Strategic low-cost energy investment opportunities and challenges towards achieving universal access (SDG7) in each African nation

Strategic energy planning to achieve universal access and cover the future energy needs in each African nation is essential to lead to effective, sustainable energy decisions to formulate mitigation and adaptation climate change policy measures. Africa can not afford a cost-increasing green energy transition pathway towards achieving SDG7. In this analysis, least-cost power generation investment options using energy systems analysis enhanced with geospatial data for each African nation are identied, considering different levels of electricity consumption per capita (Low, High) and costs of renewables (New Policies, Renewable Deployment scenarios). The power generation capacity needs to increase between 211GW (NPLs) and 302GW (RDHs) during 2021-2030 to achieve SDG7 in Africa, leading to electricity generation to rise between 6,221PJ (NPLs) - 7,527PJ (NPHs) by 2030. Higher electricity consumption levels lead to higher penetration of fossil fuel technologies in the power mix of Africa. To achieve the same electricity demand levels, decreasing renewables' costs can assist in a less carbon-intensive power system, although higher capacity is needed. However, Africa is still hard to achieve its green revolution. Depending on the scenario, grid-connected technologies are estimated to supply approximately 85%-90% of the total electricity generated in Africa in 2030, mini-grid technologies roughly 1%-6%, and stand-alone technologies 8%-11%. Solar off-grid and solar hybrid mini-grid technologies play an essential role in electrifying the current un-electried settlements in residential areas. Natural gas will be the dominant fossil fuel source by 2030, while the decreasing costs of renewables make solar overtake hydropower. Higher penetration of renewable energy sources in the energy mix creates local jobs and increases cost-eciency. Approximately 6.9 million (NPLs) to 9.6 million (RDHs) direct jobs can be created in Africa by expanding the power sector during 2020-2030 across the supply chain. Increasing the electricity consumption levels in Africa leads to higher total system costs, but it is estimated to create more jobs that can ensure political and societal stability. Also, the decreasing costs of renewables could further increase the penetration of renewables in the energy mix, leading to a higher number of jobs. in this study investigated the implications of the future energy transition in Africa on the electricity supply mix, the total system costs, achieving SDG7 and its sub-targets, the associated environmental and socio-economic implications. The different universal access scenarios were examined to analyse the economics and demographic factors as demand drivers to boost electrication in each African nation. A critical dimension of achieving SDG7 in Africa, except the different electricity consumption per capita levels each African country wants to achieve, is the evolution of renewable technology costs. Africa can not afford a cost-increasing green energy transition pathway. SDG13, SDG14) RDHs the in by 2030, of the

governments, policy makers and funders to identify proper nancing mechanisms and strategic investments [20]. They can also measure the socioeconomics (e.g., job creation) of different energy transition pathways [21], [22]. Lastly, considering the spatial dimension of energy access (location and size of un-electri ed population, geophysical parameters and technology costs) can improve the medium-to-long-term assessments of least-cost power system expansion in an energy model [23]- [25].
Several techno-economic studies have been conducted in the past in a continental, regional and national scale to address the future challenges associated with the evolution of the energy system in Africa. Trotter et.al [26] conducted a literature review on quantitative and qualitative electricity planning and implementation research approaches on Sub-Saharan Africa on a national, regional and continental scale. Their review indicates that 63% out of 306 relevant peer-reviewed journal articles favour renewable energy technologies for the identi ed challenges, mentioning success factors for electri cation in sub-Saharan Africa to include adequate policy design, su cient nance and favourable political conditions. On a continental scale, using a cost-optimization tool OSeMOSYS [27], Taliotis et al. [28] modelled the electricity system of each African country (45 in total) and linked it via electricity trade links to examine scenarios of power plant investments by exploiting trade potential in the continent during 2010-2040. They show that an enhanced trading scheme could reduce electricity generation costs. However, only the electricity sector on a national scale is modelled and not the rest of the sectors to satisfy the nation´s whole energy demand and examine the associated implications leading to changes in the electricity generation costs. Furthermore, only the electricity trade links included in the model and not other trades (e.g., gas) and the number of power-generating technologies are less than our study, decreasing the model's granularity. Bazilian et al. [29] examined various energy access scenarios in Sub-Saharan Africa by 2030 using the OSeMOSYS framework to model the electricity sector. Although in this study, a detailed analysis of the energy demand projections is presented, missing from Taliotis et al. study [28], other fuel demands are not included in the study to better capture the evolution of the energy system in Africa and the geospatial allocation of power generation technologies is missing. In another study, a cost-optimization modelling framework MESSAGE-SPLAT used to model only the power system of two regions (Eastern and Southern Africa) and examine scenarios associated with their energy transition and not the rest [30].
Puig et al. [31] suggested an action agenda for Africa´s electricity sector identifying economic opportunities associated with the power-sector reform, energy access and investing in Sustainable Development Goal 7. Expansion of investments in off-grid and interconnected clean-energy mini-grids are one of those.
Enhancing energy systems analysis with geospatial data, Mentis et.al [20] in their study developed a bottom-up geospatial electri cation tool (OnSSET) to examine least-cost electri cation strategies for each Sub-Saharan country considering investment options (grid extension, mini-grid, stand-alone systems) for different electricity consumption levels. Nevertheless, the OnSSET model, which also uses the TEMBA model as an input for the grid cost, is further improved in this study using the open-access Global Electri cation Platform (GEP) [32]. Also, the authors in their research did not present the grid-connected power generation mix for the different universal access scenarios and the analysis of the electricity demand projections primarily focused on connecting the un-electri ed population. Similarly, Longa et al. [33] applied geographic information system analysis coupled with an integrated assessment modeling to study electricity access in Africa through scenario projections until 2050, focusing on climate policies (2.0 o C). They conclude that universal electricity access in Africa cannot be achieved by 2030 but primarily until 2050. Also, they conclude that off-grid renewable energy technologies are essential to increase signi cantly the electri cation of rural areas. However, the study only examines the electricity supply system of 48 African countries and the citizen's willingness to pay for electricity. In our study, the analysis is broader covering the whole energy supply system of the African countries together with energy demand projections by 2040. Also, the focus is except from de ning the least-cost electri cation mix for each country also informing indicators relevant to SDG7 and its sub-targets and provide insights on job creation potential in Africa. Other studies also focused either on a regional or a national scale analysis. Falchetta et.al [34] modelled only the electricity supply system of East Africa (OnSSET) to estimate least-cost pathways to achieve universal access by 2030 under different electricity consumption levels. Nevertheless, this study doesn't consider the trade implications with the other power pools and doesn't include a detailed representation of the grid power generation mix and consequently how the grid cost is affected. Rocco et. al [35] and Pappis et al. [25] examined energy access scenarios for Tanzania and Ethiopia, accordingly modeling only the power sector using OSeMOSYS and geospatial data to determine the electricity demand projections and the optimal allocation of power generation technologies. Similarly, the analysis is conducted only on a national level, without considering the trade implications on a continental scale. Only the power sector is modelled and the scenarios are limited by only examining energy access considering policy and environmental goals and not the evolution of other fuels.
The Africa Energy Outlook 2019 [36] focused on electricity access in Sub-Saharan Africa by examining two scenarios, the Stated Policies and the Africa case, the period 2018-2040. Although the study covers the whole of Africa, the detailed electricity supply system of only 11 sub-Saharan countries was modelled using both a cost-optimization tool and a geospatial electri cation tool (OnSSET). In the Stated Policies scenario, hydropower is expected to be the primary fuel source by 2040, while its share declines as natural gas and other renewables expand. Contrary, in the Africa case, hydropower is being overtaken by natural gas. Solar PV capacity increases signi cantly, almost 40%-70% of all new capacity additions in the power system in each scenario, respectively. Geospatial analysis shows that the least-cost way to achieve universal access by 2030 is decentralized systems powering half of the electricity connections (nearly 440 million people) and extending the main grid. that although Africa continues to progress toward SDG7 still a lot of effort is required to reach the goal by 2030 (Stated Policies). The projections show that under current and planned policies and before the start of the COVID-19 crisis, about 555 million people would still lack access to electricity in 2030. Almost 20 countries would have less than half of the population electricity access while ten less than one in four (e.g., Chad, the Democratic Republic of Congo, Malawi, Niger and Somalia) [8]. To bridge the gap, almost 85 million people per year need to gain access from 2020 (the access rate to triple) to 2030. IEA´s and IRENA´s [37] scenarios estimate that achieving SDG7 would require annual investments of approximately $680 billion to renewable energy by 2030, around $45 billion spent in energy access and $625 billion to energy e ciency [8].
Lastly, few studies have been conducted to track the progress towards achieving SDGs in African countries [38], [39]. Also, another aspect to examine relevant to the energy transition on a global, regional or national level is the associated job creation potential of achieving universal access, which is missing from the literature [21], [22], [40]- [42].
Based on the existing literature, this study builds on previous efforts by Pappis et al. [43] modelling the electricity supply system of all African countries (48 in total) and linked it via electricity and trade links, including also all fuel demands and fuel exports, to examine scenarios relevant to achieving universal access ( Table 1). The results are provided on an annual basis from 2015 to 2030. This study also considers the geospatial allocation of power generation technologies missing from the previous analysis on Africa and examines different research questions.
In this study, therefore, we tried to address the research gaps in the current literature by examining the following research questions: First, how can an African country could identify energy pathways that are least-cost but also consider the nation´s priorities in terms of achieving universal access, the geospatial allocation of power generation investments, nancial capacity, technological maturity, environmental and policy constraints, and demand growth in different sectors? Second, what the outputs of these least-cost energy pathways using energy system modelling tools would mean for SDG7 and its sub-targets (rate of success) in each country? and Third, what will be the socio-economic bene ts focusing on job creation of these energy transitions in Africa. In this study, an electri cation least-cost investment outlook on a continental, regional and national scale for Africa, using energy systems analysis enhanced with geospatial data, is developed for 2015-2030. Four scenarios developed (New Policies Low, New Policies High, Renewable Deployment Low, Renewable Deployment High) focusing on different energy access consumption levels (low, high) and examining the effect of decreasing the costs of renewables in the energy transition in Africa. The four scenarios examined in this study investigated the implications of the future energy transition in Africa on the electricity supply mix, the total system costs, achieving SDG7 and its sub-targets, the associated environmental and socio-economic implications. The different universal access scenarios were examined to analyse the economics and demographic factors as demand drivers to boost electri cation in each African nation. A critical dimension of achieving SDG7 in Africa, except the different electricity consumption per capita levels each African country wants to achieve, is the evolution of renewable technology costs. Africa can not afford a costincreasing green energy transition pathway.
The analysis estimates the cost-optimal mix of electri cation technologies and fuel supply consistent with the corresponding whole energy system development pathways using the OSeMOSYS tool [27] and captures the spatial distribution (split of grid, mini-grid and stand-alone technologies) of future electricity connections in the residential areas using the open-access Global Electri cation Platform (GEP) [44]. In that way, grid, off-grid and minigrid energy systems' role in meeting SDG7 can be examined and provide insights into energy transition and its associated challenges. An accounting model was developed to create an index with indicators to inform each African country's SDG7 and sub-targets. Lastly, an input-output model was used to measure the socio-economic transition focusing on job creation in Africa. The open-source nature of this analysis (data, model, code, results) assists in the transparency and freely reproducibility of the research. The analysis provides results at continental, regional and national scale in an integrated way for the whole of Africa to inform national policy analysis (Nationally Determined Contributions [4], Clean Development Mechanism [45]), SDG7 and other closely interlinked targets (SDG1, SDG3, SDG11, SDG13, SDG14) [2] and be used by academics, researchers and policy analysts in their research and capacity building activities. The outcomes of this study can assist each African nation in strengthening risk mitigation associated with future renewable energy projects and ensuring their nancial viability. The methodology used to conduct this research is implemented in four stages. Firstly, the existing energy model for Africa developed using OSeMOSYS [47] has been further updated and calibrated to represent the off-grid technologies (mini-grid, stand-alone) and the residential electricity demand for the currently un-electri ed settlements in Africa. Also, the previous model does not consider the geographical characteristics of the resources and the residential electricity demand's spatial dimension to identify the least-cost split between on-and off-grid technologies. Thus, in the second step, the  The open-source energy modelling system (OSeMOSYS) tool is a freely available optimization modelling framework for medium to long-term energy planning [27]. It is a bottom-up modelling framework that uses linear-optimization techniques to satisfy an exogenously de ned energy demand.
OSeMOSYS has been employed in the scienti c literature [21], [33] and in academic teaching and capacity building for energy planners to provide insights on possible transformation trajectories of both country to continental-scale energy systems [49]. The objective function equation consists of the sum of discounted operational and capital costs. The energy system consists of nal energy demands distinguished between various end-use services, transmission and distribution networks, power generation technologies (on-grid and off-grid), energy trade links, conversion technologies, and technologies representing imports and extraction of energy resources. The modelling results can include power generation capacity, production by technology, operation and maintenance costs, and emissions on an annual level with a timely resolution for some of the variables.

Geospatial electri cation outlook (GEP)
The OnSSET tool considers population settlements and, for each one, determines which is the least-cost technology to meet the electricity demand (grid, mini-grid, or stand-alone). The Levelized Cost of Electricity (LCOE) is calculated for each technology option, considering the investment, operation and maintenance, and fuel costs in each African country. The technology that can meet the demand at the lowest LCOE is selected as the least-cost technology option in that settlement [20]. The OnSSET tool has been used in national electri cation studies (Malawi [50], Kenya [51], Tanzania [52], Afghanistan [53]) and regional electri cation studies for the whole of Sub-Saharan Africa ( [20], [54]). The latest Global Electri cation Platform explores least-cost electri cation strategies for 58 countries [32]. Integrating the spatial dimension of energy access (location and size of un-electri ed population, geophysical parameters and technology costs) can improve the medium-to-long-term assessments of least-cost power system expansion [23]- [25].
The formula used to calculate the Levelized cost of generating electricity for each technology is the following: OnSSET is not an energy system-wide optimization model, and it considers only the future connections of the current un-electri ed settlements in the residential areas. Thus it does not represent the evolution of the grid-based electricity generation mix, which is a modeling outcome of the OSeMOSYS tool. The analysis of the grid components is missing from GEP [32].
By linking the two tools, OSeMOSYS and GEP, the electri cation mix between grid (in terms of share), mini-grid and off-grid technology-speci c which supplies the residential electricity demands in the current un-electri ed population de ned in GEP is fed into the OSeMOSYS model to identify the leastcost electri cation mix (grid components) [20], [25], [51].

Index framework informing SDG7
In this study, the targets and the associated sub-targets of Sustainable Development Goal 7 are informed for each one of the African countries based on the modelling outputs of the energy transition for each scenario in 2030. An index framework is created based on the SDG7 targets focus, as presented in Table 3. The share of renewables and the amount of CO 2 emissions are de ned only for the power sector. The energy intensity is estimated as the total primary energy supply per GDP. The lifetime of fossil fuel resources (%) is calculated as the production of fossil fuel reserves during 2015-2030 per each country´s total amount of identi ed fossil fuel reserves. The import dependency (%) is calculated as the total net imports (imports minus exports) of coal, crude oil, oil products and natural gas as a share of the total primary energy supply in 2030.

Job creation potential
An analytical approach for the energy supply system of Africa is adopted to estimate direct energy jobs creation corresponding to the value chain of the energy transition of Africa. The methods by Rutovitz et al. (2015) [41], [42] are applied and adjusted to estimate job creation for the African continent using the associated techno-economic assumptions of the modeling inputs of this study. The Employment Factor (EF) method was applied amongst the other methods [56] primarily due to its simplicity and effectiveness in estimating direct employment associated with energy generation, storage, exibility options and transmission and distribution of electricity. The EF approach is preferable to other methods since it can be modi ed for speci c contexts and applied over a range of energy scenarios [57]. The total direct jobs are estimated considering a sum of jobs in manufacturing, construction and installation, operations and maintenance, fuel supply associated with electricity, decommissioning of energy plants at the end of their lifetimes and transmission and distribution of electricity [42]. This approach is brie y presented in Figure 1 and the methodology is further explained in detail in the Supplementary Material A.

Scenarios
The following four scenarios are examined for the modelling period 2015-2040 but providing results only until 2030, focusing on achieving universal access in Africa by 2030. The changes in the input parameters among the scenarios will provide a broader understanding of how universal access can be achieved in Africa and identify low-cost investments in power generation technologies. As the African nations aim to decrease their carbon dioxide emissions in the future (NDCs), how renewables can assist in this transition and the evolution of their costs is a challenge that needs to be examined.
The national electri cation rates of each African country are considered in this analysis. The National policies and SDG7 targets for each one of the African countries are collected and analyzed in each of the scenarios. The scenarios consider the national policies modelled as RET targets for each African country adopted until 2020. The main differences among the New Policies and the Renewable Deployment scenarios are the electri cation mix for the current un-electri ed settlements in the residential areas derived using the OnSSET-GEP tool [20], [32] and lower costs for renewables. In the model, electricity to the current un-electri ed settlements in the residential areas starts supplying from 2020 onwards. The energy transition pathways examined under the Renewable Deployment scenario assist in addressing the challenges associated with SDG7, SDG1 (no poverty), SDG3 and SDG11 (reducing impacts of air pollution) and SDG13 (tackling climate change). In addition, different electricity consumption levels (Low, High) for the unelectri ed settlements in the residential areas are considered for each of these scenarios ( Figure 2). The modeling assumptions can be found in the New Policies -High un-electri ed settlements residential demand scenario (NPHs): the only difference with the NPL scenario is the different tiers of electricity in the household's current un-electri ed settlements in the residential areas. In this scenario, the rural household areas which are currently un-electri ed reach Tier 3. Respectively, the households in urban areas get to one Tier higher than each African country's respective electricity consumption per capita.
Renewable Deployment -Low un-electri ed settlements residential demand scenario (RDLs): It aims to combat climate change by considering lower renewable energy technology costs than the New Policies scenarios and examining their effect. The electri cation mix is different than the New Policies scenario. In this scenario, the current un-electri ed settlements get Tiers of electricity similar to the NPLs scenario.
Renewable Deployment -High un-electri ed settlements residential demand scenario (RDHs): The techno-economic assumptions are similar to the RDLs scenario. The electri cation mix is different than the New Policies scenario. The demand projections are similar to the NPHs scenario meaning that the un-electri ed settlements in the residential areas will get higher Tiers of electricity than the SDLs scenario.

Results
Overall technology (installed capacity) mix Overall, to fully electrify Sub-Saharan Africa and meet the future energy needs of the continent, the total installed capacity in the continent needs to increase from 178GW in 2015 to 389GW (NPLs), 473GW (NPHs), 403GW (RDLs) and 492GW (RDHs) in 2030 depending on the scenario. This capacity growth is primarily due to changes in electricity demand levels and renewable technology costs. Speci cally, the capacity of fossil fuel technologies increased from 143GW in 2015 to 195GW (50%) in NPLs, 212GW (45%) in NPHs, 198GW (49%) in RDLs and 214GW (43%) in RDHs, in 2030. Natural gas constitutes most of the fossil-fuel installed capacity. In the opposite case, most of the renewable capacity in the continent is based on hydropower, although the capacity of solar technologies grows very fast. As the costs of renewables decrease further by 2040, the share of fossil fuels in the power system of Africa decreases even more in the Renewable Development scenarios than in the New Policies ones. Also, as the electricity demand increases between the scenarios (Low to High), higher investments in fossil fuel technologies are required to satisfy the nal electricity demand levels since renewables are not always available to generate electricity.
Solar off-grid and solar hybrid mini-grid technologies are expected to gradually penetrate the power system in Sub-Saharan Africa and play an essential role in its future energy transition. The current un-electri ed settlements will start getting electricity in 2020. This growth in electricity levels results in Most of the total installed capacity in the continent was located in SAPP of 61GW (34%) in 2015. Nevertheless, to achieve SDG7 in Africa by 2030, the EAPP is estimated to represent most of the continent´s installed capacity, around 37% in all scenarios due to currently low electricity access levels and high population increase. This energy transformation in EAPP is led by hydropower investments in Ethiopia, Sudan, Tanzania, and Egypt's fossil fuel and solar technologies growth. Also, CAPP is expected to have the higher share of renewables in the continent, mainly due to hydropower and solar potential. The overall technology mix in Africa among the scenarios from 2015 to 2030 is presented in Figure 3 and for each power pool in Supplementary Material B.
Energy supply mix -Electri cation mix To fully electrify Africa and cover the continent's future energy needs, the total primary energy supply in the continent from 674 Mtoe in 2015 increased to a range of 1,312Mtoe (RDLs) to 1,374Mtoe (NPHs) in 2030, depending on the scenario (Figure 4). In the Renewable Development scenarios, less supply of fossil fuels is needed as in the New Policies scenarios (Supplementary Material B). Although by 2030, the penetration of renewables in the electricity mix of the continent does not signi cantly change, as their cost decreases in the Renewable Development scenarios than the New Policies scenarios by 2040, this transition is more evident.
The total electricity generation in Africa increased from 2,704PJ in 2015 to 6,221PJ (NPLs), 7,527PJ (NPHs), 6,188PJ (RDLs) and 7,487PJ (RDHs) in 2030, depending on the scenario. Out of the total electricity generation, the fossil fuels share constituted 81% in 2015, decreases to 63% (NPLs), 60% (NPHs), 63% (RDLs) and 60% (RDHs) in 2030. Natural gas is estimated to be the primary fossil fuel in the continent in the next decade. In the opposite In this power pool, the role of electricity interconnectors is highlighted to achieve universal access. Speci cally, the gas-based generation is higher in the Renewable Development scenarios than in the New Policies ones, under respective electricity demand levels . This energy transition is due to the increase in the electricity supplied by gas power plants in Cameroon to satisfy part of its domestic consumption while its electricity imports from Chad decline by a signi cant margin. As electricity demand levels increase between the scenarios, DRC, except for covering part of its domestic electricity consumption from coal-based power plants and solar hybrid mini-grid systems, imports more electricity from Angola, Congo, Rwanda, and Zambia to also maintain its electricity exports at similar levels(cumulatively around 108PJ).
In EAPP, the electricity generation increases from 786PJ in 2015 to 2,241PJ (NPLs), 2,640PJ (NPHs), 2,224PJ (RDLs) and 2,633PJ (RDHs) in 2030. The fossil fuel share decreases from 79% in 2015 to 58% (NPLs), 52% (NPHs), 58% (RDLs) and 53% (RDHS) in each of the scenarios in 2030. Gas is expected to be the dominant fuel in the region in the next decade as in 2015, mostly of the gas-based electricity generation in Egypt, while coal from 2023 onwards increases its share in the electricity mix by a big margin due to coal investments in Egypt. Egypt needs to import coal in the future to generate electricity due to its limited availability of identi ed domestic coal reserves, affecting its import dependency. However, the government could use its natural gas reserves instead to strengthen the reliability of the power system and not be affected by the uctuation of fossil fuel prices. Except for fossil fuel investments in Egypt, the RET-based generation in the country is expected to increase by almost seven times. As electricity demand increases, Egypt decreases its imports cumulatively from 2015-2030 while it increases its natural-gas-based electricity generation. In Ethiopia, although the RET share increases signi cantly, speci cally hydropower, solar and geothermal, among the scenarios relatively as electricity demand increases, in the NPHs and RDHs scenarios, the country also starts producing electricity from natural gas power plants from 2028 onwards. To cover the increased fuel needs in the future (NPHs, RDHs), the country also reduces its electricity net exports to neighboring countries to even higher levels than the current onesin the NPLs, RDLs. The country also assists Kenya in achieving universal access. Hydropower and solar are the dominant fuels in Ethiopia, Kenya, Sudan, Tanzania and Uganda by 2030. , Tanzania is another country where as electricity demand increases (NPHs, RDHs) the country further exploits its domestic coal reserves to increase its coal-based electricity generation from 2022 onwards and decrease its net imports cumulatively almost by 30% (2015-2030). In the opposite case, under higher electricity demand levels (NPHs, RDHs), Rwanda increased their electricity generation by increasing their gas-based electrity generation from 2021 to increase its electricity exports primarily to Tanzania and Uganda. . However, this energy transition comes at the cost of increasing its carbon dioxide emissions.
WAPP presents the highest increase in its electricity generation between 2015 and 2030 in Africa. The electricity generation increases from 247PJ in 2015 to 847PJ (NPLs), 1262PJ (NPHs), 842PJ (RDLs), 1253PJ ( RDHs) in 2030. The fossil fuels share decreased from 72% in 2015 to 56% in NPLs, NPHs and 58% in RDLs and 56% in RDHs in 2030. Gas is the dominant fuel in the region primarily due to investments in Cote D Ivoire, Ghana, Nigeria and Ghana by 2030 with an increased share of hydropower in Nigeria, Cote D Ivoire and Guinea, and solar off-grid and mini-grid technologies. As electricity demand rises among the scenarios, Nigeria is estimated to increase its coal-based power generation in 2022. However, the country decreases its net exports between 2015-2030 to satisfy the high increase in its domestic electricity consumption leading the electricity importers Benin and Niger to increase their gas-based and hydropower generation in the future.
The energy balances on a continental level for the different scenarios the period 2015-2030 are presented in Supplementary Material B.
Investment needs for achieving universal access in Africa The total system costs of an energy system consist of the capital investments, operating and maintance and operating fuel costs for all grid-connected, mini-grid and off-grid technologies and the transmission and distribution (T&D) infrastructure. Thus, the minimum total system costs required to fully electrify Africa and cover the future electricity needs in the continent in the period 2020-2030 amount to 2,973 billion USD at the Renewable Development Low scenario (lowest electri cation level). In the opposite case, the maximum total system costs correspond to 3,489 billion USD at the  Figure 5.
The average cost of generating electricity per kWh each year in each African country's scenarios is presented in Table 4, over the periods 2015 -2020, 2020 -2030. The costs in each country may vary among the scenarios since there are cases (as presented in Section 3.2) a nation to increase its generation costs in assisting another country in satisfying its future electricity needs. The total system costs in this study are minimized on a continental scale and not on a country level. The average cost of generating electricity is the yearly ratio between the expenses incurred during that period (investment, operation, carbon tax) and the electricity generated. Higher average costs of generating electricity are primarily in the New Policies scenarios than the Renewable Development scenarios primarily due to higher penetration of fossil fuel technologies resulting in higher fuel operating costs. On the other hand, higher upfront capital investments are required for renewable technologies in the Renewable Development scenarios. Table 4. The average cost of generating electricity per kWh over the periods (2015-2030,) (cent USD/kWh). The costs are discounted assuming an average discount rate of 8%; they include the power supply grid-connected technologies. Sustainability insights for achieving SDG7 in each African nation Achieving SDG7 in each African nation by 2030 will have different implications for the targets and sub-targets associated with SDG7 and each nation´s energy transition. The modelling results below can assist each country in understanding the con icting objectives among the evolution of the power system with the energy indicators mentioned above. Speci cally, although Benin, Cote D Ivoire, Equatorial Guinea and Ghana will increase their renewable energy targets by 2030, this energy transition consumes most of their respective fossil fuel reserves to cover their future energy needs, negatively affecting their net import dependency in the future. As a result of this analysis, the evolution of some indicators: the share of renewables, CO 2 emissions, energy intensity (energy production/GDP), the lifetime of fossil fuel resources and import dependency calculated for each African country is presented in Table 5. Note: 1 Solar hybrid technologies included in renewables, 2 Total primary energy supply per GDP [59], 3 Production of fossil fuel reserves during 2015-2030 per each country´s total amount of identified fossil fuel reserves, 4 Total net imports (imports minus exports) of coal, crude oil, oil products and natural gas as a share of total primary energy supply in 2030. Negative values correspond to net exporters. The differences in the RET share may occur since the electricity generation is not always the same among the scenarios and the supply mix differs since the electricity trading scheme changes among the scenarios

Environmental implications
The total carbon dioxide emissions of the evolution of the energy system in Africa increased from 1,213Mton in 2015 to 2,797Mton (NPLs), 2,9431Mton (NPHs), 2,793 (RDLs) and 2,919Mton (RDHs) depending on the scenario in 2030. NAPP is estimated to represent most of the continent´s carbon dioxide emissions in the future, although the share of renewable energy technologies increases followed by EAPP, SAPP, WAPP and CAPP (Supplementary Material B). Higher electricity consumption levels lead to higher carbon dioxide emissions. However, lower carbon dioxide emissions are emitted as the renewable technology costs decrease (RDLs, RDHs). Thus, a decreased cost of renewables could further assist an African nation in evolving its power sector to achieve universal access and achieve part of the National Determined Contribution greenhouse gas emissions targets.

Socio-economic implications: Job creation
Previous sections show that Africa needs energy to grow, which has environmental implications and is cost expensive. However, this energy transition and electricity access can create several jobs than lost. Different levels of jobs can be created associated with the construction and installation of the power generation technologies to the fuel use and speci cally to the use of power generation technologies. In Africa, approximately 6.9 million direct jobs can be created by expanding power generation capacity and T&D network in the NPLs scenario, 8.7 million jobs in the NPHs scenario, 7.0 million jobs in the RDLs scenario and 9.6. million jobs in the RDHs scenario during 2020-2030 across the supply chain of the evolution of the power sector ( Figure 6). Of this total number of jobs in each scenario, 6.4 million jobs in NPLs, 7.8 million jobs in the NPHs, 6.5 million jobs in the RDLs and 8.7 million in the RDHs scenarios accordingly are associated with the future installation and operation of speci c power generation technologies. Solar power is expected to be the dominant technology in creating new employment opportunities (Figure 7). It is assumed that the manufacturing happens only locally and is not created from exports to other countries. In the Renewable Deployment scenarios, more jobs are created in the manufacturing (local) sector, construction&installation, and operation and maintenance but lower jobs on transmission since the future installed capacity is less due to higher penetration of off-grid renewable technologies. However, higher fuel jobs are created in the RDLs and RDHs scenarios than in the New Policies scenarios, primarily due to the fuel used in coal power plants until 2030. Increasing the share of renewables can boost employment in Africa, while fossil fuel development can support jobs in different ways. Further increasing the electricity consumption levels in Africa (NPHs, RDHs) is estimated to create more jobs. Solar hybrid mini-grid systems are not included in the analysis of job creation potential.

Conclusions
This study highlights the importance of strategic energy planning to achieve universal access in each African nation and cover its future energy needs considering their socio-economic development until 2030. Exploiting the energy resources of Africa strategically to expand its power system can lead to effective, sustainable energy decisions to formulate mitigation and adaptation climate change policy measures. The energy transition is complex and has several implications for a nation´s economy and future development. In this analysis, least-cost power generation investment options using energy systems analysis enhanced with geospatial data for each African nation are identi ed, considering different levels of electricity consumption per capita (Low, High) and costs of renewables (New Policies, Renewable Deployment). The four scenarios examined in this study investigated the implications of the future energy transition in Africa on the electricity supply mix, the total system costs, achieving SDG7 and its sub-targets, the associated environmental and socio-economic implications. A critical dimension of achieving SDG7 in Africa, except the different electricity consumption per capita levels each African nation wants to achieve, is the evolution of renewable technology costs. Africa can not afford a cost-increasing green energy transition pathway. This study develops country-speci c electri cation investment outlooks to assist each African nation's government o cials and policy makers in strengthening risk mitigation associated with future renewable energy projects and ensuring their nancial viability.
The power generation capacity in Africa needs to increase approximately 2 -2.6 times in the NPLs and NPHs scenarios accordingly, while 2.3-2.8 times in the RDLs and RDHs scenarios, to cover its future energy needs and achieve SDG7 between 2015-2030. Natural gas constitutes most of the fossil-fuel installed capacity in all scenarios. In the opposite case, hydropower is currently the dominant renewable energy technology. It will remain by 2030 if the costs of solar technologies do not signi cantly decrease in the upcoming decade. Otherwise, as it is shown in the RDLs and RDHs scenarios, solar power can be the leading power generation technology in Africa by 2030, building climate change resilience of the system. Solar off-grid and solar hybrid mini-grid technologies are expected to play an essential role in the electri cation of the current un-electri ed settlements in the residential areas.
The total primary energy supply in the continent is estimated to increase from 674Mtoe in 2015 to a range of 1,312Mtoe to 1,374Mtoe in 2030, depending on the scenario. The total electricity generation in the continent needs to increase by 2.3 times in the NPLs, RDLs scenarios and 2.8 times in the NPHs, RDHs scenarios, respectively, by 2030. The fossil fuels share constituted 81% in 2015, decreases to 63% (NPLs), 60% (NPHs), 63% (RDLs) and 60% (RDHs) in 2030. Higher electricity consumption levels lead to higher penetration of fossil fuel technologies in the power mix of Africa. However, to achieve the same electricity demand levels, decreasing renewables' costs can assist in a less carbon-intensive power system, although higher capacity needs to be installed. Depending on the scenario, grid-connected technologies are estimated to supply approximately 85%-90% of the total electricity generated in Africa in 2030, mini-grid technologies roughly 1%-6%, and stand-alone technologies 8%-11%. Solar off-grid technologies supply electricity approximately 8% (NPLs), 9% (NPHs), 10% (RDLs) and 11% (RDHs) of the total generated electricity in Africa and solar hybrid mini-grid technologies 2% (NPLs), 5% (NPHs), 1% (RDLs) and 3% (RDHs) by 2030.
Achieving lower-higher electricity demand levels and the costs of renewables could transform countries accordingly into net importers or exporters depending on their future energy choices (e.g., Tanzania). Also, based on the future energy investments of each nation, achieving SDG7 in an African country may have collective implications to several factors (e.g., the share of renewables, CO 2 emissions, energy intensity, the lifetime of fossil fuel resources, import dependency) and achieving one target may have con icting objectives with others on a local level but also a regional one. Benin, Cote D Ivoire, Equatorial Guinea and Ghana are some countries that increase their renewable energy targets by 2030. However, this energy transition consumes most of their fossil fuel reserves to cover their future energy needs, negatively affecting their net import dependency in the future. The outcomes of this analysis, achieving SDG7 in Africa, can provide insights to other SDGs such as SDG1, SDG2, SDG3, SDG6, SDG8 and SDG13 [3].
Higher penetration of renewable energy sources in the energy mix reduces dependence on imported fuels, creates local jobs, and increases cost e ciency. Although higher up-front capital investments in renewables are required, the operating fuel costs are lower in the long term. Thus, the total system costs required to fully electrify Africa and cover the future electricity needs in the continent during 2020-2030 amount to 3,000 billion USD in the Africa needs energy to grow, which has environmental implications and is cost expensive. However, this energy transition and achieving electricity access can create several jobs than lost. In Africa, approximately 6.9 million direct jobs can be created by expanding the power generation capacity and the T&D network in the NPLs scenario, 8.7 million jobs in the NPHs scenario, 7.0 million jobs in the RDLs scenario and 9.6 million jobs in the RDHs scenario during 2020-2030 across the supply chain. The increased share of renewables in the energy transition in Africa can boost job creation, while fossil fuel development can support jobs in different ways. Increasing the electricity consumption levels in Africa (NPHs, RDHs) leads to higher total system costs, but it is estimated to create more jobs. Also, the decreasing costs of renewables (RDLs, RDHs) could further increase the penetration of renewables in the energy mix, leading to a higher number of jobs. Potentially achieving climate change targets in the future (e.g., 2.0 o C, 1.5 o C), more jobs could be created in Africa, meaning that more jobs can ensure political and societal stability. Also, how the jobs will be created and spread in Africa due to the diversi cation of the energy mix in each African nation may lead to social and economic disruptions, so strategic energy planning is essential.
National and governmental institutions and universities involved in capacity-building activities could bene t from this open-source study since the provided datasets could strengthen the capacity for developing others and extending existing energy systems models.

Overall Data-model assumptions
The modelling period spans between 2015-2040, providing results on an annual basis until 2030. The last ten years of the modelling period were added to prevent the "edge effects" as they are distorted by the model, considering that as the "end-of-time". Each year is divided into four seasons and two dayparts to capture the key features of the electricity demand load pattern. The year split is de ned at a continental level, so the countries cannot have a corresponding day split (e.g., day and night). The "daypart 1" is between 09:00 -18:00 (most of the commercial and public services are supposed to operate) and "daypart 2" between 18:00-09:00. "Season 1" corresponds to the period between March-May, "Season 2" between June-August, "Season 3" between September-November and "Season 4" between December-February. Country-speci c hourly electricity demand pro les were used to develop average pro les for electricity demand for the models' temporal split [12].
Country-speci c fossil fuel reserves and renewable potential are considered in the analysis. The list of power plants considered into the model is The real discount rate is 8% on a continental level [44]. The monetary unit used is the 2015 United States Dollars (USD). The USD gross domestic product (GDP) de ator from the World Bank Group is used [11] to adjust the fuel prices reported in different years to the base year (2015).

Energy demands
The energy demands considered in the model for the different fuels are: i) electricity, ii) biofuels and waste, iii) coal, iii) natural gas and iv) oil products demand. The electricity demands are categorized in: i) "all sectors" which include the aggregated nal electricity consumption and ii) "residential electricity demands of current un-electri ed settlements" distinguished in low and high depending on the electricity consumption per capita levels. The To estimate the fuel demands in "(all sectors)" the IBM SPSS Statistics tool was used to conduct a linear regression analysis using historical data (fuel consumption, GDP, population, urbanization) between 1990-2018 and project the values until 2040, using as methods criteria the probability of F. In case the regression analysis didn't t well between the dependent variable (fuel consumption) and the independent ones (GDP, population, urbanization) then the average growth rate of the power pool was used for that country or the average growth rate of the historical values of the respective country. The "residential electricity demands of current un-electri ed settlements" start from 2020 onwards to achieve universal access by 2030. They are categorized in "Low un-electri ed settlements residential demand" and "high un-electri ed settlements residential demand". The electricity demand The assumptions used for the energy demand projections are presented below. The "residential electricity demands of current un-electri ed settlements" for each African country and each scenario are presented below. The rest of the fuel demands are in Supplementary Material A.
Assumptions used for energy demand projections (Table 6-9)    Note: The electricity demand between the two scenarios is the same, but the generation mix is different as an outcome of the Global Electrification Platform (GEP) [32].

Policies
In the New Policies and Renewable Deployment scenarios, the objective is to examine different universal access scenarios by extrapolating the current energy situation into the future and providing energy transition pathways on a national, regional and African level.

Code availability
The OSeMOSYS code used to develop the model for Africa can be found in Zenodo-Github repository [68]. The current version of the TEMBA model takes up to 60 min. to solve using a commercial-grade solver (such as CPLEX or Gurobi) and requires up to 64 GB of memory.
Competing interests: The authors declare no competing interests. Overall electri cation mix (in PJ) in Africa among the scenarios the period 2015-2030.

Figure 5
Comparison of the total system costs in the energy sector in Africa among the scenarios (in BUSD) the period 2020-2030. Figure 6