Scenario Analysis of an Electric Power System in Colombia Considering the El Niño Phenomenon and the Inclusion of Renewable Energies

Abstract

This paper develops and analyzes four energy scenarios for Colombia that consider the El Niño phenomenon and the inclusion of renewable energies in the energy generation matrix for the period 2020–2035. A comparative analysis is presented between the results of the different scenarios proposed. The most relevant finding is the use of the reserve margin as an indicator of system reliability. A scenario which included 7214 MW of large-scale non-conventional renewable energy, 10,000 MW of distributed generation, and 12,240 MW of hydroelectric power was assumed, with a reserve margin of over 50%. Additionally, it was found that for the scenarios in which a generation capacity with non-conventional renewable energies of less than 10,000 MW in 2034 was assumed, the reserve margin of the system in the seasons of the El Niño phenomenon will be less than historical records of the system. Alternatively, it was found that the scenarios in which the inclusion of at least 9600 MW of the electric power generation capacity of non-conventional renewable energies proposed by 2034 offer benefits in the reduction in greenhouse gas (GHG) emissions, which contributes to the achievement of the emission reduction objectives of the Paris Agreement.

1. Introduction

Electricity in Colombia is mainly produced by hydroelectric plants, which produce between 60% and 70% of the annual generation. The decrease in water resources due to dry seasons has been a vulnerability factor for the generation of electricity in Colombia. This research is focused in the evaluation of the vulnerability of the Colombian electricity generation sector through scenarios taking into account the El Niño phenomenon and the integration of non-conventional renewable energies.

In the present study, the uncertainties that arise in the generation of electrical energy in Colombia are considered, taking into account the changes that the El Niño phenomenon has had in its patterns of duration, periodicity, and intensity due to the effect of climate change [1,2,3], and the integration of non-conventional renewable energies.

The critical factors of the electricity generation sector in Colombia are: (i) the increase in the cost of electricity during periods of drought caused by the El Niño phenomenon [4]; (ii) the increase in the price of fossil fuels [5]; (iii) the limited reserves of natural gas [6]; (iv) the increase in GHG (greenhouse gases) emissions from the generation of electricity due to production by thermoelectric plants [5]; (v) the cost reduction trend of non-conventional renewable energies (solar and wind) [7,8,9]; and (vi) the potential for electricity production from non-conventional renewable energy sources [10,11].

The hydroelectric power generation system of the SIN consists of 28 hydropower plants, 23 water reservoirs, and 39 associated rivers [12]. For a comprehensive study of the El Niño phenomenon’s impact on the national hydroelectric system, the results and the division of the hydroelectric system of 11 aggregate reservoirs will be considered, thus determining the vulnerability and adaptation options of the Colombian energy sector regarding climate change [13]. In this study, the results of [13] were updated to 2020 and found that aggregate reservoirs with high vulnerability to the El Niño phenomenon represent 48% of the national hydroelectric system’s generation capacity.

The vulnerability of the Colombian electric generation has been shown. For example, 1992 and 1993 brought blackouts and electricity rationing. More recently, these periods caused energy spot prices to spike [14]. However, in the current context of the trend towards the reduction in greenhouse gas emissions, the elimination of fossil fuel thermal energy and considering the country’s scarce gas reserves, it is necessary for Colombia to find ways to achieve a reliable electricity supply through the inclusion of solar, wind and biomass energy on both a large and small scale.

Additionally, the authors of [15] analyzed the complementarity of solar and wind resources with water resources in the national territory. The authors of [15] conclude that at the national level, the sites with complementarity between solar and water resources are located in the eastern plains and in the region of Antioquia, while the wind resource is complementary to the water resource only in the eastern plains. At the regional level, the authors of [15] found that wind sites in the northeast show seasonal complementarity with rivers in the north.

The contribution of this study is a joint analysis of two critical uncertainties: the impact of the El Niño phenomenon on electricity generation and the integration of non-conventional renewable energies. Additionally, it was determined that integration of non-conventional renewable energies can minimize the vulnerability of electricity generation in Colombia, with the additional benefit of achieving clean electricity generation.

This paper is organized as follows: Section 2 provides the background, Section 3 describes the materials and methods of the investigation, Section 4 presents the most important results and discussion, and Section 5 concludes.

2. Background

Energy planning of electricity generation systems is a key activity in the context of climate change and sustainability. Energy planning consists of the evaluation of energy alternatives. In this way, each of the alternatives developed represents a prospective scenario. Prospective scenarios represent hypothetical visions of the future constructed from the development of a set of assumptions available in the present. These different images of the future help to understand how the decisions and actions we take today can influence our future [16].

The scenario-based strategic planning methodology was developed in the 1970s. Its main importance lies in the fact that it integrates uncertainty, volatility and complexity into the strategy process. In particular, it analyzes different possible future developments and incorporates aspects of both the industry under examination and the general environment in context. By anticipating change, organizations can be more responsive and alert to changes in the market [17].

Energy planning consists of evaluating energy alternatives. In Colombia, the UPME (Energy Mining Planning Unit) is charged by Law 143 of 1994 with publishing a document yearly with the planning and expansion of the electricity sector in accordance with the ten-year vision of the National Development Plan Project. The document contains demand projections, prospective energy scenarios, and expansion projects for the electric power distribution and transmission network.

The prospective scenarios represent different hypothetical visions of the future that are created through a systematic process in which each of the system evolution alternatives is quantitatively evaluated using energy modeling techniques [16]. In the energy field, these scenarios have been used to characterize the expected future conditions and reveal characteristics of the expansion of the electricity sector. The formulation and construction of scenarios are widely used by the Intergovernmental Panel of Experts on Climate Change (IPCC), the United Nations Framework Convention on Climate Change (UNFCCC), by international organizations such as the Economic Commission for Latin America and the Caribbean (CEPAL), the World Bank, and the Inter-American Development Bank (IDB), among others.

In Colombia, scenarios have been proposed and developed since 1977 from various perspectives. The energy scenarios are used to develop prospects for the development of the country’s energy sector. Among these studies, the book Bases for a National Energy Plan, published in 1977 by the Ministry of Mines and Energy, stands out. In it, different scenarios were developed on the growth of demand to 1990 [18], Futures for sustainable energy in Colombia published by UPME in 2000 [19], and the Colombia National Energy Plan Energy Ideario 2050 published in 2015 by the UPME [20]. In this latest report, different growth scenarios for the energy sector in Colombia were evaluated, considering how changes in economic variables (Gross Domestic Product (GDP), unemployment, inflation, foreign investment, among others) and the effects of the El Niño phenomenon would influence energy demand; also considered were changes in the supply of hydrocarbons, natural gas, and non-conventional renewable energies. In the National Energy Plan’s 2050 ideology, the penetration of different energy sources and the inclusion of energy efficiency measures are considered as critical uncertainties. It concluded that, in the scenario in which electricity replaces traditional energy sources in all uses and sectors where possible, that electricity will have a 90% share of the energy market by 2050.

In the analysis of energy demand and greenhouse gas emissions in Colombia, scenarios were developed using the application of the LEAP model. They maintained the energy demand and GHG emissions of the different sectors of the economy as critical uncertainties of the system. The study concluded that efforts to reduce emissions should be geared toward energy efficiency. Additionally, it was concluded that if the growth in energy demand is high (according to estimates made by the authors based on DANE data) and clean energy is used to generate electricity, GHG emissions will be reduced by 55.4% [21]. In the Energy Scenarios to 2050 study with Integration of Renewable Electric Energy Sources in Colombia, energy alternatives with the integration of renewable sources were considered as a critical uncertainty. The analysis concluded that GHG emissions from the electricity generation sector would be reduced by 25% and that the penetration of non-conventional renewable energies in the National Interconnected System (SIN, acronym in Spanish) would occur as of the year 2029, assuming that no greater capacity will be included in the system through gas or coal-fired thermoelectric plants and that demand will grow 3% per year [22].

Additionally, energy planning studies with scenarios have been developed internationally. Ghanadam and Koomey developed and analyzed four energy scenarios for California, taking into account variations in the integration of renewable energies, and state and national policies of the sector. The study concluded that for California to continue being a leader in energy policy and reduce its dependence on fossil fuels, as well as its GHG emissions, it needed to implement the articulated energy policies that promote energy diversity, energy efficiency, and the reduction in greenhouse gas emissions at the state and federal levels. All this could be achieved through the diversification of the electricity generation matrix, the implementation of measures to promote energy efficiency, and the inclusion of civil society in the use of more efficient appliances and cleaner technologies [23].

Alternatively, in the study Scenarios and Policies for Sustainable Urban Energy Development Based on the LEAP Model, a case study of post-industrial Shenzhen, China, a comparison was made of sustainable urban energy planning scenarios in that region. It concluded that the scenario in which the inclusion of policies to improve energy efficiency and update the energy structure was considered had the lowest economic cost and a significant impact on the subject energy system [24].

In the study of energy scenarios, Expansion Versus Greening? With long-term energy and emission transitions in Mozambique, the authors considered sustainable energy policy alternatives and reduction in GHG as critical uncertainties. The authors concluded that supply-side sustainable energy policies lead to much larger cumulative emission reductions than demand-side policies. Additionally, they showed that supply-side energy policies, including a complete decrease in thermal electricity generation capacity, would lead to shortages in the domestic electricity market, but it would certainly lead to a decrease in the export of electricity. Studies of energy scenarios have also been developed in South Korea to provide work plans for the sustainable development of the energy system. Another study of energy scenarios in South Korea aimed to provide roadmaps for the sustainable development of the energy system. In this work, they developed transition scenarios and a retrospective approach toward renewable energy for both supply and demand [25].

Emodi et al. (2019) studied the effectiveness of policies that have been implemented to curtail GHGs emissions from electricity generation by identifying possible emission reduction policies in power generation and climate change scenarios for the Australian electricity sector and applying approaches from combined backcasting and scenarios. They concluded that future low-carbon pathways lie in clean energy substitutions and innovative energy policies, while global warming presents the need to switch to clean energy technologies early [26].

3. Materials and Methods

The scenarios here presented were developed using the proposed methodology by Schwartz (1991) and adapted for the electricity sector by Ghanadan and Koomey (2005). As part of the proposed methodology, a comparative analysis of the scenarios was conducted.

The methodology proposed by Schwartz in [27] is a five-step methodology. The first step consists of the definition of the scope, then the perception analysis, with the main purpose to specify the key elements that influence the project; in the third step, the perception analysis is carried out, in order to obtain holistic maps of possible visions of the future; in the fourth step, the analysis of trends and uncertainties is carried out, and in this step the possible interrelationships between the critical uncertainties and their possible impact on the context of analysis are analyzed; and in the fifth step, scenarios are constructed around the critical uncertainties that are considered to be of high importance and uncertainty.

The methodology proposed by Ghanadan and Koomey (2005) is an adaptation to the electricity sector of the methodology proposed by Schwartz (1991)—they add a step which consists of evaluating the implications of the scenarios. In the methodology proposed for the development of this paper, the methodology proposed by Ghanadan and Koomey (2005) was modified by adding a step of comparative analysis of the scenarios developed. The methodology proposed for the development of this paper is shown in

Table 1. Proposed Methodology for the Development of this Paper. Source: adapted from [27].

(1) Identify and define the sources of uncertainty.

(2) Set a list of the most relevant criteria in the context of the analysis.

(3) Evaluate the criteria by importance and uncertainty.

(4) Select scenarios of the El Niño phenomenon and renewable energy integration.

(5) Develop the scenarios around the critical uncertainties.

(6) Analyze developed scenarios.

(7) Analyze the developed scenarios comparatively.

.

3.1. Step 1: Identify and Define the Sources of Uncertainty

In this step, the sources of uncertainty related to the topic of interest were defined based on the challenges for the Colombian energy sector in the coming decades. The energy scenarios were built around the occurrence of the El Niño phenomenon and the inclusion of renewable energies in the SIN.

Table 2. Research questions.

How is the behavior of the demand for electric energy in Colombia?

How is the current and historical behavior of the El Niño phenomenon?

How does El Niño phenomenon affect electricity generation in Colombia?

How should the energy planning of a primarily hydraulic system be done?

How has energy planning been developing in Colombia?

What are other countries’ experiences with the inclusion of non-conventional renewable energies?

What proposals and projects does Colombia have for the inclusion of renewable energies in the SIN?

How could COVID-19 affect the growth of electricity demand?

Source: authors’ elaboration.

shows some of the questions developed in this study.

3.2. Step 2: Set a List of the Most Relevant Criteria in the Context of the Analysis

The importance of this step lies in the identification of the interactions that may occur in the electric power generation system in Colombia.

Table 3. Example of important criteria in power electric generation system in Colombia.

The vulnerability of the Colombian power generation sector to the occurrence of the El Niño phenomenon is increasingly critical given that it is more intense and of longer duration.

The integration of non-conventional renewable energies.

The increase in the price of fossil fuels.

Energy storage.

Changes in governmental policies.

Source: authors’ elaboration.

shows the most important criteria.

3.3. Step 3: Evaluate Criteria by Importance and Uncertainty

The fluctuating nature of water resource availability for electricity generation means that additional resources are required to meet electricity demand. The adoption of new resources in the electricity system in Colombia depends on the interaction of several factors, such as governmental policies and technical and socio-economic conditions. The purpose of this step is to evaluate the relevance and uncertainty of the criteria identified in Step 2.

The critical uncertainties of the electric power generation system in Colombia are those with the greatest importance and uncertainty. For the scenarios developed in this paper, the criteria were set from the beginning given the interest in jointly evaluating the effect of the occurrence of the El Niño phenomenon and the gradual integration of renewables.

3.4. Step 4: Select Scenarios of the El Niño Phenomenon and Renewable Energy Integration

The purpose of this step is to develop the relationships according to the evaluation of Step 3. The critical uncertainties are the following: the growth of electricity demand considering the impact of COVID-19 on economic growth, the inclusion of non-conventional renewable energy sources, and the El Niño events.

3.5. Step 5: Develop the Scenarios around Critical Uncertainties

In this step, a story was developed for each of the scenarios, reflecting how critical uncertainties are interconnected to create a possible future. Thus, each scenario shows one of the different views of the possible potential for the occurrence of the El Niño phenomenon, and the possible integration percentage of the renewable energies of the electric power generation system in Colombia. Therefore, each of the scenarios here presented is intended to broaden the debate on alternative future paths for the electricity supply system in Colombia.

3.6. Step 6: Analyze Developed Scenarios

Each of the scenarios is analyzed based on the demand and supply of the different types of resources used to produce electricity to obtain the total prospective system capacity, the prospective energy generation, the prospective reserve margin, and the prospective CO2 emissions for the scenarios.

3.7. Step 7: Analyze Comparatively the Developed Scenarios

The purpose of this step is to compare the different evolution views of the electric power generation system in terms of the reliability of its seasonal prospective response to the occurrence of the El Niño phenomenon and the reduction in GHG emissions. In other words, the results obtained in step 6 in each of the scenarios are compared with each other, and subsequently, a conclusion is drawn as to how each scenario behaves in terms of reduced vulnerability to the El Niño and GHG emissions.

4. Results and Discussion

This section presents the development of steps 5, 6, and 7 of the proposed methodology. Initially, the national electricity forecast demand is carried out based on the impact of the COVID-19 public health situation. This demand curve is used for all the prospective scenarios. Subsequently, each of the scenarios is based on the assumptions, and finally, the main results are shown and analyzed.

4.1. Electricity Demand

Figure 1 shows the prospective electricity demand curve by the UPME and the prospective electricity demand curve in this study, which considers the expected impact of the pandemic on the growth of electricity demand [28]. Two percent was considered as the energy demand growth for the period 2020–2023 (Schwenke et al. 2013) and three percent for the period 2024–2035. UPME forecasted an electricity demand growth of more than 2.1% in all periods. In 2034, according to the UPME, the electric energy demand will be 111,399 GWh, and in this study, the electric energy forecast demand in 2034 will be 105,655.2 GWh, i.e., 5.1% lower than the electric energy forecast demand by the UPME. This is due to several factors: (i) this study’s forecast is based on the electricity demand for 2020, which is 3% lower than the forecast made by UPME; and (ii) the prospective demand growth is 2% for the period 2020–2023 and the average growth of the prospective electricity demand by UPME for that period is an average of 3.1%.

4.2. Base Scenario Analysis

The base scenario depends on the Colombian situation in 2020 and the advances in energy policy for the diversification of the electric power generation matrix, which is primarily hydropower, to the extent that in 2020 hydroelectric generation capacity represents nearly 64%. A 16% inclusion of non-conventional renewable energy sources is considered in the electric power generation matrix in Colombia for the 2020–2035 period. The base scenario shows no occurrence of the El Niño phenomenon for a point of comparison with the other scenarios.

The regulation for the integration of non-conventional renewable energy sources into the National Electric System started with Law 1715 in 2014. Subsequently, the introduction of different measures with the purpose of energy basket diversification seeks to reduce Colombia’s electric power generation system’s dependence on water resources. Then, with Law 2099 of 2021, the dynamization of the energy market is sought through the use, development, and promotion of non-conventional sources of energy, and in general to regulate the strengthening of public services of electric energy and gasoline fuel [30]. The impact of the El Niño phenomenon on hydroelectric power generation has led to a reduction in supply, the use of energy from gas and coal-fired thermoelectric plants has been used. These plants have provided the reliability required by the system and consequently, increased the costs of electric power and environmental consequences in the extraction and conversion of fossil resources.

It has been observed that the patterns of duration, intensity, and periodicity of the El Niño phenomenon have changed as a result of climate change, therefore, its behavior and the generation of hydroelectric power are uncertain because the El Niño phenomenon reduces the catchment flows in hydroelectric reservoirs. Therefore, as it has been documented in different reports, the production of electrical energy is vulnerable to the El Niño phenomenon [13,31,32].

The hydroelectric power generation system of the SIN consists of 28 hydropower plants, 23 water reservoirs, and 39 associated rivers [33]. For a comprehensive study of the El Niño phenomenon’s impact on the national hydroelectric system, the results and the division of the hydroelectric system of 11 aggregate reservoirs will be considered, thus determining the vulnerability and adaptation options of the Colombian energy sector regarding climate change [13]. These study results were updated to 2020. See

Table 4. Effective generation capacity of aggregate reservoirs.

Aggregate Reservoir	Effective Capacity (MW)	Effective Capacity (%)	Vulnerability to Climate Change
Antioquia 1	2221	20.23%	High
Antioquia 2	2001	18.23%	High
Bogota	785	7.15%	Low
Caldas	396	3.61%	Medium
Cauca	480	4.37%	High
Caribe	338	3.08%	Low
Huila	940	8.56%	Medium
Oriente 1	819	7.46%	Low
Oriente 2	2250	20.50%	Medium
Pacifico	561	5.11%	High
Tolima	187	1.70%	Low

Source: authors’ elaboration based on XM and ACOM-OPTIM information (2013).

.

Aggregate reservoirs of high vulnerability to climate change were modeled in the LEAP software with 50% reductions in capacity during the El Niño events. These reservoirs account for 47.94% of the national hydroelectric system’s generation capacity.

Additionally, it has been identified that the country has potential for energy generation with other renewable energy sources such as wind energy and solar energy. Thus, to diversify the electricity generation matrix in Colombia, the generation with renewable energies has been promoted for this purpose, the first long-term auction was held in 2019 [34]. In the auction that took place in October 2019, 10,186 MWh-day were assigned, which corresponds to 85% of the target demand for energy allocation determined by the Ministry of Mines and Energy. The weighted average price in the auction was USD 95.65/kWh. They were assigned 544 contracts to 7 generators, in 8 projects with a total effective capacity of 1298.9 MW, which are distributed as follows: 17.39% corresponds to solar photovoltaic and 82.61% to wind.

In the baseline scenario, for 2035, a capacity inclusion was assumed to have a matrix of electrical energy capacity composed as follows: non-conventional renewable energies—27.5%, solar energy—11.4%, wind—11.4%, PCH—4%, and co-generators—0.7%, and a share of hydroelectric plants of 46.5%. Currently, the level of participation of renewable solar photovoltaic and wind energy is less than 1%, in this scenario, it is considered that a trend increase in renewable energies will occur. Therefore, the increase to 15% of solar and wind energy capacity aims to reflect some growing trend of integration, such as the one presented by the UPME in scenario 1 of the Generation and Transmission Expansion Plan 2020–2034 [29]. Scenario 1 of the Generation and Transmission Plan proposes that the increase in solar and wind energy capacity of 15% will be achieved by 2023 and that in this scenario this participation will occur in 2028 because the scenario of this research implies a more conservative inclusion of solar and wind renewable energies.

The capacity of the electricity generation system in Colombia in 2020 is 17,480 MW (see Figure 2 Generation Capacity (2020–2035)). With the increases in the inclusion of non-conventional renewable energies proposed in the expansion plans and other official documents [20,34], it is observed that by 2035 the capacity of the SIN in Colombia would be approximately 26,300 MW. This means that in 14 years the increase in capacity compared to the current value would be approximately 50%.

It is observed that in 2025 the generation capacity of the system prospected in the base scenario will be 21,394 MW, and in 2030 the generation capacity will be 24,112 MW. In 2035, there will be a more diverse matrix of power generation capacity since it would have several generation sources, 27.4% of non-conventional renewable energies, and 46.5% of hydroelectric generation capacity, that is, approximately 74% of the capacity of the system would have intermittent resources and no shipping control. Figure 3 shows the curve of the electric power generation capacity of the system prospected in the base scenario between 2020 and 2035.

The energy produced by the electricity generation system in Colombia in 2020 was 80,800 GWh (see Figure 3. Generated Energy). By taking into account the increases in the inclusion of non-conventional renewable energies assumed by the authors according to current trends, expansion plans, and expansion proposals of the system, it is observed that by 2035 the energy of the SIN in Colombia would be approximately 122,300 GWh. It means that in 15 years the increase in the energy generated about the current value would be 51%.

The energy generated in the years that in scenarios 1, 2, and 3 the El Niño phenomenon prospected will be taken into account. In 2025, hydroelectric generation will be 58,098 GWh, which represents 64%. In 2030, the energy generated by the system will be 104,241.7 GWh, and hydroelectric plants will be responsible for generating 66,270 GWh.

4.3. Analysis of Scenario 1

Scenario 1 considers the situation in Colombia in 2020 and what would be the situation of the electric power generation system in 2035 taking into account: (i) the reduction in generation capacity of the national hydroelectric system during the El Niño phenomenon; and (ii) the inclusion of electric power generation capacity from non-conventional renewable energy sources. In the national electricity demand, the following are considered: (i) a reduction in the growth of the demand for electricity to 2% from 2020 to 2023 as a result of the COVID-19 public health situation; and (ii) from 2024 to 2035 the growth of the demand for electrical energy tends to the historical trends with a growth of 3%.

In this scenario, an increase in the installed capacity of solar and wind energy to 15% by 2028 will be assumed, it aims to reflect a certain level of trend integration, such as that presented by UPME in scenario 1 of the Generation and Transmission Expansion Plan 2020–2034 [29]. This growth corresponds to the diversification policies mentioned and the macro-objectives such as the United Nations Sustainable Development Goal 7 related to affordable and clean energy.

Thermal generation plants provide reliability to the electric power generation system because they partially support the system in El Niño seasons, although they have higher production prices [35]. It should be noted that the country has limited reserves of natural gas (9.2 years in 2018) [36]), and that traditional coal-fired thermoelectric plants release a large amount of GHG into the atmosphere. Alternatively, the IEA presents gas-fired thermoelectric energy as among the pillars of global energy that replace more polluting fuels, improve air quality, and limit carbon dioxide emissions [37]. As Colombia does not have enough gas reserves, it has been preparing since 2018 through the installation of regasification plants in the Atlantic and Pacific Oceans to incorporate thermoelectric plants.

The assumptions that were taken into account in scenario 1 are (i) an El Niño event in 2025 and another event in 2030. These events will reduce the generation capacity of the aggregate reservoirs Antioquia 1, Antioquia 2, Cauca, and Pacífico to 35%; (ii) the current trends of inclusion of non-conventional renewable energies, SHP—4%; solar—11.4%; wind—11.4%; cogenerators—0.7%) if considering the expected growth by conducting long-term auctions every five years [33]; (iii) the first phase of Hidroituango and hydroelectric projects that will start operating in 2022 (46.5%); (iv) current trends in thermoelectric generation (gas-fired thermoelectric—11.4%; coal-fired thermoelectric—13.8%; NDC—0.8%).

In Figure 3, it is observed that the capacity of the electricity generation system in Colombia in 2020 was 17,480 MW. With the increases in the inclusion of non-conventional renewable energies proposed in the expansion plans and other official documents [20,34], it is observed that by 2035 the capacity of the SIN in Colombia will be approximately 26,300 MW. This means that in 14 years the increase in capacity compared to the current value would be approximately 50%.

In 2025, due to the El Niño phenomenon, a 35% reduction was modeled in the aggregate reservoirs Antioquia 1, Antioquia 2, Cauca, and Pacifico. This reduction in the generation capacity of the hydroelectric system is 4216 MW. If the generation capacity of the system in 2025 in scenario 1 is compared with the generation capacity of the base scenario in the same year, it is found that the generation capacity of scenario 1 is 10.5 pp lower than that of the base scenario; the energy generated in scenario 1 presents a reduction of 12,342 GWh about the energy generated in the base scenario, as shown in Figure 4.

In 2030, the generation capacity of the hydroelectric system will be 12,240 MW in the base scenario, while in scenario 1 due to the El Niño phenomenon in 2030 a reduction is expected in the generation capacity in the aggregate reservoirs Antioquia 1 of 1498 MW, Antioquia 2 of 2033 MW, Cauca of 316 MW, and Pacific of 369 MW. In general, in 2030, large-scale hydroelectric power generation capacity is expected to reduce by 4216 MW as a result of the El Niño phenomenon. It is important to take into account that the reduction in hydroelectric generation was traditionally complemented by thermoelectrically generation. As it happened in the 2015 El Niño event. The energy generated is equivalent to a decrease of 16,341 GWh from the energy generated in the base scenario, as shown in Figure 4.

The reduction in the generation capacity of the system affects its reliability. Among the reliability measures used in electric power generation systems is the reserve margin. The reserve margin is the percentage of installed electricity generation capacity that is not being used at a given time and that, therefore, can be used under certain conditions if necessary. This backup capacity is technical and regulatory in nature (Energy and Gas Regulatory Commission (CREG) Resolution No 074 of 2006). The reserve margin is used to supply the system’s capacity at times of power plant failures, maintenance periods, and fuel shortages, among other eventualities that may cause the partial or total shutdown of an electric power generating plant.

The two periods in which the occurrence of the El Niño phenomenon was forecast, 2025 and 2030 (see Figure 5 and

Table 5. Historical reserve margin.

	1998	2010	2013	2016	2019
Reserve margin	38.1%	38%	55%	65.8%	68.3%

Source: [38] (1998 and 2010); 2013 and 2016 [39]; 2019 [40].

), will be compared with the historical data for 2010, the year in which the phenomenon in question occurred. A comparison of the 2025 reserve margin with that of 2010 shows a reduction of 2.8 pp. A comparison of the 2010 reserve margin with that of 2030 shows a decrease of 2.9 pp. Therefore, in 2025 and 2030, the system would be at risk of shortages.

After the analyses performed previously, the GHG reduction in scenario 1 was examined, with the main objective of reviewing whether the country will achieve compliance with the Paris Agreement, for which the electric power generation subsector committed to producing 4.74 million tons of CO2 equivalent in 2030 [41].

Following the analyses carried out previously, an analysis of GHG reduction in scenario 1 was performed; the main objective is to examine whether the country would achieve compliance with the Paris Agreement through this scenario. Colombia committed to reduce GHG emissions by 20% by 2030 compared to 2010 and to increase its GHG emissions reduction to 30% if it receives international support [42]. Specifically, the Ministry of Mines and Energy refers that in 2010 emissions from the mining and energy sector corresponded to approximately 29.4 million tons of CO2 equivalent [43]. The GHG reduction goal for the electric power generation subsector is 4.74 million tons of CO2 equivalent by 2030 [41]. Figure 5 presents the CO2 abatement curve, which shows the reduction in tons of CO2 equivalent emitted to the atmosphere in the period 2020–2034.

It is found that in scenario 1 (see Figure 5), emissions reductions to 2030 are insufficient to reach the government’s goal; the projected emissions reduction for this scenario will be 2.1 million tons of CO2 eq. Therefore, it can be concluded that if the long-term auctions are held every five years and if the expected trends in the sector continue, the emissions reduction goal for the electricity generation subsector proposed by the national government for the achievement of the commitment acquired in the Paris Agreement will not be achieved.

4.4. Scenario 2 Analysis

This scenario represents a vision of the Colombian power generation system assuming that in the forecast period 2020–2035 there will be an El Niño event in 2025 and another in 2030, which will reduce the power generation capacity by 35% in the aggregate reservoirs Antioquia 1, Antioquia 2, Cauca, and Pacífico. Additionally, the inclusion of large-scale solar and wind energy at 25%, the complete dismantling of coal-fired power plants, and an increase in the generation capacity of gas-fired power plants to 16% by 2034 are assumed.

The assumption of the inclusion of non-conventional renewable energies is made following the same trend presented in scenario 3 proposed in the UPME’s Expansion and Transmission Plan 2020–2034 [29], the diversification policies previously mentioned, and by taking into account the United Nations Sustainable Development Goal 7 related to affordable and clean energy. In addition, it is supported by the review of how renewable energies have been included in power generation systems in other countries. On the other hand, the assumption related to the increase in generation capacity of gas-fired thermoelectric power plants coincides with the energy policy of the Ministry of Mines and Energy that considers natural gas as the energy resource for energy transformation in Colombia [44], and with the suggestion of the IEA that considers gas as a resource for energy transition gas [37].

This scenario will consider a power generation capacity matrix made up of 36.7% with different non-conventional renewable energy sources (4% SHP, 16% wind energy, 16% solar photovoltaic energy, and 0.7% co-generators) and 46.5% hydroelectric power plants. Moreover, in 2034, the total dismantling or technological reconversion of all coal-fired thermal power plants is considered, as well as a 16% increase in the share of gas-fired thermal power plants in the electricity generation capacity matrix.

The reliability of the electric power supply is related to the El Niño phenomenon. In this context, this scenario intends to show at a prospective level the occurrence of two El Niño events in the next 15 years. The first El Niño event will occur in 2025, reducing the generation capacity of the SIN reservoirs to approximately 35%. The capacity is reduced since the volume of water captured decreases and therefore the effective capacity. It is observed that by 2035 the generation capacity of the system would be approximately 26,000 MW.

In 2025, even though an increase in non-conventional renewable energies of 857 MW was forecast to be higher than in the base scenario, and 429 MW more generation capacity was included in gas-fired thermal power plants than in the base scenario, the system’s generation capacity will not be enough to make up for the reduction in generation capacity of the hydroelectric system of 4216 MW, and the decrease in generation capacity of the coal-fired thermal power plants of 1296 MW forecast. Thus, this year’s power generation capacity will be reduced by 4232 MW compared to the baseline scenario generation capacity. In consequence, it is concluded that this year new plants would be required to enter the system to supply the demand satisfactorily.

In 2030, the occurrence of a second El Niño event was forecast, which would result in a decrease in the generation capacity of the hydroelectric system of 4216 MW compared to the base scenario. On the other hand, it is found that despite the increases in generation capacity of non-conventional renewable energy of 1714 MW and gas-fired thermoelectric plants of 857 MW, the reduction in generation capacity of coal-fired thermoelectric plants of 2593 MW and hydroelectric plants of 4216 MW will not be compensated, so that the system will present a deficit in generation capacity. To meet this year’s electricity demand, it will be necessary to add generation capacity to the system.

The energy generated is shown in Figure 4; in 2025, the energy generated by hydroelectric power plants will be complemented by non-conventional renewable energy generation and thermoelectric energy. Non-conventional renewable energies will generate 2925 GWh more than in the baseline scenario. This increase in non-conventional renewable energy generation will offset the reduction in coal-fired thermoelectric power in this scenario compared to the power generated in the baseline scenario. Therefore, the decrease of 12,342 GWh in hydroelectric power generated will not be supplied by additional generation; thus, the system will need additional energy, which can be supplied by new plants or by purchasing energy in the stock market. In 2030, the year in which a second El Niño event is expected to occur, the hydroelectric energy generated will be reduced by 16,340 GWh due to this event, and similar to what happens in 2025, the system will have to buy energy in the stock market or build new plants. It is important to take into account that the gap between the electricity generation and the electricity demand was supplied traditionally with thermoelectrically generation.

Previously, the effect of the two El Niño events that were simulated on the national power generation system was analyzed, and now the impact on the SIN reserve margin will be studied. Figure 5 shows the SIN reserve margin in the period from 2020 to 2035. In 2020, the system’s reserve margin was 58.2%, which allowed the national electricity system to provide a timely response in the seasons in which the hydroelectric system would present a deficit. In the year 2025, there is a decrease in the reserve margin due to the occurrence of the El Niño phenomenon, which is greater than that presented by the SIN in the periods 1998 and 2010. In 2025, the reserve margin will be 35.1%, therefore, it is considered necessary to increase the reserve margin in this period so that the system is not at high risk of shortage.

The behavior of the 2030 reserve margin is the same as that of 2025 since the forecast is similar to that presented by the system during the El Niño events of 1998 and 2010. Thus, the reserve margin in 2030 will be 34.9%, which will be 3.1 pp lower than the reserve margin in 2010.

Following the above analysis, the GHG reduction in scenario 2 was analyzed, with the main objective of examining whether the country will achieve compliance with the Paris Agreement through this scenario. Figure 5 shows the CO2 abatement curve for the period 2020–2035. Finally, it can be concluded that if there is an estimated reduction in emissions by 2030 of 9.7 million tons of CO2 eq. in the period 2020–2030 in the electric power generation subsector, and given that the national reduction goal for the electric power generation subsector is 4.74 million tons of CO2 eq. [41], it would meet the goal proposed by the national government and could contribute to the total emissions reduction in the mining and energy sector or other sectors of the economy.

4.5. Scenario 3 Analysis and Comparative Analysis

Scenario 3 shows how by including large- and small-scale non-conventional renewable energies, the country can mitigate the effect of vulnerability associated with the impacts caused by the El Niño phenomenon in the SIN. Regarding small-scale non-conventional renewable energies, the inclusion of distributed generation (DG), bioenergy, and photovoltaic self-generation are proposed, in such a way that the reduction in the generation capacity of the systems is obtained using the firm energy that biomass has, and the distribution of the location of the different energy generating units. Alternatively, the installation of non-conventional renewable energy, solar, and wind generation capacity on a large scale is also assumed.

Thus, this scenario proposes a transformation to 2035 in search of clean energy generation in the system through the inclusion of non-conventional renewable energies in 58% of the generation capacity in the system. This generation capacity will be complemented with 41% hydroelectric energy and 1% from non-centrally dispatched thermoelectric plants so that the proposed energy generation capacity matrix is 99.3% clean. This scenario shows the implications of the occurrence of two El Niño events in the period from 2020 to 2035 and the inclusion of solar and wind energy generation capacity of 36.7% by 2034. This scenario is inspired by scenario 5 of the Generation and Transmission Expansion Plan 2020–2034, which considers the inclusion of solar and wind power generation capacity greater than 30% by 2034, and the occurrence of two strong El Niño events follows the inclusion tenure proposed by UPME [38] (See Figure 3, Generation capacity).

In 2025, due to the occurrence of El Niño, the effective generation capacity of the hydroelectric power plants will be reduced; a decrease that will represent, in the total effective capacity of the system, 9.2% compared to the base scenario, which means that the system will require an additional capacity of 1967 MW to meet the demand in this period. Similarly, by 2030, the decrease in generation capacity is 9% and the additional capacity required is 1917 MW.

In 2025, the effective generation capacity will be reduced in the three scenarios (1, 2, and 3) as a consequence of the occurrence of the El Niño phenomenon this year. In scenario 1 the effective capacity will be reduced by 4232 MW, in scenario 2 by 4216 MW, and in scenario 3 by 1967 MW compared to the base scenario. Therefore, in scenario 3 there is a smaller reduction in effective capacity in 2025.

In 2030, the effective generation capacity is reduced by 1917 MW compared to the baseline scenario. This reduction in effective generation capacity is less than the reduction of 4216 MW in scenario 1 and 4248 MW in scenario 2. Thus, while in scenarios 1 and 2 the effective generation capacity is expected to be reduced between 19.7 and 19.9% as a consequence of the El Niño event, in scenario 3 it will be reduced by 9%. In scenario 3, the effective generation capacity will be 10.9 pp higher than in scenarios 1 and 2.

The reduction in the system’s generation capacity is also observed as a reduction in the energy generated in these same years, such that, in 2025, the energy generated shows a reduction of 0.95% in scenario 3, which is equivalent to 862.6 GWh less than in the base scenario. This decrease is due to the reduction in the generation of the national hydroelectric system and therefore must be supplied by other generation plants. In addition, it is found that the decrease in energy generated in scenario 3 is lower than the decrease in scenarios 1 and 2 compared to the base scenario. Thus, while 862.6 GWh less is expected to be generated in 2025, in scenarios 1 and 2, the reduction will be 12,340 GWh. This indicates that in scenario 3, 11,477.4 GWh less will be required to meet demand than in scenarios 1 and 2. It is important to take into account that the gap between the electricity generation and the electricity demand was supplied traditionally with thermoelectrically generation.

In the year 2030, it is expected that due to the effect of the El Niño phenomenon, the reduction in energy generated in scenario 3 will be 5400 GWh compared to the base scenario and is required to be supplied by new generation plants. Furthermore, the energy reduction generated is lower than in scenarios 1 and 2 compared to the base scenario, because while scenarios 1 and 2 show a reduction of 16,340 GWh, scenario 3 will require 10,940 GWh less to meet demand than scenarios 1 and 2. Therefore, in scenario 3 the system would need to supply less energy with other generation plants and would be more economically viable.

It was found that in this scenario distributed generation (DG) with solar and biomass in 2025 will generate 21,400 GWh and in 2030 30,500 GWh will help to make up the shortfall of energy generated by the hydropower system and thus contribute to the security of energy supply.

The reduction in generation capacity also has implications for the reserve margin. This compares the projected reserve margin in scenario 3 with the reserve margin in 1998 and 2010. The reserve margin is higher in 2025 by 14.9 pp and in 2030 by 12.7 pp than the reserve margin that the system had in the critical events of 1998 and 2010. This scenario is the only one of the scenarios that show a reserve margin higher than the historical reserve margin for the periods 1998 and 2010. In scenarios 1 and 2, the system reserve margin is lower than the historical margin in 1998 and 2010.

Following the analysis of how the capacity matrix, the energy generated, and the reserve margin will be transformed, the reduction in CO2 emissions from the electricity generation sub-sector as a result of the transformation of the electricity generation fleet in the period 2020–2035 will be analyzed. This analysis will take into account that Colombia’s 2030 reduction target for this period is 4.74 million tons of CO2 equivalent to achieve compliance with the commitment made in the Paris Agreement. Figure 5 shows the abatement curve for the base scenario, scenario 1, scenario 2, and scenario 3.

Figure 5 shows that the scenario with the greatest reduction in emissions is scenario 3. In the period 2020–2030 in scenario 3, the emissions of the electric power generation subsector will be reduced by 13.5 million metric tons of CO2 eq., a value that exceeds in 2030 the expected reduction in emissions in the electric power generation subsector in the country by 8.75 million tons of CO2 equivalent, which could contribute to the reduction in emissions in other sectors.

5. Conclusions

Colombia has a predominantly hydroelectric power generation system, where reliability is affected by the El Niño phenomenon. To find a solution to this problem, the country has introduced some market mechanisms such as the capacity charge and the reliability charge that have provided the generation system with the capacity required to meet demand. However, these mechanisms have not provided effective solutions when the El Niño phenomenon occurs, since the system sometimes has a low reserve margin in these seasons, which indicates a high possibility of energy shortages. On the other hand, Colombia is committed to reducing GHG emissions to comply with the commitment made in the Paris Agreement; however, during El Niño events, GHG emissions in power generation could increase because hydroelectricity has traditionally been complemented with thermoelectricity. UPME currently promotes the inclusion of non-conventional renewable energies; however, it is not clear to what extent this type of energy can contribute to improve the reliability of the electric power generation system.

The reserve margin with the occurrence of the phenomenon of El Niño is approximately 35%. The historical reserve margin reported in Colombia in 1998 and in 2010 is 38%, which indicates that the participation of hydroelectric generation at a large scale in Colombia represents a vulnerability factor that could even increase in the coming years according to the scenarios developed in this paper. Meanwhile, the reserve margin in scenario 3 shows that the incorporation of distributed generation of renewable energy represents a key factor to mitigate vulnerability in energy supply electricity reliably.

On the other hand, scenarios 2 and 3 will contribute to the reduction goal proposed by the national government for the achievement of the Paris Agreement. In scenario 2, the minimization of emissions would be 9.7 million tons of CO2 eq for 2030, considering the definitive closure of coal-fired power plants and integration of non-conventional renewable energies to 37% by 2035. Scenario 3 would contribute with a emission reduction by 2030 of 13.5 million tons CO2 eq by closing the thermoelectric plants and using distributed generation.