Energy Strategy Reviews 20 (2018) 35e48

Contents lists available at ScienceDirect

Energy Strategy Reviews

journal homepage: www.ees.elsevier.com/esr

Fuel saving strategies in the Andes: Long-term impacts for Peru,

Colombia and Ecuador
Mauro F. Chavez-Rodriguez a, Pablo E. Carvajal b, *, Juan E. Martinez Jaramillo c,
Alejandro Egüez d, Rosa Esperanza Gonzalez Mahecha a, Roberto Schaeffer a,
Alexandre Szklo a, Andre F.P. Lucena a, Santiago Arango Aramburo e
a
Energy Planning Program, COPPE/UFRJ e Universidade Federal do Rio de Janeiro, Cidade Universita
ria, Centro de Tecnologia, Bloco C, sala 211, Ilha do
Fundao, CEP: 21.945-970, Caixa Postal: 68.501, Rio de Janeiro, RJ, Brazil
b
UCL Energy Institute, University College London, Central House, 14 Upper Woburn Place, London, WC1H 0NN, UK
c
n Universitaria Polit

Institucio
, Colombia
ecnico Grancolombiano, Calle 57 Nº 3-00 Este, Bogota
d €nd 6, 901 87, Umeå, Sweden
Centre for Environmental and Resource Economics, Umeå University, Social Sciences Building, Biblioteksgra
e
Department of Computing and Decision Sciences, Universidad Nacional de Colombia e Sede Medellín, Carrera 80 No. 65-223, Bloque M8A, Medellín,
Colombia

a r t i c l e i n f o a b s t r a c t

Article history: Crude oil exports and imports play a crucial role in the trade balance of Peru, Colombia and Ecuador.
Received 7 June 2016 These countries are looking into fossil fuel saving measures as a way to deal with increasing demands
Received in revised form and decreasing oil production rates. However, studies about the long-term effects of these measures is
27 December 2017
lacking. Using a combination of energy simulation technics, 17 fuel saving strategies ranging from fuel
Accepted 29 December 2017
switching to the investment in underground transport and hydropower capacity expansion were
modelled and assessed to evaluate their long-term effect on fossil fuels demand reduction. Our results
show that the full implementation of strategies can account for cumulative oil savings of over 550 million
Keywords:
Fuel saving strategies
barrels of oil by 2030. Findings also point out that Ecuador and Colombia could face the transition from
Oil trade balance net oil-exporters to importers as soon as the end of the next decade.
Peru © 2018 Elsevier Ltd. All rights reserved.
Colombia
Ecuador

1. Introduction industrialization and economic diversification and create jobs and

The vulnerability of a country to oil price fluctuations depends
on whether or not it is a net oil importer or exporter [1e4]. On one
hand, oil importing countries, historical evidence suggests a
negative relationship between oil price increases and economic
development [5e7]. The higher the import bill relative to GDP, the
larger the exposure of an economy and the macroeconomic
adjustment costs to oil price increase [8]. For oil exporting coun-
tries on the other hand, windfall revenues are a significant source of
foreign currency and income. These revenues stimulate economic
activity and finance productive and social investments. However,
they can also promote unsustainable consumption that leads to
fiscal crises [9]. Some oil-exporting countries apply energy sub-
sidies to distribute benefits across the population, promote
* Corresponding author.
E-mail address: pablo.carvajal.14@ucl.ac.uk (P.E. Carvajal).
enhance competitiveness [10]. The efficiency of domestic in-
vestments based on oil revenues is constrained by the productivity
of tradable sectors [11]. For instance, scaling up public investments
too high and too fast could expose the economy to instability due to
absorptive capacity constraints, possible Dutch disease and risks of
currency depreciation [12,13]. Besides, for oil-exporting countries,
high energy consumption can lead to a rapid erosion of their export
capacity, increasing the risk of becoming net importers.
Crude oil plays an important role in the economy and trade of
Peru, Colombia and Ecuador. These so called ‘Andean Countries’
(due to the Andes Mountain range), are located in north-western
South America and are home to approximately 95 million people
(World Bank, 2015). These countries have experienced steady
economic growth over the last two decades as evidenced by
average GDP growth rates from 3% to 5% from 1990 to 2012,
reaching even higher rates in more recent years (World Bank,
2013). Such economic growth has led to increases in gasoline and

https://doi.org/10.1016/j.esr.2017.12.011
2211-467X/© 2018 Elsevier Ltd. All rights reserved.
36 M.F. Chavez-Rodriguez et al. / Energy Strategy Reviews 20 (2018) 35e48
diesel consumption, mostly associated with vehicle ownership and
thermal power generation [14e16]. In the last decade, oil produc-
tion in the Andean countries increased. Peruvian petroleum pro-
duction grew 46%, reaching 168 kbpd (thousand barrel per day) in
2013 [17,18], mainly due to the production of natural gas liquids in
the Camisea project in the Amazon region. In the case of Colombia,
for the same period, oil production increased in 145% reaching
1088 kbpd, mainly due to the exploration and discoveries in the
Llanos Orientales region [19]. In Ecuador, the production increased
in 85% reaching 526 kbpd [20] mainly due to the production in the
north-eastern Amazonia region.
Huntington [21] highlights the relevance to model energy sce-
narios as a way to better understand energy systems and provide
quantitative insights. Energy scenarios about oil supply-demand
are relevant tools for policy-makers to define targets of energy ef-
ficiency programs, fuel switching incentives and their overarching
national energy plans. The Andean countries lack of scientific
literature related to energy modeling and scenarios. Recent studies
in Colombia have focused on transport policies to improve energy
use [22], analyse measures for the short-term security supply of
natural gas [23] and renewable energy potential assessments
[24e26]. In the case of Peru, energy-related scientific literature also
addresses tapping of renewable energy resources [27], energy
poverty [28], liquefied natural gas exports [29], cogeneration po-
tential [30] and oil supply scenarios [31]. In Ecuador there are
studies related to renewable energy resources [32,33], discussions
about oil production in the Yasuni reserve [34,35] and an analysis of
the country's energy development under the Kuznets curve [36].
However, to our knowledge we could not find published scientific
literature with focus in the Andean countries regarding oil products
demand and corresponding policies related. What are the energy
strategies and policies that these governments enforce regarding
oil products consumption? How would fuel demand in Andean
countries evolve in the future with the implementation of these
policies? These are questions that this study aims to provide in-
sights for.
The objective of this study is to model a full set of demand-side
strategies and actions that are being implemented to replace and
lower oil (fuel) consumption, aiming at energy security, diversifi-
cation and mitigation of climate change in Peru, Colombia and
Ecuador up to 2030. These policies are within the Nationally
Determined Contributions (NDCs) that each country presented in
the United Nations Climate Change Conference held in Paris
in2015.1 We use a bottom-up modeling method combining several
technics that serve as input to the Long-range Energy Alternatives
Planning System (LEAP). By using bottom-up models we quantify
the effectiveness of current strategies by assessing how much oil
these would save in the long term. The analysis provides new in-
sights on the impact of national fuel-saving strategies, their trade-
offs and interactions with each other. The study also sets a
benchmark of quantitative data for regional policy-makers to
address the challenges of reaching peak oil and becoming net oil-
importing countries. The study is innovative for the region in the
sense that oil demand and demand-side policies are often over-
looked due to the traditional abundance of energy resources. This
study reveals the significant potential that implementing fuel-
saving strategies can have in terms of economic value as well as
for extending the treasured net oil-exporter status. This paper is
organized as follows. The next section compiles current fuel de-
mand strategies in the Andean countries. Section 3 presents the
methodology and details scenario storylines. Section 4 shows the
main findings. Finally, the last section highlights the main
1
http://www4.unfccc.int/submissions/INDC/.
conclusions of the work and proposes insights for further research.
2. Current fuel demand strategies in the Andean Countries
This section presents a compilation of the fuel saving strategies
and measures, which are currently being implemented in Peru,
Colombia and Ecuador, and will support the creation of scenarios
further below. The analysed strategies are divided in 4 sectors:
residential, services and industry, transport, and electricity gener-
ation and consumption. Table 1 summarizes the energy saving
strategies by country and sector. Fig. 1 displays the final energy
consumption by sector in the year 2015 for Colombia, Peru and
Ecuador. Colombia accounts the highest consumption with 1.2
million TJ, while the final consumption in Peru and Ecuador was
0.8 TJ and 0.56 TJ, respectively. The highest consumption of energy
was from the transportation sector for the three countries.
2.1. Residential sector
The strategies discussed here include only those that directly
affect the consumption of oil products, natural gas and electricity
consumption for heating and cooking purposes. Actions that reduce
electricity consumption in this sector are discussed in section 2.4.
Peru is implementing a natural gas program, which involves a
liquefied natural gas (LNG) distribution system from the Pampa
Melchorita plant to eleven cities in Peru [37]. The project plans to
increase access from 400 thousand households in 2015 to 1.2
million by 2020. In addition, the National Program of Family Cook
Stoves e “Cocina Peru” aims to distribute 1 million liquefied pe-
troleum gas (LPG) cook stoves by 2016, replacing the use of fire-
wood within the framework of the National Plan for Universal
Access to Energy 2013e2022 [38].
In Ecuador, the government is financing the substitution of LPG
cook stoves by electrical induction cook stoves. In 2013, 92% of LPG
supply was consumed by the residential sector [39]. The intro-
duction of 2.6 million electrical induction cook stoves is planned
until 2017 to reduce consumption of LPG in the residential sector
[40]. This measure seeks to reduce an approximate spending of US$
522 million per year on LPG subsidies, 75% of which is imported
[41]. In the case of Colombia, there were no policies considered for
the residential sector.
2.2. Services and industrial sector
In Peru, the Referential Plan for Energy Use 2009e2018 [42]
focuses on the optimization and modernization of 60% of the
country's industrial boilers by 2018. This will be achieved through
the implementation of good practices in operation (2% savings), air-
fuel ratio control (3% savings), and purges automation (1%) [43],
which will save the use of fuel oil and natural gas used for steam
production.
In Colombia, the following initiatives will affect industrial en-
ergy demand: a) electricity savings in mechanical drive through the
replacement of motors and improvements in operation; b) opera-
tional optimization of boilers; and c) efficient lighting. The imple-
mentation of these actions are estimated to save 11% of the sector's
final energy consumption [44]. For Ecuador, no fuel oil policies for
the industrial sector were considered in this study.
2.3. Transport sector
For the case of the transport sector, the policies consider effects
on the demand of oil products, natural gas and electricity
consumed for mobility purposes. In Peru, the electric transport
systems of Lima and Callao were approved in December 2010 [45].
 M.F. Chavez-Rodriguez et al. / Energy Strategy Reviews 20 (2018) 35e48 37

Table 1
Modelled fuel-saving policies by country and economic sector.

Peru Colombia Ecuador

Residential Sector - Natural gas in households (by 2015, 0.4
million new connections, and by 2020, 1.2
million new connections) [47]
- “Cocina Peru”: Distribution of 1.0 million LPG
cookers kits until 2016.
(R.M. 203-2013-MEM/DM).
Industrial, Public and - Operational improvements in 60% of
Services Sector industrial boilers by 2018.
Transport Sector - “Natural Gas Dissemination” in Transport
(250 NGV stations by 2015, and 400 NGV
stations by 2020) [47].

5 new Metro Lines in Lima.
- Monorail in Arequipa.
Electricity supply and - Lighting Program (1.5 million and 0.5 million
demand LCFs in households and public buildings by
2016; 0.1 million LEDs in public lighting).
- NUMES Plan for Power Generation (Aims an
energy mix for Power Generation in 2040:
40% NG; 40% Hydro; 20% Renewable Energy)
- PROURE measures (an estimated
reduction of 11% and 10% in the energy
intensity for the industrial and services
sector) [44].
- PROURE- Improvement of the transport
fleet as well as the establishment of
better driving techniques(an estimated
reduction of 1.29% in the energy intensity
for all transport technologies) [44].
- Increase of biofuels blends: E-20, and B-
20 by 2020 [53].
- Metro de Medellin's expansion and a new
metro for Bogota [54,55].
- Replacement of 32 million incandescent
lamps by LFC [44].
- Substitution of 2 million old refrigerators
by efficient refrigerators [44].
- “Efficient Cooking” Program: substitution
of 2.6 million LPG cookers by electrical
induction cookers by 2017 [58].
- Quito Metro.
- Establishment of biofuel blends:
“Ecopaís” program, E-5 by 2017, and
biodiesel mandatory blend, increasing
from 5% (2013) to 10% (estimated by
2020) [49].
- Hydro Power Generation Expansion
(more than 90% of electricity produced
by hydro power plants in 2017).

A first line initiated operation in 2011 and its expansion was in-
augurated in 2014. A second line is under construction and other
four lines are in the process of study/bidding. In addition, a
monorail transport system for Arequipa's metropolitan area is
currently under evaluation [46]. Finally, the government aims at a
goal of 400 natural gas vehicles (NGV) stations by 2020 under the
framework of the natural gas program [47].
In Ecuador, public transport initiatives like Quito's metro and
Cuenca's tram system will use electricity from 2020 onwards, dis-
placing the use of liquid fossil fuels. In 2010, the Government of
Ecuador launched a pilot program to have a 5% bioethanol blend for
gasoline and intends to extend the program to a national scale by
2017 [48,49]. By the same token, on September 2012, Executive
Decree 1303 formally declared as a national interest the develop-
ment of biofuels to promote the agricultural sector. It also estab-
lished a mandatory biodiesel blend of 5% and a mandate to increase
the blend to 10% in 2020.
In Colombia, the main initiatives in the transport sector are the
technological upgrade of the automotive fleet and promotion of
better driving techniques, leading to a reduction in energy intensity
(GJ/distance) of 0.33% and 0.96%, respectively [44]. Additionally
[50], provides policy guidelines to promote sustainable production
Fig. 1. Final energy consumption by sector in 2015 (TJ).
of biofuels in Colombia [51,52]. Accordingly, Decree 2629 of 2007
stipulates that the new vehicles would have to be adapted to use
interchangeably gasoline or a blend composed of 80% gasoline and
20% ethanol (E
20) [53]. In the case of diesel vehicles, these should
be adapted to use a minimum of 20% biodiesel (B-20) [53]. Future
projects for Medellin include new tramways and new railroads [54].
A metro system is projected for 2020 in Bogota, covering 26.5 km
and with an estimated demand of 800,000 passengers per day [55].
2.4. Electricity supply and demand
In Peru, the National Plan for Universal Access to Energy
2013e2022 [38] considers a program to replace incandescent
lamps by compact fluorescent lamps (CFL). Furthermore, a law
project to progressively eliminate the import, distribution, pro-
duction, retailing and use of incandescent lamps by 2020 is being
analysed [56]. In Ecuador, a similar law was launched in 2008
which set a tax on incandescent lightbulbs and introduced more
that 16 million CFL which phased-out the majority of incandescent
light bulbs in the residential sector. In Colombia, initiatives to
replace light bulbs and domestic refrigerators are expected to affect
electricity demand in the residential sector [44]. The objective of
the first is to replace 32 million incandescent light bulbs with CFL in
low-income households, reaching 20% of total light bulbs in
Colombian households. The objective of the second is to replace
and scrap two million refrigerators. Similarly, for the commercial
sector, an energy efficiency program in lighting, refrigeration and
air conditioning is estimated to save 10% of the electricity
consumed by the commercial sector [44]. The governments of these
countries have made the decision between CFL and LED with their
pocketbook. CFL are cheaper while LED costs more at purchase, but
last longer. Another argument is that people decide that CFL
lighting is closer to what they're used to and therefore choose these
bulbs to replace incandescent bulbs as they blow out.
For electricity supply, Peru's Ministry of Energy and Mines has
commissioned to tender 1100 MW of hydroelectric power plants
that should come on-line by 2018/2019 [47]. As a long run strategy,
according to MEF (2011), the structure of the electricity by 2040
38 M.F. Chavez-Rodriguez et al. / Energy Strategy Reviews 20 (2018) 35e48
mix should be: 40% hydroelectricity, 40% natural gas, and 20% non-
conventional renewable energy (NCRE). In the case of Ecuador, the
main strategy to change the power generation mix is to increase the
share of hydropower. By 2017, large hydroelectricity facilities
known as “flagship” projects will add 2700 MW to the grid. There is
also a potential for using natural gas to replace diesel units [57].
Both in Peru and Ecuador, new hydropower is expected to replace
thermal generation as well as expand overall total capacity.
3. Methods
Our goal is to assess the ability of oil demand reduction policies
to respond to the challenging oil supply-demand balance that An-
dean countries may face in the near future by developing “what if”
energy supply and demand scenarios. To achieve this, we used LEAP
as the core model,2 but complemented with econometric fittings
and other top-down methods. To assess oil supply curves, for
example, a Multi e Hubbert model was used to build oil supply-
demand curves. The results derived from LEAP simulations are
not a product of any kind of optimization runs. An optimization
model would indicate the least-cost option of supply and demand
to cope with energy services related to oil products [61,62] How-
ever, given the lack of information regarding prices and the het-
erogeneity of policies for each country, an energy system
simulation model, such as the ones proposed here, allows easily to
evaluate the impacts of energy policies already in course on the fuel
supply-demand balance [63]. Fig. 2 shows the different data sour-
ces and methodological approaches used, which will be explained
in greater detail in following sections. Data was collected from
national energy balances, energy consumption surveys, national
census, infrastructure reports and national energy plans. The
baseline year is 2010 while all scenarios have the same time frame:
the first simulation year is 2011 and the last one is 2030. Detailed
information on sources used for modeling of the base year is found
in the Supplementary Material. Some of the activity parameters
such as transport fleet projections were calculated exogenously and
then input to LEAP. Using energy intensities, activity level and the
shares for each fuel, LEAP can estimate the final consumption of
fuels by sectors. This fuel consumption is aggregated for each
country for the oil supply-demand balance, the oil supply was
projected using the Multi-Hubbert model which will be explained
as follows.
3.1. Oil supply projections
This study used Multi-Hubbert modeling of Refs. [31] and [64]
for projecting oil supply in the cases of Peru and Colombia,
respectively, which is a curve-fitting variant of the Hubbert model
[65]. The main assumption behind this model is that it is limited by
the effects of information (positive) and depletion (negative).
Through this Multi-Hubbert model [66], oil production modeling
was refined by incorporating further resource developments re-
flected in more than one cycle and a peak. The following equations
describe this approach:
X
N
2:Qmi
P¼
i¼1
1 þ cosh½
bi :ðt
tmi Þ
2
LEAP has been applied in different Latin American Countries. See, for instance,
[60].
X
N
URR ¼ URRi
i¼1
In which URR is the ultimate recoverable reserve; Qm peak
production; tm the time of the peak; b and c are parameters which
account for the slope of the curve and the estimated average useful
life of fields, respectively; i represent each cycle, and N the total
number of cycles modelled. We consider 2P reserves
(proven þ probable reserves) as the URR for the conservative sce-
nario and a higher URR for the optimistic scenario. For Ecuador, a
similar3 Multi-Hubbert analysis was made based on global histor-
ical production taken from Ref. [20] and reserves data from
Ref. [67]. The resources used in the composition of the URR are only
limited to conventional oil resources, as shale and tight oil explo-
ration in these countries is still scarce (see Table 2).
3.2. Demand modeling: baseline and policy scenarios
Two scenarios were modelled with a time horizon of 2030: a
baseline scenario based on historic trends, and a policy scenario,
which reflects the effect of the implementation of fuel-saving
strategies (as described in section 2). Although some of these
strategies are already underway and, therefore, could be considered
as baseline, there is a large degree of uncertainty as to the extent to
which they will actually be fully implemented. Also, this paper
endeavours to evaluate the effects of these measures in terms of
reductions in the consumption of oil products and, as such, a
baseline in which these measures are not considered is needed for
comparison basis. For the residential and transport sectors, a
bottom-up approach was adopted using sectorial surveys to model
energy services, whenever possible. A top-down approach was
employed to model industrial, services and public sectors according
to the countries' energy intensity by energy sources. Power gen-
eration capacity expansion was based on either maintaining the
fuel and technology mix constant or considering national plans
[68e70]. To model fuel consumption in the power sector an
approach based on technical characteristics such as effective ca-
pacity, energy production and conversion efficiency by technolo-
gies and by sources was adopted.
3.2.1. Residential sector demand
For the baseline scenario, household energy consumption was
based on official population/household projections4 (activity effect)
and on a constant energy intensity and fuel mix (intensity and
structure effects). For the policy scenario in Peru, the natural gas
program for increasing natural gas connections for households by
2016 was modelled displacing LPG. The policy initiative to
distribute 1 million LPG cookers by 2016 was considered to replace
firewood and biomass. For water heating, the same fuel proportion
estimated for cooking was used [74]. For Ecuador, the policy sce-
nario considers, the substitution of 2.6 million LPG cookers by
electrical induction cooks stoves by 2017. The values for con-
sumption of LPG and electrical induction cook stoves by household
were based on Ref. [75]. No strategies for the residential sector were
modelled for Colombia.
3.2.2. Transport sector demand
The total energy demand for the transport sector that is used as
an input for LEAP was calculated using the ASIF methodology [76],
3
See Chavez-Rodriguez et al. (2015) and Gonzalez et al. (2015) for further details
about the Hubbert approach used.
4
Refs. [71], [72], and [73].
 M.F. Chavez-Rodriguez et al. / Energy Strategy Reviews 20 (2018) 35e48 39

Fig. 2. Methodological steps followed in the study.

Table 2
Ultimate recoverable resources adopted in the Multi-Hubbert modeling by country.

Country URR-Conservative Scenario (Million barrels) URR-Optimistic Scenario (Million barrels)

Peru 3.89 7.41

Colombia 26.02 28.44
Ecuador 6.78 10.25

Source: Chavez-Rodriguez et al. (2015), MRNNR (2014), OLADE (2014a), Gonzalez et al. (2015).

in which energy demand is estimated according to the following
equation:
XX
Ei ¼ Ai;j;k Si;j;k Ii;j;k
j k
Where i is type of fuel, Ei is fuel consumption (gal/year, m3/year), j is
vehicle mode, k is vehicle type, A is activity (use of the vehicle e
km/vehicle-year), S is structure (number of vehicles by modal e
vehicles) and I is energy intensity (average efficiency by vehicle e
gal/km, m3/km).
The projections of vehicle fleets in Peru and Ecuador were
derived from the model of Ref. [77]. Vehicle ownership is modelled
using a Gompertz model based on per-capita income fitted using
pooled time-series (1960e2002) and cross section data for 45
countries [77]. This approach also explicitly models the vehicle
saturation level as a function of the country's urbanization and
population density. GDP per-capita projections were obtained from
Ref. [78]. For Peru and Ecuador, share of vehicle types was projected
based on historical data using autoregressive models. For Colombia,
the projections of vehicle by types of Ref. [79] were adopted. Since
most of vehicles in these countries are imported, this study
assumed fuel consumption per mile targets established by the US
Environmental Protection Agency [80,81] and extrapolated up to
2030 to project the improvements in fuel economy of vehicles in
the baseline scenario. In Peru's baseline scenario, the NGV fleet was
kept constant after 2013 and the policy scenario modelled new 400
NGV stations, as this is the metric of the government to expand
natural gas use in transport [47]. In this scenario it was assumed
that the ratio natural gas vehicle to NGV station would remain 760,
which resulted from a linear least square fit based on data from
Ref. [82] for the 2005e2013 period. This increase in NGV vehicles
was distributed according to each type of vehicle licensed in 2012
[83]. To estimate the impact of alternative modes, passenger-km/
vehicle-year and load factor (passenger/vehicle) values reported
by Refs. [84] and [85] were taken as reference.
40 M.F. Chavez-Rodriguez et al. / Energy Strategy Reviews 20 (2018) 35e48
For Peru's policy scenario, the passenger demand for the Metro's
second line was obtained from Ref. [86]. As for the other 4 new
Metro lines, we used the passenger demand estimated in Ref. [87].
The length of each Metro Line was taken from Ref. [88] assuming an
average travelled distance equal to half the length of the respective
lines. For the Arequipa monorail, the information indicated by
Ref. [46] was used for the passenger demand and the length of the
projects. The start of operation of Line 2 was projected according to
Ref. [86]. The passenger-km displacement of the metro projects was
divided as follows: 80-70% public transport users, and 20e30%
particular vehicle users, depending on the deployment of new
lines.5 The distribution of passenger-km shifted to the metro sys-
tem was deducted from vehicles (excluding the NGV vehicles) ac-
cording to their participation in the passenger-km by category
(public transport and particular use). However, the effects of new
metro lines on traffic congestion reduction were not considered.
The same approach was used for estimating passenger-km for the
metro of Quito, in Ecuador. Biofuel blends were modelled for
Ecuador considering that for the baseline scenario, the percentage
of ethanol commercialization in 2012 remained constant. For the
policy scenario, the E
5 program target would be extended to a
national scale by 2017 and a biodiesel blend of B-10 would be
achieved by 2020. For the policy scenario of Colombia, a 1.29%
reduction of the energy intensity for all technologies in the trans-
port sector was assumed to be reached in 10 years, according to
Ref. [44]. Additionally, the mandatory ethanol and biodiesel mixes
stated in Ref. [53] were assumed. New metro lines and other pro-
jects such as gondola lift systems, tramways and a metro for Bogota
were projected in the policy scenario. We adopted estimates of
projected passenger users and average distances travelled regis-
tered by Refs. [90] and [91] for Medellin and Bogota projects
respectively. Data from the companies in charge of the projects
were used to estimate service demand [55,92]. The modal shift,
which is the share of the type the vehicle according to type of use,6
was distributed according to their share in 2010. For jet-fuel and
bunker oil consumption for air and maritime transport, a top-down
methodology was used for all three countries. For the case of Peru
and Colombia, correlations between jet-fuel consumption and GDP
were used. In Colombia, this was also performed for bunker con-
sumption. For the case of bunker in Peru and fuel oil and jet-fuel in
Ecuador, intensities based on national energy balances [15,39] and
GDP were used. No strategies were identified in these countries for
reducing the use of jet-fuel or bunker oil.
3.2.3. Industry and services sectors
The baseline scenario considers current energy intensity and the
future projections for GDP in the three countries [78]. For Peru, the
policy scenario includes operational improvements of 60% of
boilers by 2016 and a reduction of the energy intensity of 3.6% for
the consumption of natural gas and fuel oil in the industrial sector
[42]. For Colombia, the policy scenario considers a reduction of 11%
in the industrial sector and 10% for the commercial sector in the
energy intensity respectively for all energy sources in a period of 10
years [44]. No strategies were found or modelled for Ecuador.
3.2.4. Electricity supply and demand
For all three countries, future electricity demand considered in
5
The assumed modal shift from public transport is higher than the reported by
Ref. [89] for the Madrid metro (50%), the new metro in Athens (56%), the Croydon
Tramlink (69%) and the Manchester Metrolink (50%). However, it is lower than the
one considered for the Copenhagen metro (72%) and the Quito metro (85%).
6
Example, the modal shift of private vehicles in transport is the distribution
between sedans, SUVs, station wagon, etc.
the baseline scenario is based on national plans which consider a
BAU path in the mid-term [69,70,93]. Although national power
expansion plans of the Andean countries intend to increase the
share of hydropower and NCRE in the electricity generation mix,
the baseline scenario is based on a conservative deployment of
hydropower and larger shares of thermoelectricity. In the short
term, for all countries, the electricity generation mix is fixed and
determined by projects that have been committed and are under
construction. For the policy scenario, this study considered that the
government intentions would be materialized through the adop-
tion of incentive mechanisms, e.g. feed-in tariffs, auctions, that
would increase the share of NCRE. For the case of Peru, this means a
capacity expansion rate that would result in an electricity genera-
tion mix of 40% hydro, 40% natural gas and 20% NCRE in 2040 [69].
For Ecuador, the policy scenario assumes the expansion according
to Ref. [68], including hydropower plant Coca Codo Sinclair
(1,500 MW), the largest in the country so far and to be fully
commissioned in 2017. For Colombia, the baseline and policy sce-
nario was based on the expansion plan of Ref. [68] and have the
same shares of participation of technologies since no policies have
been set in place for NCRE deployment. Fig. 3 shows the mix of
sources for power generation in all three countries for the horizon
period and according to the base and policy scenarios.
On the demand-side, for Peru, this study modelled the
replacement of 1.5 million and 0.5 million incandescent lamps by
CFLs in households and public buildings respectively, as well as the
installation of around 100 thousand LEDs for public lighting by
2016. For Colombia, the replacement of 32 million incandescent
lamps by CFL was considered. Also in Colombia, the replacement of
2 million old refrigerators by efficient ones in a period of 10 years
was modelled. For public lighting, the values reported by Ref. [94]
for HPS lamps (1179 kWh/year) and LEDs (242 kWh/year) were
applied in both countries. Finally, in Ecuador, the public subsidy for
the acquisition of 2.6 million electrical induction cook stoves by
2017 was considered, leading to an increase in electricity con-
sumption of 100 kWh/month-household [75].
4. Results
4.1. Oil products demand
Baseline scenarios indicate high growth rates for petroleum
products demand in the Andean countries. Fig. 4 shows total fuel
demand by types of oil products, natural gas and alternatives
including biodiesel and ethanol. By 2030, Colombia is the greatest
consumer reaching 250 MMBOE/year,7 followed by Peru over-
passing 200 MMBOE/year and Ecuador with 170 MMBOE. While
Colombia and Peru have overall greater demands offossil fuels in
general (oil and gas), Ecuador has the greatest demand for only oil
liquid products. Ecuador would become the major consumer
among the three countries with 160 MMBOE/year, followed by
Colombia (151 MMBOE/year) and Peru (140 MMBOE/year) in 2030.
Despite Ecuador being the smallest county in terms of surface area
and population among the three countries, it has the largest oil
product demand in the long term, this is explained mainly by the
boost in fuel oil consumption in thermal power plants, in addition
to transport and industrial demand, another reason is the low share
that natural gas has in this country's energy mix (see blue shaded
area in Fig. 4). In Peru, freight transport growth is the main driving
force behind the increasing demand for diesel, but growth of SUV
7
MMBOE: Million barrels of oil equivalent.
 M.F. Chavez-Rodriguez et al. / Energy Strategy Reviews 20 (2018) 35e48 41

Fig. 3. Capacity mix evolution by energy sources projected for Peru, Colombia and Ecuador.

Fig. 4. Oil products, biofuels and natural gas consumption in the baseline scenario for Andean countries.10 Notice that natural gas is not considered in this study as and oil product.
While Colombia and Peru have overall greater demands of fossil fuels in general (oil and gas), Ecuador has the greatest demand for only oil products mainly due to low natural gas
shares in its energy matrix.

sells also contributes to the increase of gasoline consumption8 in the country. Growth rates in Colombia are lower than in the other
two countries due to two factors: firstly, because Colombia has a
more saturated market9; secondly, the high use of natural gas in all
sectors partially mitigates the increasing demand for oil products.
8
This is an ongoing trend in Peru. For example, in 2012 the fleet of SUVs and Fuel demand side measures assessed in the policy scenario can
trucks increased 17% and 8%, respectively [95].
9
Colombia reports highest vehicle/person rates, 183 vehicles/1000 person in
2012, while Peru and Ecuador reports 70.9 vehicles/1000 person and 97.2 vehicles/
10
1000 person for the same year [96]. Numeric values of these projections are shown in the Supplementary Material.
 42
Table 3
Accumulated strategy impacts in energy consumptions by sources from 2012 to 2030 (kboe). Positive sign means that the strategy increases the consumption of certain fuel/resource, a negative sign meaning the opposite.

Country No. Policy GSL DSL FO LPG BDSL ETH NG CO FWD HYD NCRE Fuel Fuel Net
reduction increase change

Peru 1 Natural Gas in
17023 18,991
17023 18,991 1968

households
2 “Cocina Peru”: 1 16,143
56608
56608 16,143
40465
million LPG cookers
3 Operational
594
3417
4011 0
4011
improvements
industrial boilers
4 Natural Gas in
7450
1454
1256 23,636
10160 23,636 13,476
Transport

M.F. Chavez-Rodriguez et al. / Energy Strategy Reviews 20 (2018) 35e48

5 NUMES Plan for
17230
4569 16,580
26842 92,877
48641 92,877 44,236
Power Generation
6 Subway Lines in
4732
6484 62
2462 5210 902 1596
13678 7770
5908
Lima. Monorail in
Arequipa
7 Lighting Energy
383
54
3270 121
3707 121
3586
Efficiency Program
Total by fuel type
12182
25551
5155
4598 0 0 41,150 0
56608
25940 94,594
153,828 159,538 5710

Colombia 1 Replacement of 32
2743
17199
14709
5852
40503 0
40503

million CFL
2 Substitution of 2
426
2481
2239
751
5897 0
5897
million old
refrigerators
3 PROURE - Industrial
6394
5968
2034
95433
57681
202
16845
539
185,096 0
185,096
and services sector
4 PROURE- Transport
9326
6289
547
811
2253
19226 0
19226
fleet and better
driving
5 Metro of Medellin
3368
945 41
82
293
805 559 264
5493 864
4629
and for Bogota
6 Increase of biofuels
86445
59001 59,001 86,445
145,446 145,446 0
blends: E-20, and B-
20 by 2020
Total by fuel type
99139
72629
9096
2034 58,372 85,341
118,171
74070
202
23184
539
401661 146,310
255,351

Ecuador 1 Hydro Power
10002
160,718
72050 106,623
7035
249,805 106,623
143,182
Generation
Expansion
2 Efficient Cooking 3036
59084 6640 25,582
59084 35,258
23826
Program
3 Quito Subway
1061
2005 120 641 1867
3066 2628
438
4 Biofuel blends
27532
68009 68,009 27,532
95541 95,541 0
Total by fuel type
28593
80016
157,562
59084 68,009 27,532
64769 0 0 134,072
7035
407496 240,050
167,446

Total change by
139,914
178,196
171,813
65716 126,381 112,873
141,790
74070
56810 84,948 87,020
962,985 545,898
417,087
fuel for the
Andes region
 M.F. Chavez-Rodriguez et al. / Energy Strategy Reviews 20 (2018) 35e48
Fig. 5. Energy consumption by source between baseline and policy scenarios for Andean countries. Negative values show the amount of fuel saved by the polices compared to the
baseline (zero axis). NCRE and Hydro refer to electricity production by hydropower and non-conventional sources.
partially alleviate the increase in oil demand of the baseline sce-
nario. Table 3 shows the results of modeling fuel saving measures
for oil demand, according to fuel and policy measure. Notice that a
variety of fossil fuels and alternative energy sources are presented,
and not only the sources that are impacted by policy measures. The
reason for this is that we have modelled the development of the
whole energy system for the three countries, therefore in-
terdependencies and trade-off between some measures can be
observed. For instance, while the efficient cooking program in
Ecuador would save LPG, it would also increase electricity con-
sumption through electric stoves and increase fuel consumption for
thermal-power generation.
In Peru, the NUMES Plan for the deployment of sustainable
energy has a significant potential to reduce diesel consumption.
The five new metro lines in Lima and the monorail in Arequipa, as
well as the natural gas program are also relevant for reducing
diesel, gasoline, and LPG consumption. Peru is willing to invest
more in this kind of projects when compared to Colombia and
Ecuador. New metro lines in Colombia affect gasoline demand more
than that of diesel, since the share of gasoline buses in the country
is high. In Peru, the natural gas program for the residential sector
more than compensates the increase in LPG consumption, which
was caused by the ‘Cocina-Peru’ program, which disseminates LPG
cookers to reduce firewood use. Impacts of the lighting energy ef-
ficiency program and the improvements in boilers operation were
modest.
In Colombia, major reductions in oil products demand seem to
be achieved by increasing share of ethanol and biodiesel blends. In
addition, the aggressive targets of Colombian energy efficiency
programs (e.g. PROURE) would lead to oil products savings in po-
wer generation, industrial boilers and heaters, and transport.
Interestingly, 74% of the cumulative impacts of PROURE up to 2030
(6901 kboe of gasoline and 4654 kboe of diesel) derive from better
driving techniques and implementation of good practices. These
are low-cost measures, with an investment of US$ 0.2 million [59].
In Ecuador, hydropower generation expansion has significant
impacts on lowering fuel oil and natural gas consumption for po-
wer generation. In addition, the introduction of induction cook
stoves seems to be a good strategy in the short-term to save LPG
and at the same time absorb the surplus power from hydropower
plants. However, should expected precipitation levels decrease and
43
consequently hydropower generation decrease, the electrical in-
duction cook stoves would cause a dramatic increase in fuel oil,
natural gas and diesel consumption for power generation. Imple-
mentation of the efficient cooking program has however a series of
caveats and complications if the subsidy of LPG remains un-
changed. Households who have access to cheap LPG (US$ 0.1/kg)
would not have an incentive for switching to electricity. This
perhaps explains low implementation records compared to pro-
posed targets. According to Ref. [97], between June 2014 and
November 2015, a total of 223 thousand induction cook stoves
where commercialized in the country. However, the expected
target by the end of 2016 was 1 million cook stoves and 2.6 million
by 2017 when Coca Codo Sinclair enters full operation. In this sense,
a low introduction of electric cook stoves could actually be bene-
ficial if hydropower projects do not enter operation according to
schedule.
Changes in energy consumption between the policy and the
baseline scenarios are depicted in Fig. 5, where the increase or
decrease of and energy source can be seen up to 2030. In Peru,
reductions are achieved mainly for diesel, LPG, firewood and gas-
oline. Hydropower also sees a reduction since it is replaced partially
by natural gas and NCRE. Bagasse also sees an increase in con-
sumption as policies to use biofuels are implemented. In Colombia,
important reductions are observed in gasoline, diesel, natural gas
and coal, and these cumulative savings are the largest among the
three countries, bordering 35,000 kboe by 2030. Increases are
observed however in ethanol and biodiesel consumption due to
fossil fuel blending measures. In Ecuador fuel oil, LPG, diesel and
natural gas consumption savings are emphasized. Natural gas re-
duces due to decreases in production and pessimistic expectations
of further resources. Consumption reduction in Ecuador almost
matches Colombia's bordering 30,000 kboe, this mainly due to the
impressive shift to hydropower and also the use of biodiesel blend.
4.2. Oil supply-demand balance
This section presents the discussion regarding the oil supply and
demand balance considering two oil supply scenarios (conservative
or optimistic) and two oil demand scenarios (baseline and policy
scenarios). Peru has a permanent oil supply deficit during the
modeling horizon (Fig. 6). This highlights the need for a broader
44 M.F. Chavez-Rodriguez et al. / Energy Strategy Reviews 20 (2018) 35e48
Fig. 6. Supply-demand curve for the baseline and policy scenarios for Peru.
fuel-saving strategy portfolio and a strategy to secure international
oil imports at reasonable prices.11 Peruvian oil production is seen to
grow in both scenarios. In a conservative scenario, Peru would
reach a peak of 33 MMboe/year (90.4 kbpd) in 2030 and of 72
MMboe/year (197 kbpd) in 2030 in the optimistic scenario.
In spite of maintaining an oil trade deficit, the implementation
of fuel saving strategies would result in economic benefits for Peru
saving an accumulated of 47.5 MMboe by 2030, however when
assessed annually savings are not that significant and maintain
under 10% annually. Considering a price of US$ 70 per barrel of oil,12
and conservatively assuming that oil products and crude oil have
the same prices,13 cumulative savings would account for US$ 3.3
billion. However, fuel-switching strategies should be carefully
assessed, since they could result in a dramatic increase in natural
gas consumption. It is worth analysing the availability of natural gas
resources as well as the investment needed to convert them into
reserves. Currently, Peru is an exporter of liquefied natural gas.
However, the country may become a natural gas net importer as
early as mid 2030s [29]. In addition, even the conservative oil
supply scenario would struggle to be accomplished in Peru if the
regulatory framework is not improved, the national oil company is
constrained for upstream activities and the environmental conflicts
are not coped with [31,99].
Interestingly, according to our findings, Colombia and Ecuador
have the theoretical capability of contributing to fill out the Peru-
vian oil gap, at least during the next decade. This is seen in Fig. 7
and Fig. 8, which show the oil balance for Colombia and Ecuador.
11
It is out of the scope of this study to detail the issue of how to address Peruvian
oil imports vulnerability. However, a possible strategy would be to invest in more
sophisticated refineries able to process heavy South American crudes, such as
Venezuelan streams.
12
For the sake of simplicity, the price of US$ 70 per barrel of oil will be considered
for the three countries. An average price of US$ 70 was considered for the modeling
period based on the current price of US$ 50 in 2017.
13
Refined product prices move in tandem with crude oil prices. From September
2011 to September 2016, WTI prices fell 47% and gasoline prices, while WTI fell 48%
during the same period. The situation is similar for other refined products as well.
Heating oil spot prices have fallen 54% in the past five years, and diesel prices have
fallen 52% during the same period [98].
For Colombia, the conservative oil supply curve reaches a peak as
soon as 2016 with an oil production of 391 MMboe/year
(1070 kbpd) while in the optimistic scenario the peak would be
reached in 2019 with 459 MMboe/year (1258 kbpd). Nevertheless,
Colombia would still benefit from a relatively positive oil supply
balance in both cases. Besides, Colombia has the possibility of
delaying the country's peak oil by adding reserves from contingent
or unconventional resources [100], not considered in the optimistic
scenario. Colombia can reduce consumption of oil products by 181.6
MMboe accumulated along the period under study, saving up to
US$ 12.8 billion. As Colombian oil saving strategies rely significantly
on increased biofuel production and consumption, it is worthy to
review the sustainable deployment of biofuels in this country and
incentivize strategies for its use [101]. Ecuador has already reached
its peak oil for both the optimistic and conservative scenarios.
Therefore, the prospects for the country's oil supply balance are not
encouraging. Actually, even under the policy scenario, oil demand
will exceed the supply as early as 2020 (conservative oil supply) or
as late as in 2028 (optimistic oil supply). This highlights the urgency
of Ecuador's strategy to diversify the country's energy mix, by
including renewable energy sources. However, an alternative
strategy could also be to increase oil production by tapping not only
governmental but also private investments. In 2004 only 37% of oil
production was carried out by the state oil producing company,
however this value rose to 75% in 2013, indicating the policy shift
towards nationalizing oil resources [14]. The current state policy is
to increase in the short and midterm the proven reserves by
including the southern Amazonian region [102] as well as imple-
menting new technologies for enhanced and improved oil recovery
in existing fields [67]. To this end, the government has organized
the latest roundtable in 2013 to attract private exploration investors
but still has not closed any deals. Ecuadorian fossil fuel-saving
strategies can save up to 325 MMboe accumulated during the
period under study, equivalent to US$ 22.8 billion. Even though
Ecuador is a net energy exporter (primary energy), the country
imports a large share of its oil products consumption [39]. This
explains why the government is currently working on the new oil
refinery project ‘Refinería del Pacífico’ to start operations in 2020
and to process between 150 and 200 kbpd [103].
 M.F. Chavez-Rodriguez et al. / Energy Strategy Reviews 20 (2018) 35e48 45

Fig. 7. Supply-demand curve for the baseline and policy scenarios for Colombia.

Fig. 8. Supply-demand curve for the baseline and policy scenarios for Ecuador.

5. Conclusions and discussion
Currently, Peru, Colombia and Ecuador aim towards imple-
menting fuel saving measures to deal with increasing oil demand
and decreasing oil supply. These include different strategies,
ranging from fuel switching to the investment in metro lines and
hydropower capacity expansion plans. If these governmental pro-
grams are implemented as planned, 47.5 MMboe, 181.5 MMboe,
and 325.0 MMboe of oil products consumption could be saved in
Peru, Colombia and Ecuador, respectively, up to 2030. This is
equivalent to a gross cumulate saving of US$ 39 billion, considering
an average oil barrel price of US$ 70.
Even under policy scenarios it is interesting to see that an
imminent transition from net exporter to net importer is inevitable.
Ecuador appears to be the country at greatest risk, which would
undergo through this transition before 2030, while Colombian oil
supply-demand balance would also be facing shortages towards the
end of this period. However, oil supply projections should be con-
trasted with results of bottom-up models, which are likely more
accurate for the short-term [104]. In any case, in terms of energy
strategy, our results highlight the need of complementing the
current fuel saving measures with more aggressive actions. This is
of course if the political energy sovereignty discourse remains as
the idea of producing enough oil or natural gas endogenously [105].
Greater efforts should be carried out in Peru for saving diesel,
including maintenance of trucks, logistics optimization, and
46 M.F. Chavez-Rodriguez et al. / Energy Strategy Reviews 20 (2018) 35e48
investment in railroads for freight transport. Besides, the country
must deal with the fact that a significant fraction of its light duty
fleet is fuelled by diesel. Promoting electric public transport
(especially metro lines) is the best option, but can take significant
time to implement. Therefore, this measure should be accompanied
by more simple mechanisms such as minimum efficiency standards
for vehicles and even higher taxes for larger cars. In Colombia,
energy efficiency standards should be defined for the light-duty
vehicle fleet. Since gasoline is also used in buses, programs
should have this focus, including fuel switching options (even new
technologies, such as hybrid buses), bus rapid transit in major cities
and new metro lines. Ecuador and Peru, show an increasing trend
for heavier cars that also calls for attention. In addition, Ecuador has
highly subsidized fossil fuels, which discourage fuel saving mea-
sures. Diminishing or eliminating these subsidies is a highly sen-
sitive issue that involves political risks, but should be taken into
consideration.
Fuel switching strategies should take into account that through
integrated energy planning processes it is possible to reshape
electricity and gas delivery systems in a way that minimizes re-
dundancies and saves capital costs. Whereas in the industrialized
world, gas and power regulators historically had little incentive to
coordinate their activities, much of developing countries have the
opportunity to avoid this and construct their new energy infra-
structure in an efficient manner [106].
There are opportunities for energy integration between Andean
countries. Firstly, oil surpluses in Colombia and Ecuador, at least
until 2025, could be diverted towards Peru, at very low trans-
portation costs. Clearly, the characteristics of Peruvian oil-refining
infrastructure and oil stream qualities should be carefully studied
in order to optimize the whole system. For instance, the increasing
demand for more stringent low-sulphur diesel pools in Peru
(50 ppm) could be met by a Peruvian refinery able to process
Colombian oil streams or a Colombian refinery adapted to supply
the specified diesel in Peru. Further research should evaluate the
best option in this case. Secondly, another opportunity for energy
integration relies on information and knowledge transfer. For
instance, the industrial and know-how capacities created by
aggressive plans to build underground transport lines in Peru could
be shared with the other Andean countries. These integration ini-
tiatives could be supported by regional platforms such as OLADE
(Latin American Energy Organization), CIER (Regional Energy
Integration Commission), CAN (Andean Community Nations) and
UNASUR (Union of South American Nations). Interestingly, the
same could be done for defining minimum efficiency standards for
light vehicles across the region, avoiding that one country imports
less efficient vehicles than the other.
The regional application of this study to the Andean region is
unique and could serve as a basis for further studies. For example,
an assessment of the economic impacts of the scenarios simulated
in this study are of great interest. Additionally, this study did not
evaluate the oil downstream industry of the Andean countries,
which composes part of the strategy to deal with oil trade deficits,
including possible integrations between them. As mentioned
before, it is not within the aim of LEAP simulations to evaluate the
cost effectiveness of policies, i.e., to reach a target at the least cost
for society. In this case, the simulated policies were taken from the
countries' energy agendas, and it is difficult to assess if these pol-
icies have been a result of a cost-effectiveness analysis between
different alternatives to lower oil (fuel) consumption, aiming at
energy security, diversification and mitigation of climate change.
Therefore, further studies could assess the cost-efficiency of these
fuel saving strategies if both private and social costs and benefits
over time could be obtained. For this, not only engineering models
devised to test what-if scenarios of energy flows are necessary, but
also optimization models to derive least-cost energy options for the
Andean countries would be interesting to assess. The authors of this
study are now preparing versions of these models in order to assess
least-cost pathways of development for the energy system as well
as generating further insights on price sensitivities for oil and how
these affect the deployment of fuel-saving strategies. Finally,
another aspect for further research is to assess possible rebound
effects of the policies in question, and incorporate such effects in
our modelling [107]. This will contribute to a more tuned outcome.
Acknowledgments
We thank the financial support from CNPq/TWAS (Processes
190318/2011-2) for the full time graduate fellowship that spon-
sored this research for the first author. We thank the Ecuadorian
Secretariat of Higher Education, Science and Technology for the full
time graduate fellowship for the second author. We also would like
to acknowledge the support of the Climacap (Integrated CLimate
Modeling and CAPacity building in Latin America) project for the
early stages of this study. Finally, we thank to Prof. Joyce Dargay for
her transport model data and Jorge Leo
n for his comments.
Appendix A. Supplementary data
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.esr.2017.12.011.
References
[1] A. Cologni, M. Manera, Exogenous oil shocks, fiscal policies and sector real-
locations in oil producing countries, Energy Econ. 35 (2013) 42e57.
[2] L. Kilian, Exogenous oil supply shocks: how big are they and how much do
they matter for the U.S. Economy? Rev. Econ. Stat. 90 (2) (2008) 216e240.
[3] R.E. Looney, Socio-economic planning in an era of falling oil revenues: the
case of Venezuela, Soc. Econ. Plann. Sci. 22 (2) (1988) 57e70.
[4] S.F. Schubert, S.J. Turnovsky, The impact of oil prices on an oil-importing
developing economy, J. Dev. Econ. 94 (1) (2011) 18e29.
[5] A.S. Blinder, J.B. Rudd, The Supply-shock Explanation of the Great Stagflation
Revisited, National Bureau of Economic Research, Dec. 2008. Working Paper
14563.
[6] J. Dargay, D. Gately, The response of world energy and oil demand to income
growth and changes in oil prices, Annu. Rev. Energy Environ. 20 (1) (1995)
145e178.
[7] J.D. Hamilton, Oil and the Macroeconomy since World War II, 1983.
[8] E. Gupta, Oil vulnerability index of oil-importing countries, Energy Pol. 36 (3)
(Mar. 2008) 1195e1211.
[9] M. Mehrara, The asymmetric relationship between oil revenues and eco-
nomic activities: the case of oil-exporting countries, Energy Pol. 36 (3) (Mar.
2008) 1164e1168.
[10] O. Damette, M. Seghir, Energy as a driver of growth in oil exporting coun-
tries? Energy Econ. 37 (May 2013) 193e199.
[11] R. Cherif, F. Hasanov, Oil exporters' dilemma: how much to save and how
much to invest, World Dev. 52 (2013) 120e131.
[12] A. Berg, et al., Public Investment in Resource-abundant Developing Coun-
tries, 2012.
[13] G.A. Davis, J.E. Tilton, The resource curse, Nat. Resour. Forum 29 (3) (2005)
233e242.
[14] MICSE, Balance Energetico Nacional 2015 (Base 2014), 2015. Quito - Ecuador.
[15] Ministerio de Energía y Minas, Balance Nacional de Energía 2012, 2013.
[16] UPME, Boletín Estadístico de Minas y Energía 2008-2012, Unidad de
Planeacio
n Minero Energe
tica, Bogota, Colombia, 2013.
[17] Anuario Estadístico de Hidrocarburos 2013, Ministerio de Energía y Minas
del Perú, Lima, Peru, 2014.
[18] MINEM, Anuario Estadístico de Hidrocarburos 1999, Ministerio de Energía y
Minas del Perú, Lima, Peru, 2000.
[19] ACP, Informe Estadístico de Petro
leo, Asociacio

n Colombiana de Petro
leo,
2014.
[20] OLADE, SIEE, Sistema de Informacio
n Econo
mica-Energe
tica, 2014.
[21] H.G. Huntington, J.P. Weyant, J.L. Sweeney, Modeling for insights, not
numbers: the experiences of the energy modeling forum, Omega 10 (5) (Jan.
1982) 449e462.


[22] J.E. Martínez-Jaramillo, S. Arango-Aramburo, K.C. Alvarez-Uribe, P. Jaramillo-


Alvarez, Assessing the impacts of transport policies through energy system
simulation: the case of the Medellin Metropolitan Area, Colombia, Energy
Pol. 101 (2017) 101e108.
[23] J. Villada, Y. Olaya, A simulation approach for analysis of short-term security
 M.F. Chavez-Rodriguez et al. / Energy Strategy Reviews 20 (2018) 35e48 47

of natural gas supply in Colombia, Energy Pol. 53 (2013) 11e26. [60] OLADE, Manual de Planificacio
n Energe
tica, 2014.
[24] J. Arias-Gaviria, B. van der Zwaan, T. Kober, S. Arango-Aramburo, The pros- [61] S. Jebaraj, S. Iniyan, A review of energy models, Renew. Sustain. Energy Rev.
pects for small hydropower in Colombia, Renew. Energy 107 (2017) 10 (4) (2006) 281e311.
204e214. [62] I. Keppo, M. Strubegger, Short term decisions for long term problems e the
[25] A.A. Radomes Jr., S. Arango, Renewable energy technology diffusion: an effect of foresight on model based energy systems analysis, Energy 35 (5)
analysis of photovoltaic-system support schemes in Medellín, Colombia, (2010) 2033e2042.
J. Clean. Prod. 92 (2015) 152e161. [63] E. Worrell, S. Ramesohl, G. Boyd, Advances in energy forecasting models
[26] H.-E. Edsand, Identifying barriers to wind energy diffusion in Colombia: a based on engineering economics, Annu. Rev. Environ. Resour. 29 (1) (2004)
function analysis of the technological innovation system and the wider 345e381.
context, Technol. Soc. 49 (2017) 1e15. [64] E.M. Gonzalez, A. Szklo, A.F.P. Lucena, Aplicacio
n del modelo Multi-Hubbert
[27] M. Lo
pez, M. Veigas, G. Iglesias, On the wave energy resource of Peru, Energy para la elaboracio
n de escenarios de produccio
n de petro
leo en Colombia,
Convers. Manag. 90 (Jan. 2015) 34e40. 2015.
[28] S. Groh, The role of energy in development processesdthe energy poverty [65] Nuclear Energy and the Fossil Fuels, 1956
penalty: case study of Arequipa (Peru), Energy Sustain. Dev. 18 (2014) [66] J. Laherrere, Multi-hubbert Modeling, 1997.
83e99. [67] MRNNR, Reservas y Proyecciones de Produccio
n de Petro
leo, 2014. Quito,
[29] L. Leung, G.P. Jenkins, An economic evaluation of Peru‫׳‬s liquefied natural gas Ecuador.
export policy, Energy Pol. 74 (Nov. 2014) 643e654. [68] MEER/CONELEC, Plan Maestro de Electrificacio
n 2012 -2021, CONELEC,
[30] R. Gonzales Palomino, S.A. Nebra, The potential of natural gas use including Quito, Ecuador, 2012.
cogeneration in large-sized industry and commercial sector in Peru, Energy [69] Elaboracio
n de la Nueva Matriz Energe
tica Sostenible y Evaluacio
n Ambi-
Pol. 50 (Nov. 2012) 192e206. ental Estrate
gica, como Instrumentos de Planificacio
n, Ministerio de Econ-
[31] M.F. Chavez-Rodriguez, A. Szklo, A.F.P. de Lucena, Analysis of past and future omía y Finanzas, Lima, Peru, 2011.
oil production in Peru under a Hubbert approach, Energy Pol. 77 (Feb. 2015) [70] UPME, Plan de Expansio
n de Referencia Generacio
n e Transmisio
n 2012-
140e151. 2025, Unidad de Planeacio
n Minero Energe
tica, 2012.
[32] G. Villacreses, G. Gaona, J. Martínez-Go
mez, D.J. Jijo

n, Wind farms suitability [71] INEI, Estimaciones y Proyecciones de Poblacio
n Urbana y Rural por Sexo y
location using geographical information system (GIS), based on multi- Edades Quinquenales, Según Departamento, 2000-2015. Boletín Especial No.
criteria decision making (MCDM) methods: the case of continental 19, Instituto Nacional de Estadística e Informa
tica, Lima, Peru, 2009.
Ecuador, Renew. Energy 109 (2017) 275e286. [72] DANE, Estimaciones 1985-2005 y proyecciones 2005-2020 de hogares
[33] M.R. Pelaez-Samaniego, G. Riveros-Godoy, S. Torres-Contreras, T. Garcia- nacional y departamental por a
rea, Departamento Administrativo Nacional
Perez, E. Albornoz-Vintimilla, Production and use of electrolytic hydrogen in de Estadística, 2010.
Ecuador towards a low carbon economy, Energy 64 (2014) 626e631. [73] INEC, PROYECCION
NACIONAL POR EDADES SIMPLES, AREA
Y SEXO,
[34] M.C. Vallejo, R. Burbano, F. Falconí, C. Larrea, Leaving oil underground in PERIODO 2010-2020, 2012.
Ecuador: the Yasuní-ITT initiative from a multi-criteria perspective, Ecol. [74] A. V
asquez, E. Ruiz, F. Coelho, H. Palacios, Informe de Resultados: Encuesta
Econ. 109 (2015) 175e185. Residencial de Consumo y Usos de Energía - ERCUE 2012, Oficina de Estudios
[35] B.K. Sovacool, J. Scarpaci, Energy justice and the contested petroleum politics Econo
micos, OSINERGMIN e Perú, Lima, Peru, 2014.
of stranded assets: policy insights from the Yasuní-ITT Initiative in Ecuador, [75] MEER, Programa Emblema
tico Coccio
n Eficiente, 2014.
Energy Pol. 95 (2016) 158e171. [76] L. Schipper, C. Marie-Lilliu, R. Gorham, Flexing the Link between Transport
[36] A. Robalino-Lo
pez, J.-E. García-Ramos, A.A. Golpe, A.
Mena-Nieto, System and Greenhouse Gas Emissions: a Path for the World Bank, International
dynamics modelling and the environmental Kuznets curve in Ecuador Energy Agency, Paris, 2000.
(1980e2025), Energy Pol. 67 (2014) 923e931. [77] J. Dargay, D. Gately, M. Sommer, Vehicle ownership and income growth,
[37] PROINVERSION, Presentacio
n Proyecto ‘Masificacio
n del Uso del Gas Natural worldwide: 1960-2030, Energy J. 28 (4) (2007) 143e170.
a Nivel Nacional, 2013. [78] IMF, World Economic Outlook Database, Feb-2015.
[38] El Peruano, R.M. No. 203-2013-MEM/DM, May-2013, pp. 495589e495593. [79] N. Di Sbroiavacca, H. Dubrovsky, Metodología y prospectiva a partir de
Lima, Peru. escenarios enege
ticos (2008-2030) realizados con el modelo LEAP: El caso de
[39] MICSE, Balance Energe
tico Nacional 2013, Ministerio Coordinador de Sec- Colombia, CEPAL, Santiago, Chile, 2011.
tores Estrate
gicos - MICSE, Quito, Ecuador, 2014. [80] EPA, EPA and NHTSA Set Standards to Reduce Greenhouse Gases and
[40] MEER, Proyectos de Energia Renovable en el Ecuador, 2015. Improve Fuel Economy for Model Years 2017-2025 Cars and Light Trucks,
[41] BCE, BOLETIN ESTADISTICO MENSUAL NO. 1949-IEM-414, Banco Central del United States Environmental Protection Agency, Office of transportation and
Ecuador, 2014. Air Quality, Washington, DC, USA, 2012. Regulatory Announcement.
[42] Resolucio
n Ministerial No. 469-2009-MEM/DM, Ministerio de Energía y [81] EPA, EPA and NHTSA Finalize Historic National Program to Reduce Green-
Minas del Perú, Lima, Peru, Oct. 2009. house Gases and Improve Fuel Economy for Cars and Trucks, United States
[43] MINEM, Resolucion Ministerial No. 469-2009-MEM/DM. Peru, 2009. Environmental Protection Agency, Office of transportation and Air Quality,
[44] MINMINAS, Programa de uso racional y eficiente de energía y fuentes no Washington, DC, USA, 2010. Regulatory Announcement.
convencionales e proure. plan de accio
n 2010-2015, Ministerio de Minas y [82] Infogas, Evolucio
n de Vehículos Activos vs Acumulado de Estaciones de
Energía, Bogota, Colombia, 2010. Servicio, 2014. Lima, Peru.
[45] El Peruano, Decreto Supremo No. 059-2010-MTC, Dec-2010, [83] MTC-OGPP, NUMERO DE VEHICULOS POR TIPO, SEGÚN ANO ~ Y TIPO DE
pp. 431578e431579. Lima, Peru. COMBUSTIBLE: 2007 Y 2012, Ministerio de Transportes y Comunicaciones,
[46] MTC, Transporte Público Masivo en Ciudades Intermedias del Perú, 2013. Lima, Peru, 2014.
Lima, Peru. [84] R.G. Palomino, Oportunidades e Alternativas para a Mudança da Matriz
[47] MINEM, Presentacion del Ministro de Energia y Minas a la Comision de Energe
tica Peruana: O papel do Ga
s Natural, Universidade Estadual de
Energia y Minas del Congreso de la Republica, 2013. Campinas, Campinas, Brazil, 2012.
[48] Andes, Hasta 2017 gasolina extra sera
reemplazada por Ecopaís, m
as limpia y [85] L. Fulton, H. Kaneko, IEA Mobility Model Lite (MoMo Lite), 2010.
beneficiosa para Ecuador, Oct-2014. [86] PROINVERSION, Estudio de Preinversio
n a Nivel de Factibilidad de la Línea 2
[49] MCPEC, Ecopaís: Una Muestra Del Cambio De La Matriz Productiva, 2014. y Tramo de la Línea 4 del Metro de Lima. Volumen I - Resumen Ejecutivo,
[50] Documento CONPES 3510: Lineamiento de política para promover la Ministerio de Energía y Minas del Perú, Lima, Peru, Jul. 2013.
produccio
n sostenible de biocombustibles en Colombia, Consejo Nacional de [87] JICA, Encuesta de Recoleccio
n de Informacio
n B
asica del Transporte Urbano
Política Econo
mica y Social República de Colombia Departamento Nacional

en el Area Metropolitana de Lima y Callao. Informe Final, Agencia de Coop-
de Planeacio
n, Bogota, Colombia, 2008. eracio
n Internacional del Japo
n, Lima, Peru, 2013.
[51] DNP, Lineamiento de Política para Promover la Produccio
n Sostenible de [88] Consorcio Metro de Lima, La Red B
asica de Metro de Lima, 2013.
Biocombustibles en Colombia, Consejo Nacoinal de Política Econo
mica y [89] G. Vuk, Transport impacts of the copenhagen metro, J. Transport Geogr. 13
Social, Departamento Nacional de Planeacio
n, Bogota, Colombia, 2008. (3) (2005) 223e233.
[52] UPME, Biocombustibles en Colombia, Unidad de Planeacio
n Minero Ener- [90] Alcaldia de Medellin, Medellín, la ciudad de la movilidad inteligente. Balance
ge
tica, Bogota, Colombia, 2009. de Gestio
n: Enero de 2008 - noviembre de 2011, Secretaría de Transportes y
[53] MINMINAS, Decreto 2629 de 2007, Ministerio de Minas y Energía, Bogota, Tra
nsito de Medellín, Medellin, Colombia, 2011.
Colombia, 2007. [91] Alcaldia Mayor de Bogota, MOVILIDAD EN CIFRAS 2010, 2011. Bogot
a,
[54] Plan Maestro 2006-2030, Empresa de Transporte Masivo del Valle de Aburr
a Colombia.
Limitada, Medellin, Colombia, 2014. [92] Metro de Medellín, Plan Maestro 2006-2030, Empresa de Transporte Masivo
[55] Metro en Bogota
, En seis an ~ os bogota
tendra
metro [Online]. Available: del Valle de Aburra
Limitada, Medellin, Colombia, 2014.
www.metroenbogota.com/movilidad-bogota/metro-de-bogota, 2014. [93] MEER/CONELEC, Plan Maestro de Electrificacio
n 2012-2021, CONELEC, Quito,
[56] DGEE/MINEM, Oficio No. 2376-2013-MEM/SEG, 2013. Ecuador, 2012.
[57] ARCONEL, Transacciones del Sector Electrico Ecuatoriano - Estadistica del [94] NOWAXLED, Tabela Comparativa HPSxLED, 2014.
Sector Electrico Ecuatoriano 2014 B, 2014. [95] MTC-OGPP, PARQUE VEHICULAR NACIONAL ESTIMADO, SEGÚN CLASE DE
[58] MEER, Programa Emblema
tico Coccio

n Eficiente. VEH
ICULO: 2003e2012, Ministerio de Transportes y Comunicaciones, Lima,
[59] MINMINAS, PROGRAMA DE USO RACIONAL Y EFICIENTE DE ENERG
IA Y Peru, 2014.
FUENTES NO CONVENCIONALES e PROURE. Plan de Accio
n 2010-2015, [96] CAN, Parque Vehicular en la Comunidad Andina 2004-2013, Comunidad
Ministerio de Minas y Energía, Bogota, Colombia, 2010. Andina, Lima, Peru, 2014.
48 M.F. Chavez-Rodriguez et al. / Energy Strategy Reviews 20 (2018) 35e48
[97] El Comercio, Cocinas chinas ponen en apuros al productor local, Actualidad-
Economía, Nov 2015.
[98] Market Realist, How much are refined product prices affected by crude oil
Prices? - market realist [Online]. Available: http://marketrealist.com/2016/
11/refined-product-prices-affected-crude-oil-prices/, 2016. Accessed: 19-
Mar-2017.
[99] J. Haselip, Transparency, consultation and conflict: assessing the micro-level
risks surrounding the drive to develop Peru's Amazonian oil and gas re-
sources, Nat. Resour. Forum 35 (4) (Nov. 2011) 283e292.
[100] Technically Recoverable Shale Oil and Shale Gas Resources: an Assessment of
137 Shale Formations in 41 Countries outside the United States, U.S. Energy
Information Administration, Washington, USA, 2013.
[101] C. Castiblanco, A. Moreno, A. Etter, Impact of policies and subsidies in agri-
business: the case of oil palm and biofuels in Colombia, Energy Econ. 49
(2015) 676e686.
[102] SHE, Ronda Suroriente Ecuador: Aspectos Te
cnicos, Jurídicos, Econo

micos Y
Socio-ambientales, Secretaría de Hidrocarburos Ecuador, Quito, Ecuador,
2013.
[103] The Oil and Gas Year Ecuador 2013, The Oil and Gas Year, 2014.
[104] R.W. Bentley, The expected dates of resource-limited maxima in the global
production of oil and gas, Energy Effic. 3 (2) (Dec. 2009) 115e122.
[105] J. Jewell, Energy Security and Climate Change Mitigation: the Interaction in
Long-term Global Scenarios, Central European University, 2013.
[106] H.D. Saunders, The unexploited potential for natural gas to greatly increase
energy efficiency, Energy Effic. 8 (2) (Aug. 2014) 403e415.
[107] S. Sorrell, J. Dimitropoulos, M. Sommerville, Empirical estimates of the direct
rebound effect: a review, Energy Pol. 37 (2009) 1356e1371.