Energy planning under uncertainty remains a critical challenge in developing economies, where data limitations and climate variability complicate long, term investment decisions. In Sierra Leone, these challenges are particularly acute given low electricity access 36% and heavy reliance on hydropower with strong seasonal variability. This study addresses the limitations of deterministic energy modelling, which typically assumes perfect foresight and produces single optimal pathways that may not remain valid under uncertain futures. To overcome this, we integrate an OSeMOSYS, based technoeconomic optimization model with a Robust Decision Making (RDM) framework to evaluate resilient energy strategies. The approach is applied to the case of the Bumbuna Hydroelectric Expansion, including the proposed upstream Yiben reservoir. Key contributions include the systematic validation of least, cost model outputs under deep uncertainty through hundreds of parameter variations, and the identification of critical uncertainty drivers using scenario discovery techniques (PRIM). Results show that scenarios incorporating the Yiben reservoir (Scenarios 6 and 7) achieve up to ,100% renewable electricity generation by 2030, reduce reliance on heavy fuel oil generation, and deliver lower system costs and Levelised Cost of Electricity compared to alternatives without storage expansion. Furthermore, these scenarios consistently demonstrate robustness across a wide range of futures, clustering in low, cost, high, renewable outcome spaces. The findings highlight hydropower storage as a key enabler of system reliability and decarbonization. From a policy perspective, the study underscores the importance of integrating uncertainty analysis into national energy planning and supports prioritising cascade hydropower development to enhance energy security

Laos’ updated Nationally Determined Contribution (NDC) prioritises decarbonisation and energy sustainability, yet progress is challenged by Laos’ reliance on imported petroleum for transport. This prompts government effort to promote biofuels and electric vehicles (EVs). This study applies the CLEWs (Climate, Land, Energy, and Water systems) framework within the Open-Source Energy Modelling System (OSeMOSYS) to assess interlinkages between renewable energy, transport, and land-use in Laos over the 2020-2050 period. Four national-strategy based scenarios were modelled: (i) baseline least-cost, (ii) 10% biofuel share, (iii) combined biofuel and EV targets (30% penetration among cars and 2-3 wheelers), and (iv) combined policy scenario with cropland constraint. Results shows that hydropower remains the dominant electricity source until 2050, with solar and waste-based generation emerging as a lucrative option after 2040. Biofuels are policy-driven rather than cost-optimal, while a 30% EV target for cars is economically feasible but requires policy intervention for 2-3 wheelers. Combining all policy measures achieves the lowest emissions but limits crop production. The study suggests supporting the target with land allocation for biofuel crops, biofuel feedstock diversification, and incentives for EV. As renewables become the least-cost power source, Laos could also consider coal phase-out by 2045. Overall, findings underscore the need for integrated policy planning within the CLEWs nexus to guide future national strategies.

In many low- and middle-income countries, energy planning processes remain centralised, technocratic, and disconnected from the lived realities of informal settlements, thereby institutionalising distributive, procedural, and recognition injustices before infrastructure decisions are implemented. This study conducts a global-to-local scoping review of 102 studies following PRISMA-ScR guidelines. Through thematic analysis, the review reveals that energy injustices in informal settlements are systemic outcomes of fragmented governance, exclusionary planning paradigms, and regulatory misalignment, which collectively curtail fundamental human capabilities for health, education, and livelihood. In response, the study develops and proposes the Energy Justice Triangle (EJT), an integrative framework linking three interdependent vertices: Capabilities & Well-being, Justice Principles, and Governance & Institutions. Applied through the lens of Lusaka’s informal settlements, the EJT demonstrates how opaque, grid-centric planning systems reproduce spatial and socio-economic exclusion. Crucially, the paper argues that dismantling these barriers requires a shift toward 'Glass Box' planning through a transparent approach that marries technical modelling standards, such as the U4RIA principles (Universal, Open, and Accessible), with the foundational tenets of Energy Justice. By utilizing the Climate Compatible Growth (CCG) 'flatpack' approach to institutionalize these open-source practices into local energy planning capacity, the study advances the EJT as a normative guide for justice-oriented energy system transformation. By connecting governance reform with modelling innovation, the study advances the EJT as a normative guide for justice-oriented energy system transformation. When planning assumptions, data inputs, and scenario choices are made more transparent, justice considerations can be more explicitly incorporated into planning processes.

Energy system modelling to explore development pathways has advanced significantly in recent decades as a critical tool and widely integrated methodology for energy planning procedures within Kenya. Modellers, analysts and academics have applied such tools to support international, national, regional and local organisations and policymakers within Kenya to make evidence‐informed energy policy and development decisions. As the body of energy modelling literature examining Kenyan applications expands and evolves, it is helpful to investigate research developments to guide current and future modelling research. This paper employs a systematic literature review, examining 79 studies, tracking the progress, challenges, gaps and trends related to energy modelling research related to Kenya over the last decade. We show that current energy modelling research with Kenyan applications typically operates outside of Kenya, largely conducted by European research institutions, with most articles consisting of no Kenyan co‐authorship. Six key research themes were identified across the examined articles: electrification, clean cooking, emissions, resources, energy transitions and grids. Additionally, three methodological themes of mixed methods, soft‐linking and policy implications were outlined. Three main limitations were defined across the existing literature, including data, model scope and looking beyond techno‐economic assessment. Furthermore, two key future research directions were identified as increasing geographical resolution and integration of social considerations.

Energy devolution in Kenya has had mixed and fragmented progress with over 15 counties yet to begin their County Energy Plan development. Yet, with the Ministry of Energy and Petroleum (MoEP) upcoming deadline for CEP production, and the following attempts to begin the first iteration of the integrated national energy plan (INEP), focus on county to national planning, data dialogues and modelling practices are of key critical current importance. Additionally, questions remain on how to translate the rich, bottom up, qualitative and narrative needs-based insights outlined within county energy plans, to an aggregate quantitative integrated national energy model. Here we apply mixed methodologies to energy modelling to overcome limitations of a strictly quantitative approach. We propose developing the existing national demand modelling socio-economic scenarios of baseline, ambitious, and reserved, to include a narrative driven county specific additional scenario, co-developed with 26 local county energy experts and decision-makers. Through the co-development of energy demand projections for integrated national energy planning using the case study of Taita Taveta, we present energy demand projections to assist policy makers at the regional, national, and international levels in energy modelling and policy pipelines to support both county energy planning and integrated national energy planning. Here we show how qualitative narratives can be integrated into modelling processes, alongside looking beyond technical and economic approaches to consider social, cultural, and behavioural factors. The produced energy demand projections provide a novel exploration of translating county energy plans into a standardised format for county-to-national integrated and inclusive energy modelling.

This briefing examines how artificial intelligence is entering national energy planning and why its use requires explicit governance, assurance and capacity-building. Its scope covers AI applications in forecasting, scenario exploration, optimisation acceleration, digital twins, generative-AI-enabled workflows, stress testing and decision support for increasingly complex, climate-exposed and cross-sectoral energy systems. It argues that AI can improve planning speed, probabilistic analysis and exploration of deep uncertainty, but should complement—not replace—physics-based models, institutional judgement or public accountability. The key finding is that the principal risks of AI in energy planning are institutional rather than purely technical. Poorly governed AI can create opaque model authority, vendor dependence, loss of public-sector capability, weak traceability, data-governance failures, and erosion of national sovereignty over models, data and decisions. The paper therefore proposes the Shared AI Governance for Energy Systems (SAGES) framework as a non-binding, principles-based assurance architecture for governments, regulators and international partners. SAGES centres on five commitments: AI as an empowering partner rather than a default solution; human and institutional accountability for final decisions; proportional, purposeful and testable trust; respect for national sovereignty and institutional maturity; and collective progress towards just energy futures. It recommends practical implementation through a Secretariat and Working Task Groups covering vendor engagement, demonstration tools, sovereignty and IP safeguards, capacity building and model literacy. The briefing concludes that responsible AI adoption requires traceability from data to decision, stress testing, independent challenge, public-sector capability, and international cooperation that supports national ownership without transferring decision rights away from states.

This report examines the prospective role of nuclear and Small Modular Reactors (SMRs) within the complex energy transitions of Low and Middle-Income Countries (LMICs), offering a replicable energy planning framework and informing potential UK Government support. It underscores that while variable renewable energy (VRE) sources are foundational for decarbonisation, their high penetration can introduce significant system-level costs, often uncaptured by simplistic Levelised Cost of Electricity (LCOE) metrics, necessitating complementary firm, dispatchable capacity. Modelling indicates that SMRs could integrate meaningfully into fossil-fuel heavy systems (with limited hydro and geothermal as so-called “firm” renewable capacity) from approximately 2037–2045. This integration would be driven by carbon constraints and reliability needs, potentially contributing over 20% of generation by 2055. Conversely, in hydropower- dominated systems (that provide “firm” renewable capacity), SMRs are projected to play a more limited role, typically less than 2% of generation, as existing hydro and increasingly cost-effective battery storage largely meet demand. The viability of SMRs is contingent upon specific “triggers”, notably their capital costs falling below approximately $6,000–7,500 MUSD/GWe, alongside carbon prices exceeding $73–78 USD/ton, respectively. (At lower carbon prices, fossil-fuel alternatives provide the required “firm” capacity). Note that these numbers will change over time (as the energy system is dynamic) and will require updating annually. Construction delays significantly escalate total system costs, primarily by necessitating reliance on more expensive, often fossil-fuel-based, backup generation and incurring higher capital costs over time.

The intersection of energy systems optimization modelling and policy evaluation remains critically underdeveloped in developing-country contexts, where data discrepancies, institutional capacity constraints, and the rapid growth of decentralized solar photovoltaics (PV) routinely confound government energy planning. This paper presents a five-step integrated methodology combining open-source energy system optimization modelling with systematic policy evaluation and government data auditing, applied comprehensively to Pakistan's energy system with 2024 as the base year and projections to 2050. The five steps comprise: (1) Multi-Source Data Discrepancy Diagnosis; (2) OSeMOSYS Model Construction and Sector-Coupled Calibration; (3) Systematic Testing and Audit of Government-Published Energy Reports; (4) Discrepancy Reporting and Source-of-Error Attribution; and (5) Scenario-Based Policy Calibration and Reform Pathways. Application to Pakistan reveals structural discrepancies averaging 18-34% between government-published energy statistics, quantifies a previously unreported 12.4 GW gap between officially counted and operationally active generation capacity, and diagnoses the systematic exclusion of 3-8 GW of decentralized rooftop solar from national energy balances. Scenario analysis projects that an ambitious transition pathway could achieve 72% renewable electricity by 2035 and deliver net system cost savings of USD 340 billion over 2025-2050 relative to the Business-as-Usual trajectory when fossil fuel import savings are included. The five-step methodology is designed for replication across developing nations facing comparable data and institutional challenges.

Linear and mixed-integer optimisation models are widely used across economics and engineering to study resource allocation, infrastructure planning, and energy-system transitions. Algebraic modelling languages such as GNU MathProg (GMPL), AMPL, and GAMS let researchers write these models close to their mathematical form, keeping them transparent and reviewable even without extensive coding experience. However, as models grow in scale, translating algebraic formulations into solver-ready sparse matrices becomes a major computational bottleneck. This paper introduces MOSOX, a Rust-based command-line tool and library that compiles a targeted subset of GMPL model and data files into sparse matrices, covering the constructs required by the OSeMOSYS energy-system model family. It expands sets, parameters, variables, objectives, and constraints into matrices, exports them in standard MPS format, and can solve models directly via the HiGHS solver. On OSeMOSYS benchmarks, MOSOX compiles matrices up to 6.5 times faster than GLPK's glpsol while also reducing peak memory use on the largest tested model. By combining fast, low-memory compilation with the readability of GMPL and solver-independent output, MOSOX - developed within the Climate Compatible Growth Program - supports reproducible, auditable, and automatable optimisation workflows for large-scale energy-system modelling.

The rapid growth of Vietnam’s transport sector presents challenges for sustainable energy and transport planning, particularly due to rising fuel consumption and associated carbon emissions. To help address these challenges, this paper presents a comprehensive dataset designed to support the modelling of future transport demand, energy use, and CO2 emissions in Vietnam. The dataset covers population, GDP, passenger and freight activity, vehicle stock, energy intensity, load capacity, and CO2 factors across nine transport modes: motorcycles, cars, buses, light-duty vehicles, heavy-duty vehicles, rail, inland waterways, maritime, and aviation. These are further disaggregated into nine fuel types: petrol, diesel, compressed natural gas, electricity, biofuel, fuel oil, hydrogen, ammonia, and jet fuel. Data were compiled from national statistics, government reports, online databases, academic journals, and media sources. Structured for use with open-source modelling tools, the dataset supports analyses of transport demand and carbon accounting, offering researchers, policymakers, and consultants a resource to evaluate long-term decarbonisation pathways and inform evidence-based policymaking.

Transport is often overlooked in energy planning, and this is the case in Vietnam. Yet the transport sector is rapidly growing and is therefore a vital part of the country’s energy transition. This study adopts an evidence-based policymaking approach, using the Model for Analysis of Energy Demand (MAED) to project transport activity, energy demand, and emissions to 2050 in support of Vietnam’s net-zero commitment. It provides the first comprehensive, scenario-based assessment of Vietnam’s transport sector using national data and policy assumptions, offering a quantitative foundation for integrating transport into energy planning and facilitating informed stakeholder dialogue. By modelling scenarios that combine net-zero pathways with avoid-shift-improve (ASI) measures and conducting sensitivity analyses on population and GDP projections, the study develops new insights for Vietnam’s transport sector. Results show that net-zero can reduce energy demand by nearly 60% in passenger transport and 19% in freight, compared to 2024 values. Additionally, ASI measures can further lower these values by 19% and 41%, respectively. In terms of transport activity, ASI measures can reduce passenger activity by 50 billion passenger-km and freight by 115 billion tonne-km in 2050, compared to a reference case. Uncertainty in population growth has a minor impact on transport activity and emissions, due to the narrow range of projected population outcomes. In contrast, uncertainty in GDP growth has a much stronger influence, reflecting the wider range of economic projections. Modelling suggests that this variation significantly affects freight activity levels, cumulative emissions, and the timing of the peak emissions.

Ghana’s electricity sector faces persistent challenges of transmission losses, regional inequities, and limited renewable integration despite achieving a high national electricity access rate of 89%. While national-level energy models have provided valuable insights for Ghana, spatial bottlenecks and urban-rural disparities remain insufficiently examined. This study develops a multi-regional optimisation energy system model that segments Ghana into four clusters (Coastal, Central, Northern, and Northern Remote) based on access rates, grid infrastructure, and existing and potential power generation.The model incorporates inter-regional transmission and distribution losses, government renewable targets, and urban-rural demand differences to simulate four scenarios from 2015 to 2070: Business-as-Usual (BAU), Transmission and Distribution Loss Target (TDL), Government Target (GT), and a combined pathway (GT+T). Results show that long-term power system planning dominated by conventional generation leads to high total system costs (USD 95 billion by 2070), largely driven by fossil fuel operating expenditures. In contrast, the combined implementation of grid efficiency improvements and renewable integration could reduce total system costs by up to 39%.Under this integrated pathway, installed capacity transitions from a fossil-dominated system to a diversified mix led by solar deployment in rural areas. Transmission system analysis shows that pathways without efficiency interventions place significant stress on central transmission corridors, whereas efficiency-focused strategies promote decentralised transmission flows. Increased renewable deployment diversifies supply sources, reduces reliance on long-distance transmission, and improves spatial equity in electricity access.This study shows that integrating grid efficiency with renewable deployment provides the least-cost and most resilient pathway for Ghana’s long-term power system development.

Energy system modelling for sustainable development has advanced significantly as a critical tool for designing cost-effective energy transitions. Modellers and analysts have used these tools to support international organizations and policymakers in crafting and making decisions about energy policy. Open-source frameworks have been instrumental in this progress, enhancing stakeholder engagement, transparency, and public acceptance. Among these, the Open-Source Energy Modelling System (OSeMOSYS) stands out as a key example, widely applied in energy transition and planning studies. As the body of OSeMOSYS literature rapidly expands, it is essential to track research advancements to guide both current and future modellers. This paper presents a systematic literature review, exploring the applications, developments, and research trends related to OSeMOSYS over the last 10 years. The findings highlight a significant growth in OSeMOSYS-based research, with an annual increase of approximately 28 %, and most applications occurring in Africa and Latin America, though largely conducted by European institutions. Six key application areas were identified, such as capacity expansion planning in the power sector, transport sector planning, and sector coupling opportunities. Nine categories of complementary methods commonly integrated with OSeMOSYS were also categorized, including power sector performance, stakeholder engagement, and geospatial assessments. A thorough review of code enhancements demonstrates the framework's adaptability to various fields, such as flexibility assessment, hydropower systems, and storage modelling. Furthermore, seven key future research directions were identified: operational feasibility, uncertainty evaluation, temporal and spatial resolutions, technological detail, storage modelling, and macroeconomic impacts. This paper aims to serve as a comprehensive resource for modellers, analysts, and users, offering insights into research questions, complementary methods, available code enhancements, and potential future directions for the use of OSeMOSYS.

Despite abundant renewable resources, Indonesia’s heavy reliance on coal for electricity generation results in high emissions and an electricity surplus, which could hinder the integration of renewable energy. To that end, this study involves a novel framework application to softlink OSeMOSYS and FlexTool, evaluating the potential of energy storage to enhance power system flexibility and facilitate a higher share of renewables in Indonesia’s future power sector. Six scenarios covering 2015 to 2070 were modelled using a soft-linking approach combining OSeMOSYS for capacity planning with FlexTool for operational analysis. These scenarios include Business as Usual, Nationally Determined Contributions, and Just Energy Transition Partnership, each modelled with and without energy storage. Results reveal that achieving Indonesia’s climate targets will require significant reliance on solar energy, which could lead to substantial curtailment and load losses. Incorporating energy storage could mitigate these issues, thus enabling a renewable energy share of 64.45% by 2060 while reducing total system costs and CO2 emissions. They also highlight that energy storage only becomes cost-effective when renewable energy shares are high, with hydrogen storage likely to surpass pumped hydro storage after 2050. Based on these findings, three key policy recommendations are proposed: implementing a moratorium on new coal-fired power plants, establishing a dedicated financing facility for energy storage, and developing a clear energy storage roadmap with specific targets.

Nigeria has one of the greatest electricity deficits globally and, even in areas connected to the central grid, struggles to provide reliable power across the nation. Frequent system collapses and widespread reliance on diesel generation present a burden for Nigerian households and the economy as a whole. One causal factor in these collapses is capacity inadequacy owing to reduced plant availability as plants are frequently non-operational due to maintenance or other management issues. Using a combination of OSeMOSYS and FlexTool modelling, this study shows the significant burden that persistently unavailable plants present for decarbonisation of the Nigerian energy system. Modelling which includes reliability improvements nearly halves total system costs and emissions versus business-as-usual. Further, Nigeria is unable to meet its 2021 Nationally Determined Contributions (NDCs) without such improvements, indicating that increasing plant availability and reducing diesel generator use must be prioritized in policy to support national implementation of these targets.

As global environmental challenges increase, the need for integrated energy modelling to facilitate data-driven decision-making in energy policy and finance is critical. However, most existing integrated frameworks are limited in their applicability, granularity, and accessibility, risking the exclusion of developing countries from the global energy transition. Social dimensions are also often insufficiently addressed, and financial planning is not integrated, leaving gaps between technical analysis, social considerations, and actionable investment pathways. To address this, the article presents the Integrated Model for Policy, Actions and Collaborative Climate Transitions (IMPACCT), a new comprehensive framework that soft-links seven significant open-source tools—MAED; OnSSET; OSeMOSYS, including CLEWs and SIBs; FlexTool; PathCalc; MINFin; and FINPLAN—for the first time. IMPACCT estimates energy demand from electrified and unelectrified populations and calibrates the least-cost capacity mix to meet demand while accounting for land availability, water use, carbon emissions, and social factors. The capacity mix is further refined to ensure power system flexibility, and the technical outputs are visualised in an engaging interface. Financial strategies at national and utility levels complete the framework, supporting the practical realisation of the technical plans. By outlining a new process with open-source, user-friendly interfaces, this paper increases accessibility and ease of use, supports capacity building in developing countries, and facilitates collaboration across institutions and disciplines. It delivers a significant leap in energy modelling, social inclusion, and financial planning, advancing a more integrated approach to sustainable development. Overall, IMPACCT enables more transparent and collaborative decision-making, accelerating financial mobilisation for a just energy transition.

This paper provides an analysis of water heating within South Africa's residential energy demand. Using MAED, this work models four scenarios to examine the impact of varying degrees of governmental intervention and policy implementation, specifically SANS10400-XA2, on water heating energy consumption from 2020 to 2040 across income groups. Results show that solar thermal use, with effective support, could reduce low-to-middle income households' reliance on carbon-intensive fuels by similar to 30 %. Findings also indicate that strong governmental support is essential to reach 16 % of electricity savings across income groups. In the reference scenario, high-income households rely entirely on electricity, while LMI households use 31 % electricity and utilize fossil fuels. Under basic support, LMI electricity use drops by 50 %, but fossil fuels increase relative to BAU. Moderate support continues LMI reliance on fossil fuels due to infrastructure limitations. High support, causes both electricity and fossil fuel use to decrease for LMI. In pursuit of equitable access to sustainable energy technologies, the level of aid determines whether LMI households utilize 0 %, 9 %, or 35 % of solar thermal energy to meet their final water heating needs. To realize the potential energy efficiency benefits and emissions reductions, targeted policies and infrastructure improvements are crucial. If less wealthy households are not considered, they may struggle to transition to cleaner energy solutions and remain dependent on polluting fuels, leading to negative health and environmental implications. Results from this study indicate that supporting solar thermal technologies and improving water access can enhance living standards for vulnerable populations while simultaneously mitigating climate change effects.