Different Visions 2030–2040 mark a transformative shift from hydrocarbon reliance towards a diversified, sustainable energy future, driven by the need for enhanced energy security and economic sustainability. For instance, Oman vison defined an ambitious target: reducing the oil sector’s GDP contribution from 39% (2017) to 8.4% by 2040. This also aims to increase renewable energy consumption to 35–39% of total energy use, with an interim goal of 20% by 20301. A critical component is developing robust solar and wind infrastructure and significantly scaling clean hydrogen production from 32,500 metric tons to 1 million annually by 2030, and further to 8.5 million by 20502. This entails a complete reimagining of Oman’s energy infrastructure, including smart grids, energy storage, and modern distribution networks with a strong commitment to environmental protection and green technologies, positioning Oman as a regional leader in sustainable energy1.
To achieve these goals, the Oman Electricity Transmission Company (OETC) is actively integrating large-scale renewable projects, such as the 500 MW Ibri II Solar IPP, with further solar expansions in Manah (1000 MW), Ibri III (500 MW), and A Kamil (500 MW), alongside wind projects in JBB (100 MW), Duqm (200 MW), Ras Madrakah (200 MW), and Dhofar II (100 MW) planned for connection to the Main Interconnected System (MIS) by 20303. While this significant renewable penetration supports sustainability objectives, it introduces substantial technical challenges. Primary concerns include the inherent intermittency of solar and wind power generation and the reduction in system inertia due to the displacement of conventional synchronous generators, both of which critically affect grid stability and operational reliability3.
Energy Storage Systems (ESSs) present crucial opportunities to address these challenges, enhancing renewable energy integration in Oman, lowering operational costs, and reducing fossil fuel consumption by managing intermittency and stabilizing the grid4,5. Current research highlights various ESS technologies. These technologies vary in their applicability: lithium-ion batteries offer high efficiency but face scalability and environmental issues; Pumped Hydro Energy Storage (PHES) provides large-scale storage but is geographically limited; Compressed Air Energy Storage (CAES) also presents a viable option for large-capacity, long-duration storage, contingent on geographical suitability; flywheels excel in rapid response but lack long-term capacity; and thermal storage suits specific applications6. Emerging innovations include solid-state batteries, hydrogen for long-duration storage, and Artificial Intelligence/Machine Learning techniques for optimized management6.
Literature review
Previous studies have explored optimal sizing and location of ESS with different objectives.
Authors of7 introduced an advanced distributed planning framework for integrated electricity and natural gas systems (IEGS), explicitly incorporating Regional Integrated Energy Systems (RIESs) and addressing the realistic scenario of multi-stakeholder ownership across gas networks (GN), electricity networks (EN), and RIESs. Departing from conventional centralized joint planning approaches, the proposed model leveraged the Alternating Direction Method of Multipliers (ADMM) to enable decentralized coordination, allowing each agent to independently optimize their objectives while ensuring convergence on shared operational variables such as transmission flows. Validated through a case study using the IEEE 24-bus and Belgian 20-node systems, the model demonstrated improved adaptability and responsiveness to localized variations, along with effective negotiation-based planning among stakeholders. Sensitivity analyses further assessed the influence of penalty parameters and convergence thresholds on algorithm performance. However, the model faced several limitations, including convergence challenges due to the non-convexity of mixed-integer programming (MIP), high computational burden compared to centralized models, and an observed increase of 4.86% in overall planning costs due to distributed inefficiencies. Additionally, the framework relied on simplified assumptions, such as static market conditions and uniformly cooperative stakeholder behavior, which may not fully capture the complexity of competitive or deregulated energy environments. These limitations suggested the need for future research to enhance convergence robustness, integrate economic uncertainty, and extended the model’s applicability to larger and more dynamic multi-energy systems. Authors of8 presented a comprehensive framework for multi-objective optimization of an interactive buildings-vehicles energy sharing network that leveraged grid-responsive strategies, diverse renewable integrations, and hybrid storage systems. By employing the advanced Pareto archive NSGA-II algorithm, the study effectively minimized equivalent CO2 emissions and import costs while maximizing energy flexibility. The proposed system achieved notable performance improvements over conventional isolated configurations, including a 7.5% reduction in emissions (from 147.4 to 136.4 kg/m2·a), an 8.5% decrease in grid import costs (from 212.7 to 194.6 HK$/m2·a), and substantial enhancements in energy flexibility metrics, such as shifting 52.48% of surplus renewable energy compared to 33.6% in the baseline scenario. A key innovation was the integration of quantifiable flexibility indicators like RSR and GSR, offering valuable insight into system responsiveness and storage effectiveness under dynamic conditions. Nevertheless, limitations remained in the deterministic modeling of vehicle behavior, the omission of battery production and degradation impacts, and a lack of in-depth economic feasibility analysis regarding capital investment and ROI.
Complementing this work, authors of9 offered a multidisciplinary perspective on urban energy systems by emphasizing the cross-sectoral integration of buildings, transportation, and power grids to support the development of sustainable and climate-resilient smart cities. The study underscored the importance of technologies such as bidirectional EV charging, hydrogen refueling stations, and building prosumers in enhancing energy efficiency and system resilience. By incorporating climate modeling, artificial intelligence, energy resilience indicators, and techno-economic-environmental evaluations, the authors proposed a holistic roadmap for city-scale planning and energy management. However, this analysis was primarily conceptual and lacked empirical validation, detailed methodological descriptions of optimization or AI frameworks, and region-specific economic assessments. Furthermore, while societal and regulatory barriers were acknowledged, they were not explored in sufficient depth.
In10, another study advanced a co-simulation framework combining MATLAB and TRNSYS to enable multi-objective optimization and component sizing of a solid oxide fuel cell-based combined cooling, heating, and power (SOFC-CCHP) system for green buildings. Utilizing a semi-empirical surrogate model of the SOFC, the study optimized the battery, electrolyzer, and SOFC subsystems to simultaneously enhance energy efficiency and reduce annual costs while accounting for performance degradation. Results indicated that increasing the size of the electrolyzer and SOFC improved energy efficiency by 13.64% and 2.19%, respectively, with annual costs ranging between $67,230 and $73,250. A novel application of ANOVA and Pareto-based decision-making added robustness to the design process. Despite its methodological strengths, the study relied on idealized gas assumptions, excluded water crossover effects, and did not account for uncertainties in input data or operational variability, limiting its applicability in real-world dynamic environments.
In a related context, authors of11 conducted an in-depth study on sustainable interactive energy sharing districts that incorporated electrochemical battery storage. This work explored multi-directional energy exchanges among PV systems, wind turbines, buildings, EVs, and microgrids, and integrated both semi-empirical and machine learning models for battery lifetime estimation. The study proposed deterministic and stochastic control frameworks, emphasizing stakeholder-based optimization approaches and employing multi-criteria decision-making tools such as Shannon entropy, Euclidean distance, and fuzzy logic. It provided valuable insights into system design, battery sizing, carbon reduction, and cost-effectiveness, with a strong focus on improving energy flexibility and resilience. Nonetheless, the research was largely simulation-based and lacked empirical validation. Moreover, while battery degradation was thoroughly addressed, economic trade-offs and real-time data limitations in machine learning applications were not fully examined.
Building upon the theme of battery longevity, another study12 proposed a hierarchical control strategy that integrated thermal and electrical storage to reduce battery cycling ageing and improve energy flexibility. By applying depth-of-discharge (DoD)-based control and utilizing surplus renewable energy for recharging thermal systems such as air conditioning and hot water tanks, the strategy significantly mitigated battery degradation and extended system life. Results demonstrated an increase in relative battery capacity from 94.21 to 95.46% annually and system cost reductions of up to 19.33% for lead-acid batteries. However, this work remained theoretical, lacking real-world implementation or sensitivity analyses that would consider market conditions, infrastructure constraints, or policy incentives.
In13, a climate-adaptive design framework for Zero-Energy Buildings (ZEBs) was introduced, proposing an integrative “kWp-kWh-m2” approach that coupled PV generation with optimally sized battery storage across various Chinese climate zones. The study employed a life-cycle analysis model incorporating economic, environmental, and policy parameters under different Representative Concentration Pathway (RCP) scenarios. It revealed that battery integration significantly improved renewable penetration, with decarbonization potential reaching up to 70%, especially in centralized systems. The authors provided strategic battery sizing recommendations—ranging from 3.75–4 kWh/kWp for centralized systems to 1.75–2.75 kWh/kWp for distributed setups, while highlighting the influence of electricity pricing, carbon intensity, and climate variations. Despite its comprehensive modeling and policy relevance, the study lacked empirical case studies and broader regional comparisons, which limited its global generalizability.
Furthering the agenda for sustainable electrification, authors of14 proposed a lifecycle carbon neutrality strategy grounded in an electricity-driven circular economy. The study explored five scenarios integrating renewable energy, vehicle-to-everything (V2X) interaction, and EV battery cascade utilization. Among these, the synergistic scenario combining V2X and second-life battery reuse proved most effective, achieving negative carbon intensity and strong economic viability across varied climate zones. Moreover, the analysis extended globally, supported by policy instruments such as production and investment tax credits. However, the study did not fully consider the transitional dynamics of power grid decarbonization or the integration of large-scale centralized renewables like offshore wind, suggesting the need for further research into broader energy contexts.
In15, a novel lifecycle design methodology for renewable-battery-consumer energy systems was introduced, focusing on battery capacity optimization through the “M-value” (matching degree) method. This approach offered a significant improvement over the conventional “U-value” (uniformity) method by aligning PV generation more effectively with building energy demand. The framework accounted for performance degradation, carbon emissions, and economic outcomes, and established a scalable area-kWp-kWh design guideline. Simulation results across different building types and climate zones in China demonstrated lower lifecycle carbon intensity and improved net present value (NPV). Nevertheless, the scope was limited to solar PV and Li-ion batteries, excluding other renewable technologies and the potential contributions of electric vehicles to energy balancing.
Other researchers addressed the optimal sizing and location of photovoltaic generation systems (PVGS) and battery energy storage systems (BESS) to enhance power loss reduction, voltage profile improvement, and voltage unbalance in an unbalanced distribution system. It employs a refined genetic algorithm for optimization, considering investment costs and operational constraints. Other researchers presented a novel optimization framework for the optimal siting and sizing of distributed renewable generation and energy storage systems. It utilized four distinct load models and three metaheuristic approaches, with the Elephant Herding Optimization (EHO) emerging as the best performer for voltage stability and real power loss reduction. The findings suggest that the Modified Ant Lion Optimization (ALO) is most effective under ideal conditions, particularly when wind and irradiance percentages are 60% or greater16,17. While authors of18 presented a deterministic planning model for the optimal sizing and location of battery energy storage systems (BESS) to support renewable energy sources. It utilized a mixed-integer nonlinear programming approach to minimize investment and operational costs while satisfying system constraints. The model was tested on a modified IEEE 14-bus system, indicating its applicability for future scenarios. In19, researchers discussed a method for optimal sitting and sizing energy storage (ES) systems to support renewable energy integration, focusing on minimizing the expected operating and investment costs. It analyzes various parameters, including the maximum number of storage locations and renewable generation capacity. It was evaluated using a realistic model of the Western Electricity Coordinating Council (WECC) interconnection, consisting of 240 buses and 448 lines. While in20, Authors established an optimization method for determining the optimal sizing and location of energy storage plants to support renewable energy. This approach involved pre-selecting locations based on load, renewable energy capacity, and distances to neighboring nodes. The multi-objective artificial bee colony algorithm was employed to maximize peak-shaving profit, minimize investment costs, and reduce active network loss. This approach effectively enhanced renewable energy utilization by optimizing both the location and capacity of energy storage facilities. Authors of21 focused on determining the optimal capacity and location of energy storage systems (ESS) to support grid stability in the presence of renewable energy sources (RESs). It emphasized that ESS can provide virtual inertia, mitigating frequency fluctuations caused by the low inertial characteristics of RESs. The developed model considered power grid constraints, including voltage, angle, and line capacity limits, and was evaluated using the New England IEEE 39-bus system to ensure cost-effectiveness and stability. While in22, researchers found that Energy storage systems can significantly improve frequency response in power systems with high renewable penetration, particularly in scenarios with low inertia due to synchronous generator displacement. Also found that the optimal placement of storage could mitigate frequency disturbances, enhancing overall system reliability during transients. Authors of23 deduced that Storage technologies could absorb surplus renewable energy, reducing the need for additional capacity by 24–44% under favorable economic conditions. While authors of24 discussed that the integration of storage not only lowers greenhouse gas emissions but also decreases overall energy costs by optimizing the use of available renewable resources. They deduced that geographic and seasonal variations in renewable energy availability necessitate tailored storage solutions to meet local demand effectively. In25, researchers concluded that although the benefits of energy storage in supporting renewable energy integration are evident, significant challenges persist, particularly the high costs of certain storage technologies and the need for supportive policies to enable widespread adoption.
Hydrogen literature
Green hydrogen, produced through water electrolysis using renewable energy, plays a vital role in decarbonizing energy and transport sectors by enabling flexible integration of intermittent renewables. Hydrogen fuel cell vehicles (HFCVs) are central to this integration, supported by technological advances and expanding refueling infrastructure. Despite this progress, significant technological, economic, and regulatory barriers hinder the growth of the hydrogen economy. This review highlights current achievements and ongoing challenges in deploying HFCVs and scaling green hydrogen production, and proposes a roadmap for future research, policy, and investment to advance hydrogen as a key component of sustainable energy and transport systems26. This section presents a brief critical review of hydrogen optimization with renewable energy integration, including comments on the limitations of existing studies.
Authors of27 developed an integrated renewable energy–refinery hydrogen management system that combined energy storage and direct utilization to enhance hydrogen utilization efficiency in refinery operations. They introduced a novel superstructure that encompassed green hydrogen production via water electrolysis and hydrogen compression powered by wind energy, supported by underground hydrogen storage (UHS) units. A nonconvex MINLP model was formulated to optimize the system’s design, incorporating mass and energy balances, compressor configurations, and cost functions. Through a hierarchical algorithm, the study demonstrated that the proposed approach reduced total annual costs by 21.21% under abandoned wind pricing and 13.49% under normal pricing, with further reductions projected for 2035 and 2050. It was also found that using wind power for hydrogen compression yielded more favorable economics than standalone green hydrogen production, especially as wind electricity costs decreased. Overall, the study emphasized the feasibility of directly integrating fluctuating renewable energy into refinery hydrogen systems while minimizing energy conversion losses. Despite its innovative contributions, the research relied on idealized assumptions regarding system behavior, constant hydrogen purity, and geographic availability of UHS units. Moreover, the lack of empirical validation and sensitivity analysis under broader economic and policy uncertainties limited the real-world applicability of the proposed model.
In28, authors conducted a comprehensive methodological review of renewable hydrogen system modeling and optimization, with a particular focus on integrating machine learning to enhance system performance and decision-making. The study critically evaluated limitations in traditional modeling approaches and highlighted the need for dynamic representations of solar PV, electrolyzer, and fuel cell performance. It demonstrated that machine learning techniques—such as neural networks and support vector machines—significantly improved forecasting accuracy, component behavior modeling, and load profile development, particularly in data-scarce regions like the Global South. Additionally, the authors proposed a novel metric—Levelized Value Addition (LVA)—to incorporate socio-economic considerations into hydrogen system evaluations. An integrated, evidence-based multi-criteria decision-making (MCDM) framework combining both multi-objective and multi-attribute methods was also introduced. By identifying best practices in optimization, load profile creation, and economic assumption modeling, the review offered a structured pathway toward more realistic, data-driven, and socially inclusive system designs. However, the work remained largely conceptual, lacking empirical case studies to validate the proposed frameworks. Furthermore, although machine learning integration was extensively discussed, practical challenges such as data scarcity, interpretability, and computational complexity were not fully resolved. The application of socio-economic metrics like LVA also lacked real-world demonstrations across different contexts, limiting their immediate implementation potential.
Similarly, authors of29 proposed a two-stage distributed robust optimization model for scheduling a hydrogen production system based on renewable energy sources (H2-RES), taking into account the uncertainties in solar and wind power generation as well as the flexibility of electric loads. The first stage optimized the capacity configuration of key components—electrolyzers, hydrogen compressors, and hydrogen storage tanks, with the aim of minimizing investment costs. The second stage focused on dynamic scheduling of system operation under worst-case scenarios using the Column and Constraint Generation (C&CG) algorithm to minimize operating costs. The model incorporated flexible load modeling, including transferable and reducible loads, and was applied to a real hydrogen project in Zhangjiakou, where it demonstrated improved economic performance and better alignment with actual system configurations. The study effectively highlighted the importance of flexible demand response and robust planning in enhancing the reliability and cost-effectiveness of hydrogen systems integrated with renewables. Nevertheless, it was constrained by its heavy reliance on simulation and the absence of operational data for validation. Behavioral and regulatory factors influencing real-world user participation in flexible load schemes were not deeply addressed, and the worst-case focus of the uncertainty treatment potentially oversimplified the stochastic nature of renewable fluctuations.
In a broader system context, authors of30 developed a comprehensive modeling and optimization framework for a large-scale renewables-based hydrogen system that encompassed the entire production–storage–transportation–utilization chain (PSTUH2S). The study considered multiple hydrogen production sources—renewables, fossil fuels, and grid power—as well as various storage options, transport methods, and sectoral demands spanning industry, power, construction, transport, and aerospace. A multi-objective nonlinear optimization model was formulated to simultaneously maximize economic benefit, increase renewable energy consumption, and minimize carbon emissions. Solved using a hybrid approach combining nonlinear programming, the CPLEX solver, and piecewise time series simulation, the model was applied in northwest China for the period 2025–2035. The results demonstrated significant economic and environmental gains, including cost savings up to USD 865.87 million and reductions in renewable energy curtailment and emissions. Sectoral analysis revealed industrial hydrogen as the dominant demand, though rapid growth potential was observed in power and transportation sectors. Despite its extensive modeling, the study depended on deterministic forecasts and assumed fixed operational parameters—such as hydrogen blending ratios and ideal electrolyzer behavior—potentially limiting its responsiveness to real-world variability. Additionally, socio-institutional considerations such as regulatory frameworks, market dynamics, and infrastructure readiness were not sufficiently explored.
Lastly, authors of31 conducted a detailed parametric and optimization study on solar/wind-based hybrid renewable hydrogen production in Kuqa, China. They explored six scenarios under both grid-connected and off-grid configurations using HOMER and TRNSYS simulations to evaluate performance under varying energy conditions. The PV/WT (G5) system was identified as the most effective grid-connected configuration, while the PV/WT/FC (S6) system was optimal for off-grid applications. Both systems demonstrated strong technical and economic performance, with low energy costs, reduced capacity shortages, and improved hydrogen utilization. The study further optimized system configurations by refining load profiles and adjusting component sizes, leading to reduced grid dependency in G5 and lower excess energy rates in S6. These findings underscored the value of careful component-load matching to enhance system efficiency in arid, resource-abundant settings. However, the research remained confined to simulation and lacked empirical testing. Operational assumptions, such as consistent component behavior and static user demands—may not hold in real-world deployments. Furthermore, the study’s narrow focus on techno-economic metrics excluded policy, regulatory, and environmental dimensions necessary for scalable and sustainable deployment.
Research gap
While existing literature provides valuable insights into the optimal sizing and siting of energy storage systems (ESS) to support renewable energy integration, several limitations persist that highlight the need for further research. Many studies rely on deterministic or idealized scenarios, often neglecting the impact of uncertainties in renewable generation, demand variability, and market conditions. Additionally, most models assume static network conditions and do not fully capture the dynamic behavior of power systems under high renewable penetration. The optimization methods employed, though diverse, are often tested on small or simplified test systems, limiting their scalability and applicability to real-world, large-scale grids. Furthermore, several approaches overlook the coordination among multiple stakeholders, regulatory constraints, and economic incentives required for practical implementation. High investment costs and lack of detailed techno-economic analysis for different storage technologies also remain underexplored. These gaps underscore the need for more comprehensive, robust, and scalable models that incorporate uncertainty, system dynamics, and policy frameworks to support effective ESS planning in future power systems.
This paper presents a novel multi-objective optimization framework designed specifically to support Oman’s Vision 2040 energy transition goals. The study systematically evaluates how various energy storage systems (ESS), including pumped hydro storage, compressed air energy storage, batteries, and hybrid configurations, perform across different renewable energy investment scenarios. The framework simultaneously optimizes three critical objectives: maximizing renewable energy integration, minimizing carbon emissions, and enabling green hydrogen production from surplus energy.
Unlike previous approaches, the model incorporates realistic storage charge/discharge efficiency losses and explicitly accounts for transmission line flow constraints as a proxy for N-1 security requirements. This allows for determination of maximum renewable penetration levels that maintain system reliability while identifying the most suitable storage technologies for Oman’s specific needs. Moreover, this study includes a detailed techno-economic assessment of each storage configuration, which is often overlooked in the literature. By validating our framework on the IEEE 9-bus test system, we demonstrate its robustness and applicability, offering practical insights for grid planners and policymakers. Furthermore, the scalability of the proposed optimization framework was evaluated using IEEE 30, IEEE 39, IEEE 57, and IEEE 118-bus systems, demonstrating its computational efficiency and practical applicability on larger grid models.
The rest of this paper is organized as follows: The paper’s contributions are provided in Section “Contributions”. The model formulation is described in Section “Model formulation”. The results and discussion are covered in Section “Results and discussion”. The Scalability is proved in Section “Computational scalability”. Finally, Section “Conclusion” includes conclusions and suggested future studies.
