Abstract
Background and potential impacts
The proportion of the population without access to electricity worldwide has been decreasing over the last decades, reaching about 9% of the total world population, but maintaining 17% for the rural population in 2021, especially in Sub-Saharan Africa, Southeast Asia, and Latin America [1], [2]. It represents progress towards meeting SDG-7 [3], [4] however, as progress is made towards this target, difficulties increase because much of this population is located in remote and off-grid localities, where plants with small capacities and operating at low load factors are required, which increases energy costs [5]. In addition, it is a challenge for governments in different countries, such as Colombia, where, by the year 2020, 6% of the population, most of them in the NIZ, will not have access to electricity (14% of this population), a value similar to the average of other developing countries [6].
Identifying the energy potentials of renewable energy sources such as residual biomass represents an opportunity to project sustainable energy solutions for electrification in these localities in Colombia. Potentially, this energy resource can represent between 14% and 29% of the country’s electricity demand, according to the 2019 national demand of 259 PJ [7]–[9] Therefore, it could be an environmentally viable solution for reducing greenhouse gas (GHG) emissions in the country, contributing to the fulfillment of the country’s international commitments in the fight against climate change to reduce these emissions by 51% by 2030[10].
Methodology
Energy service data analysis
Initially, nine technical and commercial variables on electricity service in the ZNI localities were obtained from the Unique Information System (SUI) of the Superintendence of household public services (SSPD) [11], reported by the service operators. Data were cleaned, leaving only stratum one residential users (about 99% of users) and only localities with less than 500 kW of installed capacity. Localities with implemented or planned photovoltaic solutions were also discarded. It results in 1553, 1618, 1608, and 1589 localities, respectively, for four moments: Mar-2019, Sep-2019, Mar-2020, and Sep-2020. Missing data was filled in different forms, as recommended in [12]. COE missing data were filled according to values in localities with the same service operator, and billing was filled using energy data and COE. For diesel unit and transport costs, missing data were filled according to localities with the same fuel collection site and municipality. For some localities with more than one Genset, only one was taken according to with maximum power reported for these localities in CNM telemetry reports [13]. The analysis of the four moments results in 1489 localities with data at each moment, showing an invariable behavior of the data according to the P-value test, which allows selecting only one data set, being the selected one the one corresponding to Sep-2020.
The nine variables were normalized by the Z-score method to avoid the dominance of those with higher ranges or scales. Subsequently, they have been dimensionally reduced by Principal Component Analysis (PCA), and some outliers have been removed by the percentile method. The first three principal components add up to a variance higher than 90% and are the ones selected to analyze the data. Finally, using the k-medoids method and the Partitional Around Medoids (PAM) algorithm, using a Square Euclidean distance, clustering the data has been made according to the values obtained for the 3 PCs representing the highest variance.
Analysis of waste biomass data and its energy potential
The residual biomass data were obtained from the municipal agricultural assessments (EVA) projected for 2020. [14] for all the municipalities that contain the localities of the sep-2020 dataset, for a total of 1522 localities with energy service data and residual biomass data considered for energy use [7]. The EVAs' relevant data for estimating residual biomass's energy potential are the cultivated area, the yield per hectare, and the annual tons of production. These data, together with the product/residue ratio taken from [7] and a residual biomass harvesting factor of 0.8 [15], [16], allow estimating the theoretical energy potential of these residual biomasses, similar to other published works. [17], [18]. The lower calorific value of each waste is established according to different references and estimates [7], [19]–[24]. Finally, the technical energy potential for biomass gasification is estimated using a conversion efficiency of 13%, the minimum value referenced for downdraft biomass gasifier systems coupled to internal combustion engines [25], [26], the most suitable for small-scale power generation from biomass [27]–[29].
The residual biomass energy potential data are then obtained for each municipality with localities in ZNI and normalized similarly to the energy service data. Allocation of energy potential for each locality is made by weighting the number of stratum one users of the service in each locality of each municipality, assuming that the area of their agricultural activities is proportional to the number of users and people in each one.
Results and conclusions
The analysis of energy service data in the localities of the ZNI has made it possible to determine clusters that group these localities into five typologies according to their technical and commercial energy service data. Furthermore, the characteristics of each cluster have made it possible to establish two deficit indicators, one for users served and the other for diesel fuel deficit. Finally, a weighted aggregation was used to generate an energy service deficit indicator for each locality according to its typology or cluster.
The analysis of residual biomass energy potential data has made it possible to establish a possible energy margin indicator between the potential for each location and the theoretical electrical energy to be supplied by the operators in each location.
These two indicators have been weighted-aggregated, and using adjustment factors; it has been possible to establish an indicator of the relevance of energizing localities in ZNI from biomass gasification based on public data and analysis of these data. This indicator allows identifying those localities where this possible energy solution deserves to be analyzed in terms of its feasibility and sustainability, achieving greater precision when considering possible projects in these localities. Likewise, this methodology can be replicated for other renewable energy sources such as photovoltaic or thermal solar energy or livestock waste. Finally, the results show how the data analysis is a helpful tool for studying energy solutions aligned with the fulfillment of national goals in the fight against climate change in isolated localities.
The proportion of the population without access to electricity worldwide has been decreasing over the last decades, reaching about 9% of the total world population, but maintaining 17% for the rural population in 2021, especially in Sub-Saharan Africa, Southeast Asia, and Latin America [1], [2]. It represents progress towards meeting SDG-7 [3], [4] however, as progress is made towards this target, difficulties increase because much of this population is located in remote and off-grid localities, where plants with small capacities and operating at low load factors are required, which increases energy costs [5]. In addition, it is a challenge for governments in different countries, such as Colombia, where, by the year 2020, 6% of the population, most of them in the NIZ, will not have access to electricity (14% of this population), a value similar to the average of other developing countries [6].
Identifying the energy potentials of renewable energy sources such as residual biomass represents an opportunity to project sustainable energy solutions for electrification in these localities in Colombia. Potentially, this energy resource can represent between 14% and 29% of the country’s electricity demand, according to the 2019 national demand of 259 PJ [7]–[9] Therefore, it could be an environmentally viable solution for reducing greenhouse gas (GHG) emissions in the country, contributing to the fulfillment of the country’s international commitments in the fight against climate change to reduce these emissions by 51% by 2030[10].
Methodology
Energy service data analysis
Initially, nine technical and commercial variables on electricity service in the ZNI localities were obtained from the Unique Information System (SUI) of the Superintendence of household public services (SSPD) [11], reported by the service operators. Data were cleaned, leaving only stratum one residential users (about 99% of users) and only localities with less than 500 kW of installed capacity. Localities with implemented or planned photovoltaic solutions were also discarded. It results in 1553, 1618, 1608, and 1589 localities, respectively, for four moments: Mar-2019, Sep-2019, Mar-2020, and Sep-2020. Missing data was filled in different forms, as recommended in [12]. COE missing data were filled according to values in localities with the same service operator, and billing was filled using energy data and COE. For diesel unit and transport costs, missing data were filled according to localities with the same fuel collection site and municipality. For some localities with more than one Genset, only one was taken according to with maximum power reported for these localities in CNM telemetry reports [13]. The analysis of the four moments results in 1489 localities with data at each moment, showing an invariable behavior of the data according to the P-value test, which allows selecting only one data set, being the selected one the one corresponding to Sep-2020.
The nine variables were normalized by the Z-score method to avoid the dominance of those with higher ranges or scales. Subsequently, they have been dimensionally reduced by Principal Component Analysis (PCA), and some outliers have been removed by the percentile method. The first three principal components add up to a variance higher than 90% and are the ones selected to analyze the data. Finally, using the k-medoids method and the Partitional Around Medoids (PAM) algorithm, using a Square Euclidean distance, clustering the data has been made according to the values obtained for the 3 PCs representing the highest variance.
Analysis of waste biomass data and its energy potential
The residual biomass data were obtained from the municipal agricultural assessments (EVA) projected for 2020. [14] for all the municipalities that contain the localities of the sep-2020 dataset, for a total of 1522 localities with energy service data and residual biomass data considered for energy use [7]. The EVAs' relevant data for estimating residual biomass's energy potential are the cultivated area, the yield per hectare, and the annual tons of production. These data, together with the product/residue ratio taken from [7] and a residual biomass harvesting factor of 0.8 [15], [16], allow estimating the theoretical energy potential of these residual biomasses, similar to other published works. [17], [18]. The lower calorific value of each waste is established according to different references and estimates [7], [19]–[24]. Finally, the technical energy potential for biomass gasification is estimated using a conversion efficiency of 13%, the minimum value referenced for downdraft biomass gasifier systems coupled to internal combustion engines [25], [26], the most suitable for small-scale power generation from biomass [27]–[29].
The residual biomass energy potential data are then obtained for each municipality with localities in ZNI and normalized similarly to the energy service data. Allocation of energy potential for each locality is made by weighting the number of stratum one users of the service in each locality of each municipality, assuming that the area of their agricultural activities is proportional to the number of users and people in each one.
Results and conclusions
The analysis of energy service data in the localities of the ZNI has made it possible to determine clusters that group these localities into five typologies according to their technical and commercial energy service data. Furthermore, the characteristics of each cluster have made it possible to establish two deficit indicators, one for users served and the other for diesel fuel deficit. Finally, a weighted aggregation was used to generate an energy service deficit indicator for each locality according to its typology or cluster.
The analysis of residual biomass energy potential data has made it possible to establish a possible energy margin indicator between the potential for each location and the theoretical electrical energy to be supplied by the operators in each location.
These two indicators have been weighted-aggregated, and using adjustment factors; it has been possible to establish an indicator of the relevance of energizing localities in ZNI from biomass gasification based on public data and analysis of these data. This indicator allows identifying those localities where this possible energy solution deserves to be analyzed in terms of its feasibility and sustainability, achieving greater precision when considering possible projects in these localities. Likewise, this methodology can be replicated for other renewable energy sources such as photovoltaic or thermal solar energy or livestock waste. Finally, the results show how the data analysis is a helpful tool for studying energy solutions aligned with the fulfillment of national goals in the fight against climate change in isolated localities.
| Original language | American English |
|---|---|
| State | Published - 2022 |
| Event | 8th Encuentro Latinoamericano de Economía de la Energía - ELAEE : Energy Transition in Latin America - Universidad Jorge Tadeo Lozano, Bogota, Colombia Duration: 20 Nov 2022 → 22 Nov 2022 Conference number: 8 https://www.utadeo.edu.co/es/8elaee |
Conference
| Conference | 8th Encuentro Latinoamericano de Economía de la Energía - ELAEE |
|---|---|
| Abbreviated title | ELAEE |
| Country/Territory | Colombia |
| City | Bogota |
| Period | 20/11/22 → 22/11/22 |
| Internet address |
Enfoques Temáticos Institucionales
- Transición energética
Research Results
- Scientific events with appropriation component with Quality A
