Arrow's conditions (Arrowvian social
welfare), both when the judgments of each individual are consistent
and when they are not consistent. Thus, the consistency in the
(5) judgments of the survey participants is tested based on the calculation
of the Consistency Ratio (CR), which results equal to 0.3%. Following
this procedure, the performance of each alternative is estimated under
each criterion based on the indicators stated in the Section 2.4. Results
according to each criterion and a ranking of feasibility are presented in
Table 2 below.
Based on these results, we used the ratio between estimations of
average growth rates in order to nd the weights of the environmental
and economic indicators. On the other hand, under the technical and
social criteria, the weights were estimated based on the ratios between
the values of the (CENS) and employment number respectively. If
the calculated ratio is greater than nine, then the weight is set equal to
nine. Likewise, if the result is a decimal number between one and nine,
the weight is set equal to the immediate superior integer. Table 3 shows
the pair-wise comparison matrices and the weights assumed for each
criterion on each scenario. In turn, the calculations of the priority
vector of each scenario are performed based on each criterion. These
last results are presented in Table 4.
The results suggest that under the technical criterion, the strategy
5
Dataset of Environmental parameters can be found at: https://goo.gl/g6bAF5
317
G. Blanco et al. Energy Policy 108 (2017) 312321

Table 1 Table 5
Composite priorities of criteria. Composite priority vector.

Criterion Eigenvector Alternatives (A1) (A2) (A3) (A4)

Economic 0.24 Priority 0,24 0,20 0,27 0,29

Technical 0.21
Social 0.20
Political 0.18
Environmental 0.17
relevance of a multi-criteria approach can be assimilated, due to the
max = 5.01, CI = 0.003, CR = 0.003 fact that dierent solutions arise according to dierent point of views.
This issue, in emerging economies such as Paraguay, might constitute
Table 2 one of the biggest barriers that the policy makers should cope with in
Results of the assessment of each criterion under each alternative. order to promote an inclusive public policy based on a holistic view.
Based on the results of Table 1 and Table 4, the composite priority
CRITERION/ALTERNATIVES (A1) (A2) (A3) (A4)
vector is estimated (see Table 5) and A4 has the highest priority,
Technical (MUSD) 29.47 29.47 12.81 32.27 followed very closely by A3. Therefore, the best alternative for Paraguay
Environmental (%) 3.95 3.95 5.85 6.03 -from this approach-is to use its hydropower surplus to promote a large
Economic (%) 3.8 4.31 4 6.8 industrial development, generating labor demand through the high
Social (Number of jobs) 0 0 7000 152339
Political (Risk level) 4 2 3 1
development of industrial parks of around 180 MW which can sum up
to reach an accumulated demand of 1100 MW.

Table 3
The pair-wise comparison matrices and the weights for each criterion and alternative.

Technical Environmental Economic Social Political

A1 A2 A3 A4 A1 A2 A3 A4 A1 A2 A3 A4 A1 A2 A3 A4 A1 A2 A3 A4

A1 1 1 1/2 3 1 1 2 3 1 1/2 1/2 1/3 1 1 1/5 1/9 1 2 2 4

A2 1 1 1/2 2 1 1 2 3 2 1 1/2 1/2 1 1 1/5 1/9 1/2 1 1/2 2
A3 2 2 1 3 1/2 1 2 2 2 1 1/2 5 5 1 1/7 1/2 2 1 3
A4 1/3 1/2 1/3 1 1/3 1/3 1/2 1 3 2 2 1 9 9 7 1 1/4 1/2 1/3 1

Table 4
Priority vector and consistency ratio of alternatives under each criterion.

Alternatives/ Criterion Technical Environmental Economic Social Political

(A1) 0,25 0,35 0,12 0,05 0,43

(A2) 0,22 0,35 0,19 0,05 0,18
(A3) 0,42 0,19 0,27 0,19 0,29
(A4) 0,11 0,11 0,42 0,70 0,10
Consistency Ratio 0.02 0.004 0.03 0.09 0.02

A3 has the highest priority. This fact can be explained by considering
that the network interconnection (needed in order to supply electricity
to the Aluminum smelter) helps to reduce the congestion in the
transmission grid and consequently the (CENS). We assume that
this transmission reinforcement is conducted by the Aluminum
Company as a part of its facility investment.
From Table 4, A1 and A2 have the highest priority under the
environmental criterion. Under these alternatives, there are no con-
tributions to the baseline GHG emission; therefore, they show the best
performance according to this indicator. Conversely, A4 presents a
signicant better performance under the social and economic criteria
due to the fact that this scenario encourages GDP growth by generating
more jobs (see Table 2).
Finally, when highlighting the criterion political feasibility, A1 has
the highest priority. This could partially help explain the historical
evolution of the policy development in the past. Historically in
Paraguay, the A1 alternative was the dominant policy, from where is
possible to infer that the prevalent criteria among the policy-makers
was the Political Feasibility. In addition, the consistency ratios are
shown in Table 5. It can be seen that none of the criterion has a
consistency ratio above the threshold of 10%.
As aforementioned, according to dierent criterion, dierent stra-
tegies have the highest priority level. At this point, the importance and
3.1. Sensitivity analysis
We believe that an analysis of implications of possible perturbations
on the prioritization by the decision-makers of the dierent criteria is
of paramount importance. It is important to bear in mind that, even
though the criterion ranking presented in Section 3 was estimated
based on a weighted preference of the stakeholders, the nal decision
in a policy making process is usually made by a single political
authority, which may or may not take into consideration the integral
vision of the main stakeholders of the sector.
In this vein, we conducted a sensitivity analysis 6 by systematically
exploring all the possible combinations for prioritizing criteria .7 In this
way, we analyzed the performance of the policy strategies under
scenarios where dierent criteria are prioritized and others where a
6
A document containing an exhaustive sensitivity analysis can be found at:
goo.gl/ntdEUt
7
The 41 possible combinations for prioritizing criteria were extensively analyzed.
Firstly taking a one by one criterion to be prioritized, then two by two, three by three and
so on. The pairwise comparison of the prioritized criteria with the remainders, in each
case, was increased in a level of three (moderate importance). Additionally, scenarios
where only a mono criteria analysis was conducted were analyzed. Likewise a four-
criteria analysis, where a single criteria was excluded under each scenario, was
considered.

318
G. Blanco et al. Energy Policy 108 (2017) 312321

Fig. 2. Sensitivity analysis Strategy map.

T: Technical, S: Social, En: Environmental, Ec: Economic, P: Political

Labels Description

T-S-En-Ec-P Labels that represent the evaluation criteria, which do not carry the symbols (*) or (**), indicate that there is a greater preference for the mentioned criteria
T-S-En without excluding any. Example: If T-S appears on the label then the Technical and Social criteria have a higher preference (Weight = 3) than the rest (Weight =
T 1). This happens in the same way for all other combinations that do not contain (*) or (**).
T-S-En-Ec- In the case where the labels, which represent the evaluation criteria, bear the symbol (**), they indicate that they represent the results of the 4-criteria evaluations.
P** This evaluation excludes criteria that do not appear on the label. Example: T-S-En-Ec-P** represents an AHP evaluation, which only takes into account the
Technical, Social, Environmental and Political Feasibility criteria, excluding the economic criterion of the study.
T* Labels containing the symbol (*) indicate that a single-criterion analysis was performed. This analysis excludes all other alternatives. Example: The T* label
indicates that an analysis was made based only on the Technical criterion, excluding the other criteria.

given criteria is excluded of the analysis. Likewise, the results for a
mono-criteria analysis are shown. The Fig. 2 exposes the results of
these analyses and it can be understood as following: considering; for
example, taking the T-P scenario (Technical and the Political
criterion), A3 turns out to be the best strategy. On the other hand, if
we consider, for instance, as priority the En-Ec-P scenario
(Environmental, Economic and Political criteria) then A1 turns out to
be the best alternative. A similar analysis can be done for any
combination of prioritized criteria, resulting in a given strategy for
each case.
Likewise, the labels, which represent the evaluation criteria, bear
the symbol (**), they indicate that they represent the results for 4-
criteria evaluations. This evaluation excludes criteria that do not
appear on the label, i.e. for instance T-S-En-P** represents an AHP
evaluation, which only takes into account the Technical, Social,
Environmental and Political Feasibility criteria, but excluding the
Economic criterion of the study. Finally, labels containing the symbol
(*) indicate that a mono-criterion analysis was performed. This analysis
excludes all other alternatives, i.e. the T* label indicates that an
Hence, Fig. 2 portrays a map of ecient policies according to the
possible vision orientations of the decision-maker authority. Thus, it
can be seen, that A4 is the most predominant option for most of the
scenarios. In second place, A3 is the policy alternative with better
performance under most of the scenarios where Technical criterion is
prioritized. Finally, A1 is the most convenient strategy just under few
scenarios where the Political criterion is considered as relevant. In
order to quantify the magnitude among the alternatives, the area of
each strategy is presented in the map (Table 6).
In this sense, both A4 and A3 can make the most of the hydropower
surplus in order to balance the energy matrix and develop the
industrial sector of Paraguay. A4 and A3 can drive a positive energy
transition in Paraguay by allowing deep and wide technological and
institutional transformations in energy end-use. The only case where
Table 6
Composite priority vector.
Alternatives (A1) (A2) (A3) (A4)

analysis was made based only on the Technical criterion, excluding the Area strategy map 0,23 0,16 0,28 0,33
other criteria.

319
G. Blanco et al.
A1 could make sense would be where political criterion is the most
relevant, or in other words, where decision-makers pursue an alter-
native which tends to minimize the political risk of the energy policy.
Finally, a very important nding of this work is that the electricity
export policy A2 is not the best option under any prioritization
criterion at the analyzed price. Thus, this strategy is dominated by
the others alternatives and should not be selected under the electricity
price considered in this study.
8
4. Conclusions and policy implications
Of the four policy options under consideration, we found that both
A4 and A3 have the potential to unleash a positive and sustainable
change in the energy matrix and are likely to drive a positive energy
transition in Paraguay by allowing deep and wide technological and
institutional transformations in energy end-use. In this sense, A4 is the
best option, followed by A3. Regarding A4, the results of our work
suggest that the high development of relative small industrial clusters
can generate positive spillovers for the overall society. As Paraguay has
made signicant progress in the macroeconomic front, A4 is a realistic
option as for instance, emerging economies generally experience
energy-related transitions at lower levels of income, with faster rates
of change in conditions over time, and tend to use less energy per
capita than developed countries at lower systematic environmental
impact per capita. Even more, the sensitivity analysis implies that
option A4 is the most robust, as it creates more job opportunities and
contributes the most to the creation of wealth, even considering the
potential negative or unfavorable implications this choice carries with
it. On the basis of these results, we conclude that A4 is a viable policy
option which can allow a sustainable energy transition in Paraguay.
Also, A4 might enable the country to avoid the rent-seeking trap, as
opposed to options A1 and A2 which would perpetuate this behavior.
Furthermore, based on other experiences in emerging economies, it is
highly likable that, by choosing A3, the country could be compromising
most of its hydropower surplus to an industrial enclave that is not
known for allowing the installation of downstream industries.
Furthermore, we argue that the combination of MCDA/AHP can
help governments in emerging economies as they face increasing
pressures to deal with rising social, economic, and environmental
problems. Even more, we argue that the combination of MCDA/AHP in
the specic case of Paraguay can help policy makers make informed
decisions for sustainable energy use that are more inclusive and make
better use of its surplus of hydropower. Finally yet importantly, this
approach makes for a very relevant tool for revealing the policy maker
s hierarchy of criteria through the policy formulation process.
5. Critical review of the proposed model and further
research
For the proposed decision problem, and in view of our objective to
provide a tool to aid policy makers in the decision making process, the
proposed methodology and its level of complexity, proved useful for the
posed decision problem. Considering more complex elements, such as
new technologies (energy storage, electric mobility, etc.), which should
also be considered and could probably enrich the analysis, would
render the model too complex for our present objective. However, we
reckon that considering them in separate research projects could prove
very enriching and be a great contribution to understand the implica-
tions of the case of Paraguay. For instance, technologies related to fossil
fuel replacement, considering the great dependence of the Paraguayan
Energy Sector on this Energy carrier. Particularly given that the fossil
8
The core section of this analysis has already had a straightforward impact on the
energy policy of Paraguay for the period 20162040, which has been enacted by decree
as from 10 October 2016 (Decree 6092/2016). This energy policy took this research
project as one of the cornerstones of its vision.
320
Energy Policy 108 (2017) 312321
fuel demand is dominated by the transport sector -and to an extent the
industry sector which has specic characteristic and complexities
which were not addressed in-depth in this project as it is mainly
focused on the reutilization of the hydropower surplus. A project in the
lines of the paper by Shakya and Shrestha (2011), based for instance on
the work conducted by Sauer et al. (2015) could be an interesting
approach for this. In fact, the work of Sauer et al. (2015) even suggests
an interesting potential sector for regional productive integration and
industry policy incentives, as the electric car industry using resources
from Bolivia and Paraguay in the case of study, could generate a
positive loop encouraging job creation and replacement of the tradi-
tional fossil fuelled-based car eet.
Taking the AHP/MCDA approach as starting point has advantages
such as its simplicity, which makes it easier to explain to policy-makers
and for them to employ it. In this sense, we believe the chosen model
boosts a balanced complexity-performance trade-o. By considering
some more complex hypotheses, although the results would have been
more adjusted to real life situations, we believe there would have been a
large trade-o between the model's complexity and the usefulness of
the model as an accessible and practical tool to employ in policy
making settings. However, we are working on the thoroughness of our
model, and believe it represents a promising path for further works. As
part of a larger research program, we are currently trying to answer the
same question on the decision problem regarding the use of hydro-
power surplus in Paraguay by means of an Analytic Network Process
(ANP) (Saaty and Vargas, 2013). Further research could explore,
among others, which industrial clusters should be developed in
Paraguay in order to make the most of the industrial policy, for
instance by looking into concepts such as economic complexity and
product space (Hidalgo et al., 2007). Furthermore, on account of the
relevance of the political feasibility criteria for policy making in
emerging economies, looking at the politics of policy in detail
(Meadowcroft, 2009, 2011; Normann, 2015) could help shed light on
the factors which in the end play a very prominent role in the success of
the implementation of policy decisions.
Acknowledgement
The authors are very grateful to the Paraguayan National Council of
Science and Technology (CONACyT) for nancial support through the
projects 14-INV-290, 14-POS-032 and the PRONII program. Likewise,
the authors acknowledge fruitful contribution of Horacio Ojeda,
Manuel Garca, Tanja Sinozic and Arturo Gonzalez during the devel-
opment of this work. Finally, gratitude is also extended to one
anonymous reviewer for her/his comments and suggestions, which
were very helpful to improve this article.
References
Amarilla, R., Buzarquis, E., Domaniczky, J., Barn, B., Blanco, G., 2015. November.
Analysis of the energy sector of Paraguay. Energy balance in terms of useful energy in
2011. In: Proceedings of the 2015 IEEE Thirty Fifth Central American and Panama
Convention (CONCAPAN XXXV) pp. 17. IEEE.
Arajo, K., 2014. The emerging eld of energy transitions: progress, challenges, and
opportunities. Energy Res. Social. Sci. 1, 112121. http://dx.doi.org/10.1016/
j.erss.2014.03.002.
Banco Central del Paraguay, 2011. Incorporacin de las Binacionales a las Cuentas
Nacionales y a la Balanza de Pagos de Paraguay. http://www.ssme.gov.py/vmme/
images/noticias/Incorporacion_de_Ingreso_Binacionales_a_PIB%28rev%29.pdf.
(accessed 10 July 2016).
Beccali, M., Cellura, M., Mistretta, M., 2003. Decision-making in energy planning.
Application of the Electre method at regional level for the diusion of renewable
energy technology. Renew. Energy 28 (13), 20632087. http://dx.doi.org/10.1016/
S0960-1481(03)00102-2.
Browne, D., O'Regan, B., Moles, R., 2010. Use of multi-criteria decision analysis to
explore alternative domestic energy and electricity policy scenarios in an Irish city-
region. Energy 35 (2), 518528. http://dx.doi.org/10.1016/j.energy.2009.10.020.
CCEE, 2015. Cmara de Comercializao de Energia Eltrica. http://www.ccee.org.br.
Center for International Development at Harvard University, 2016. "The Atlas of
Economic Complexity". http://www.atlas.cid.harvard.edu. (accessed 7 September
2016).
G. Blanco et al.
Congreso Nacional - Sala Bicameral, 2011. La calidad de las Inversiones, Resumen
Audiencia Pblica Inversion de Rio Tinto Alcan. 60 pp http://www.ssme.gov.py/
vmme/pdf/riotinto/IIAudiencia/audiencia-5.pdf Accesed May 23, 2016.
Folch, C., 2015. The cause of all Paraguayans? Dening and defending hydroelectric
sovereignty. J. Lat. Am. Caribb. Anthropol. 20 (2), 242263.
Goumas, M., Lygerou, V., 2000. An extension of the PROMETHEE method for decision
making in fuzzy environment: ranking of alternative energy exploitation projects.
Eur. J. Oper. Res. 123 (3), 606613. http://dx.doi.org/10.1016/S0377-2217(99)
00093-4.
Grubler, A., 2012. Energy transitions research: insights and cautionary tales. Energy
Policy 50, 816. http://dx.doi.org/10.1016/j.enpol.2012.02.070.
Hidalgo, C.A., Klinger, B., Barabasi, A.-L., Hausmann, R., 2007. The product space
conditions the development of nations. Science 317 (5837), 482487. http://
dx.doi.org/10.1126/science.1144581.
Hobbs, B.F., Horn, G.T.F., 1997. Building public condence in energy planning: a multi
method MCDM approach to demand-side planning at BC gas. Energy Policy 25 (3),
357375. http://dx.doi.org/10.1016/S0301-4215(97)00025-6.
Huesemann, M., Huesemann, J., 2011. Techno-Fix: Why Technology won't Save us or
the Environment xxviii. New Society Publishers, Gabriola B.C., 435.
International Labour Organization -ILO, 2014. Slight decline and key challenges in
informal employment in Paraguay. Notes on Formalization of the Regional Oce for
Latin America and the Caribbean. http://www.ilo.org/americas/sala-de-prensa/
WCMS_245890/langen/index.htm.
Kern, F., Howlett, M., 2009. Implementing transition management as policy reforms: a
case study of the Dutch energy sector. Policy Sci. 42 (4), 391408. http://dx.doi.org/
10.1007/s11077-009-9099-x.
Klein, S.J.W., Whalley, S., 2015. Comparing the sustainability of US electricity options
through multi-criteria decision analysis. Energy Policy 79, 127149.
Marcotullio, P.J., Schulz, N.B., 2007. Comparison of energy transitions in the United
States and developing and industrializing economies. World Dev. 35 (10),
16501683. http://dx.doi.org/10.1016/j.worlddev.2006.11.006.
Markandya, A., Halsnaes, K, Lanza, A., Matsuoka, Y., Maya, S., Pan, J., Shogren, J.,
Motta, R.d., Zhang, Z., 2001. Costing methodologies. In: Metz, B. (Ed.) Climate
change 2001. Mitigation: contribution of Working Group III to the third assessment
report of the Intergovernmental Panel on Climate Change / edited by Bert Metz
[et al.]. Published for the Intergovernmental Panel on Climate Change, Cambridge,
451198.
Markard, J., Raven, R., Truer, B., 2012. Sustainability transitions: an emerging eld of
research and its prospects. Res. Policy 41 (6), 955967. http://dx.doi.org/10.1016/
j.respol.2012.02.013.
Markard, J., Suter, M., Ingold, K., 2016. Socio-technical transitions and policy change
advocacy coalitions in Swiss energy policy. Environ. Innov. Soc. Transit. 18, 215237
. http://dx.doi.org/10.1016/j.eist.2015.05.003.
Meadowcroft, J., 2009. What about the politics? Sustainable development, transition
management, and long term energy transitions. Policy Sci. 42 (4), 323340. http://
dx.doi.org/10.1007/s11077-009-9097-z.
Meadowcroft, J., 2011. Engaging with the politics of sustainability transitions. Environ.
Innov. Soc. Transit. 1 (1), 7075. http://dx.doi.org/10.1016/j.eist.2011.02.003.
Ministerio de Industria y Comercio, 2011. Equipo de Trabajo Institucional ETI. Proyecto
de Instalacin de una Planta de Aluminio en Paraguay. http://www.ssme.gov.py/
vmme/pdf/riotinto/diegozavala.pdf. (accessed 23 May 2016).
Miller, C.A., Richter, J., OLeary, J., 2015. Socio-energy systems design: a policy
framework for energy transitions. Energy Res. Social. Sci. 6, 2940. http://
dx.doi.org/10.1016/j.erss.2014.11.004.
321
Energy Policy 108 (2017) 312321
Moodys Investor Services, 2015. Moodys upgrades Paraguays government bond rating
to B(A1) from B(A2), and changes the outlook to stable from positive. https://www.
moodys.com/research/Moodys-upgrades-Paraguays-government-bond-rating-to-
B(A1)-from-B(A2)PR_320602. (accessed 10 July 2016).
MOPC, 2015. Balance Energtico Nacional 2014 en trminos de Energia Final, Asuncion.
http://www.ssme.gov.py/vmme/pdf/balance2014/Balance%20Energetico
%20Nacional%202014-Final-3.pdf. (accessed 10 May 2016).
Normann, H.kon E., 2015. The role of politics in sustainable transitions. The rise and
decline of oshore wind in Norway. Environ. Innov. Soc. Transit. 15, 180193.
http://dx.doi.org/10.1016/j.eist.2014.11.002.
Pohekar, S.D., Ramachandran, M., 2004. Application of multi-criteria decision making to
sustainable energy planninga review. Renew. Sustain. Energy Rev. 8 (4), 365381.
http://dx.doi.org/10.1016/j.rser.2003.12.007.
Ramanathan, R., Ganesh, L.S., 1995. Energy resource allocation incorporating
qualitative and quantitative criteria: an integrated model using goal programming
and AHP. Socio-Econ. Plan. Sci. 29 (3), 197218. http://dx.doi.org/10.1016/0038-
0121(95)00013-.
Ren, J., Sovacool, B.K., 2015. Prioritizing low-carbon energy sources to enhance China's
energy security. Energy Convers. Manag. 92, 129136. http://dx.doi.org/10.1016/
j.enconman.2014.12.044.
Saaty, T.L., 1994. How to make a decision: the analytic hierarchy process. Interfaces 24
(6), 1943. http://dx.doi.org/10.1287/inte.24.6.19.
Saaty, T.L., Vargas, L.G. (Eds.), 2013. Decision Making with the Analytic Network
Process: Decision Making with the Analytic Network Process. Springer, US, Boston,
MA.
Sauer, I.L., Escobar, J.F., da Silva, M.F., Meza, C.G., Centurion, C., Goldemberg, J., 2015.
Bolivia and Paraguay: a beacon for sustainable electric mobility? Renew. Sustain.
Energy Rev. 51, 910925.
Shakya, S.R., Shrestha, R.M., 2011. Transport sector electrication in a hydropower
resource rich developing country: energy security, environmental and climate change
co-benets. Energy Sustain. Dev. 15 (2), 147159.
Skikos, G., Machias, A.V., 1992. Fuzzy multicriteria decision making for the evaluation of
wind sites. Wind Energy 6 (4), 213228.
Stein, E.W., 2013. A comprehensive multi-criteria model to rank electric energy
production technologies. Renew. Sustain. Energy Rev. 22, 640654. http://
dx.doi.org/10.1016/j.rser.2013.02.001.
World Bank, 2014. New World Bank Group Strategy to Support Reduction in Extreme
Poverty and Inequality in Paraguay. http://www.worldbank.org/en/news/press-
release/2014/12/16/new-world-bank-group-strategy-to-support-reduction-in-
extreme-poverty-and-inequality-in-paraguay. (accessed 9 July 2016).
Toledano, P., Maennling, N., 2013. Leveraging Paraguay's Hydropower for Sustainable
Economic Development. Columbia Center on Sustainable Investment, 109 pp.
http://ccsi.columbia.edu/les/2014/01/Leveraging-Paraguays-Hydropower-for-
Economic-Development-Final-CCSI.pdf. accessed 10 May 2016.
Tsoutsos, T., Drandaki, M., Frantzeskaki, N., Iosidis, E., Kiosses, I., 2009. Sustainable
energy planning by using multi-criteria analysis application in the island of Crete.
Energy Policy 37 (5), 15871600. http://dx.doi.org/10.1016/j.enpol.2008.12.011.
Voropai, N.I., Ivanova, E., 2002. Multi-criteria decision analysis techniques in electric
power system expansion planning. Int. J. Electr. Power Energy Syst. 24 (1), 7178.
http://dx.doi.org/10.1016/S0142-0615(01)00005-9.
Yedla, S., Shrestha, R.M., 2003. Multi-criteria approach for the selection of alternative
options for environmentally sustainable transport system in Delhi. Transp. Res. Part
A: Policy Pract. 37 (8), 717729.