2. SHADING LIGHT ON ENERGY INTEGRATION

There is no precise definition of energy integration in the current literature, suggesting that it should be interpreted as a process that involves at least two countries and that aims at promoting regional energy industry through a permanent installation and based on a specific agreement that guides the relationship rules between the parties (^{Oxilia, 2009}). However, it can be argued that 'energy integration and, more widely, infrastructure integration, represents the cornerstone of a new level of regional integration', that is perfectly in line to the central argument of this paper (^{Lima and Coutinho, 2006}: 363).

States deal with this sector by taking into account ordinary notions such as energy planning, diversification of energy matrix, and energy self-sufficiency (^{Santos, 2014}). We note that such concepts are equally important when linked to energy security. Particularly, the concept of self-sufficiency ends up being a great barrier to the promotion of regional energy integration.

In the case of countries in South America, particularly the Mercosur countries, the possibility of exploring synergies derived from hydrological complementarity, as well as different sources, has been highlighted (^{Moura, 2017}; ^{Ramos, 2016}). Thus, there is evidence of a strong complementarity between the different pluviometric regimes in the region, which suggests the joint planning of the dispatch of hydroelectric dams, construction of new ventures and joint management of decision-making.

Again, in the context of South America and particularly the Southern Cone, there is a diversity of financial sources coming from regional and international financial institutions that have, among their portfolio of projects, those involved in the issue of energy integration (^{Padula, 2013}; ^{Santos, 2014}), such as the Inter-American Development Bank (IADB), the World Bank (WB), the Brazilian Development Bank (BNDES), the Development Bank of Latin America (CAF) and the Financial Fund for the Development of the Río de Plata Basin (FONPLATA).

The literature often stresses the importance of Brazil in the region's energy integration process. This should be due to (i) being the largest energy market in the region, with a consistent economic model for expanding production capacity; (ii) having borders with 10 of the 12 countries in South America; and (iii) its previous expertise in national energy integration, with the creation of the National Interconnected System (SIN) (^{Santos, 2018}).

In this sense, the obstacles to regional energy integration begin to appear. In addition to the issue of Brazilian hegemony in the region, there is a significant lack of convergence and consensus on political, macroeconomic and microeconomic issues, which makes any regional integration initiative a major challenge (^{Santos, 2014}). This reality ends up being reflected in the asymmetry of development and technical and technological power between Brazil and its neighbors, especially the smaller ones (^{Biato et al., 2016}).

An obstacle to this broader integration process is the Brazilian commercial model applied since 2004, since it relies on the sale of electricity certificates (physical guarantee), defining a closed, planned and operated model in an optimized and centralized way. It is not trivial to incorporate into this model the energy imported from other countries, unless there is contractual and legal security to consider it in the long-term Brazilian energy planning.

In general, there are institutional and regulatory asymmetries that make the implementation of energy integration projects in the region too complex and costly (^{Queiroz et al., 2013}). This complexity and diversity should be reduced in order to minimize economic uncertainties, legal insecurities and political risks.

Since the regulatory framework of South American countries was based on different experiences in time and space, the current model presented by each country is quite different, especially regarding the environment, opening up to private (and foreign) capital and strategic planning of the sector (^{Santos, 2014}). Consequently, institutional conditions in the region still have a major influence over the technical, commercial and contractual relations in the integration process (^{Queiroz and Vilela, 2010}).

In order to avoid political risks and 'loss of sovereignty' in the majority of extra-national energy projects, countries seek to promote enterprises of binational nature. However, the logic of these actions has almost always been subordinated to the interests of national energy planning and not to an integrated and systemic policy for the whole region; in other words, each country individually designs its annual energy plan, its investment prospects, and its short/medium term interests (^{Santos, 2014}).

As 'side effects', energy integration can lead to reduced energy self-sufficiency (energy interdependence) and the operational autonomy of systems, which is one of the main resistance to its implementation. Moreover, given the need for harmonization of regulatory frameworks, the complexity of defining legal frameworks, treaties, norms and rules, as well as possible shocks in diplomatic relations between countries, should be highlighted. It is also possible to stress the increase in risks related to changes in the previously agreed conditions, through interventions by governments, such as through the definition of new legislation and taxation (subsidies and administered prices) and the expropriation of assets.

From the existence of different benefits and barriers to energy integration in Mercosur, we established that it would be necessary to deal with issues of commercial, operational and institutional natures. Events such as nationalization of assets (Bolivia and Venezuela), interruption of contracted energy supply (Argentina to Chile, and Venezuela to Roraima, and Petrocaribe) and request for renegotiation of the agreement signed (Paraguay from Brazil, in the case of Itaipu) created a bad and pessimistic history for the advancement of the process.

3. SCENARIZATION AND QUANTITATIVE MODELING

After briefly analyzing benefits and barriers to energy integration, this section aims to model and analyze possible scenarios of energy integration in the region, using the case of the power sector to illustrate its potential. The power sector was considered as a case study of energy modeling, given the social relevance of ensuring access to electricity at affordable prices, particularly when it comes to developing countries.

The scenarios were modeled in the OSeMOSYS-SAMBA. This model provides long-term cost-optimization of the power expansion planning of South America countries, being an open source, dynamic, bottom-up and multi-year power sector framework that allows us to deal with large-scale linear programming problems.

Data on existing regional infrastructure was used and the expansion plans of the countries. The base-year is 2013, with four scenarios built for the period 2013-2050. Features related to population growth, electricity demand, costs, hydro reservoirs, technological performance, reserve margin time zones and carbon emissions were considered. 2013 is a strategic base-year for the model because from that date we have to consider that the United Nations General Assembly unanimously declared the decade 20142024 as the Decade of Sustainable Energy for All (SE4ALL). This stresses the relevance of energy issues for sustainable development and for the elaboration of the post-2015 development agenda.

The following features are key assumptions of the model: (i) technological changes are provided by exogenous learning curves based on IEA ETP reports; (ii) time resolution is 12 months, divided into 4 intra-day periods and time horizon is 2013-2058, yearly steps; (iii) reserve margin is 15% (only dispatchable technologies are able to meet it); (iv) real discount rate is 8% and monetary values is 2013 US$; (v) there are also three time zones: 1^st: Argentina, Brazil (NE, 5 and SE) and Uruguay; 2^nd: Bolivia, Brazil (N), Chile, Paraguay and Venezuela; and 3^rd: Colombia, Ecuador and Peru; (vi) carbon electricity intensity to be reduced by 34% by 2058 (IEA, 2014); (vii) subsidies for national fuel prices are eliminated in the long-term, allowing convergence to international prices; (viii) regarding losses in T&D systems, both reduction costs and increasing efficiency of generating technologies are considered; and (ix) existing oil refining capacity and international pipelines limit countries' national supply.

The availability of natural gas for electricity generation was restricted for SAMBA scenarios, so, producing and importing countries cannot use more than 50% of the extracted/imported resource in the power sector. In addition, Argentina and Brazil are the only countries expected to develop shale gas production (due to their large reserves and land availability) and new nuclear plants.

A 34% reduction, by 2058, in the overall electricity's carbon intensity was imposed when structuring the SAMBA scenarios, following results achieved by IEA WEO (2014) for non-developed countries. For Argentina, Bolivia, Brazil, Chile and Peru a maximum installed capacity investment of large-scale electricity production using Concentrated Solar Power (CSP) plants per year is up to 1 GW, while it is 100 MW for Colombia and Venezuela. The same assumptions were applied to investments in large-scale solar photovoltaic plants.

Based on the current energy infrastructure of the region, on some projects suggested by the CIER 15 Project (because many were not implemented), on the national expansion plans of the countries, and on official data and in academic analyzes, four scenarios were proposed: reference integration scenario (RIS), weak integration scenario (WIS), moderate integration scenario (MIS) and strong integration scenario (SIS). It is important to make clear that all scenarios are prepared by the author and are created based on official energy expansion plans of the countries analyzed.

3.1. ASSUMPTIONS

Similarly to selected projects by CIER Project 15, changes in alternative scenarios can be classified as: (i) type I interconnection (operational security and opportunity exchanges); (ii) type II interconnection (operational security and energy export); (iii) use of infrastructure ('swaps'); (iv) hydroelectric with export contracts (economies of scale); and (v) binational plants. Regarding the nature of the alternative policies for each of the alternative scenarios, they can be divided into the following goals: (i) diversification of the power generation mix; (ii) consideration of socio-environmental vulnerability; (iii) increasing in international transactions; and (iv) harmonization of regulatory frameworks.

Reference integrate scenario (RIS) corresponds to business as usual (BAU) scenario, being the baseline scenario. It considers national expansion plans projected by Mercosur governments (short, medium and long-term), in addition to 23 existing international interconnections. As it can be seen, in the reference integration scenario (RIS), there are several policies and energy investments in (and among) Mercosur countries, with particular emphasis on Bolivia. This country does not currently have energy interconnections with its neighbors, despite being in the center of the South American subcontinent.

Weak integration scenario (WiS) is based on the reference integration scenario (RIS). As with RIS, its focus is also national, precisely because it does not include advances of new regional integration projects. As its name suggests, there will be no progress of any project under study. In addition to what is considered in RIS, it considered lower hydro expansion capacity and reduced investment costs of biogas (from second generation) power plants and addition of distributed photovoltaic (PV) in Brazil. Considering the already presented nature of alternative policies, it is perceived that WIS presents diversification of the power generation mix and considers socio-environmental vulnerability, without having any projects that increase international transactions and/or contribute to the harmonization of regulatory frameworks.

The maximum capacity expansion in hydro plants in Brazil was set at a lower level of up to 200 MW per year in the Northern subsystem, 100 MW in the subsystems of the South and Southeast and no hydro expansion in Brazil's Northeast. In addition, distributed PV was considered only in the electricity supply mix of Brazil to assess the impact of the penetration of this technology in 10% of total households, due to recent new regulations. Regarding the third measure, it was assumed that the long-term investment cost of new biogas power plants (US$ 2.449/kW) will converge with the investment cost of bagasse incineration plants in 2013 (US$ 1.905/ kW) in Brazil (^{Moura, 2017}).

Moderate integration scenario (MIS), as well as the weak integration scenario (WIS), is based on the reference integration scenario (RIS). The focus is on the moderate expansion of Mercosur region energy integration projects, considering national hydro projects in Bolivia and international interconnections between the countries analyzed. Considering the already presented nature of the alternative policies, we note that the moderate scenario presents diversification of power generation mix, socio-environmental vulnerability, and has (bi)national projects that increase international transactions, without any connection to the desired harmonization of regional regulatory frameworks.

As its name suggests, strong integration scenario (SiS) is the most audacious scenario. To be viable, there must be a series of changes such as: (i) political will; (ii) diplomatic engineering; (iii) institutional development; (iv) adaptation/harmonization of regulation related to cross-border trade; and (v) advancement of transmission and interconnection infrastructure. Precisely because of this, it was modeled on the moderate integration scenario (MIS), which already considers some of these prerequisites. Regarding this scenario, more than focusing on the expansion of installed capacity, we seek to optimize the use of existing infrastructure in the region. It is not limited to Mercosur countries (and their neighbors, to the extent that there are joint projects), but the analysis is extended to all of South America. Therefore, countries such as Ecuador and Colombia are considered, comprising almost all the States Parties and Associated States of Mercosur, making up almost the whole of South America.

Considering the already presented nature of the alternative policies, we can see that the SIS presents a diversification of the power generation mix, considers socio-environmental vulnerability, has (bi) national projects that increase international transactions and presupposes the desired harmonization of regional regulatory frameworks. In this sense, this is the only scenario that acknowledges all the different natures of the alternative policies considered in the model. In this scenario, there are only projects involving two or more countries. Only one extra dam is considered, although it is facing popular resistance to development (Cachuela Esperanza), given its socio-environmental impacts. At the same time, swaps are considered between Paraguay, Argentina and Chile. Finally, the scenario considers new international interconnections, with Chile, Ecuador and Peru.