This paper proposes a forecasting-driven uncertainty management framework for a hydrogen-enabled Multi-Energy Microgrid (MEMG) integrating a 14-bus IEEE power network, a District Heating Network (DHN), and a Gas Distribution System (GDS). The framework incorporates renewable energy resources, Combined Heat and Power (CHP) unit, Power-to-Hydrogen-and-Heat (P2HH) technology, and Plug-in Electric Vehicles (PEVs) to simultaneously satisfy electricity, heat, and gas demands while enhancing operational flexibility. Unlike conventional IGDT-based energy management approaches that rely on predefined nominal demand profiles, the proposed framework integrates CNN–LSTM demand forecasting with IGDT-based optimization by using the forecasted demand as the nominal operating profile for uncertainty-aware scheduling. A two-stage methodology is developed. First, a hybrid CNN–LSTM model is employed to forecast electricity, heat, and gas demands. Subsequently, Information Gap Decision Theory (IGDT) is used to manage forecasting-driven uncertainty through risk-averse and risk-seeking decision-making strategies. The forecasting model is implemented in Python, while the operational optimization problem is formulated as a Mixed-Integer Nonlinear Programming (MINLP) model and solved in GAMS. Simulation results demonstrate satisfactory forecasting performance, with sMAPE values below 15% for all energy carriers. The results further show that, compared with the risk-neutral strategy, the risk-averse strategy increases total cost, electricity transaction cost, and emission cost by 28%, 51%, and 46%, respectively, in exchange for enhanced robustness against uncertainty. Conversely, the risk-seeking strategy reduces the total cost and emission cost by 40% and 85%, respectively, while the electricity transaction balance shifts from a net cost to a net revenue due to increased electricity exports under favorable operating conditions. These findings demonstrate the effectiveness of the proposed forecasting-driven IGDT framework in balancing economic performance and operational robustness in hydrogen-enabled multi-energy microgrids under uncertainty.
Biomass district heating plants are central to renewable heat supply in rural Austria, but growing competition for limited biomass resources requires more efficient use of existing infrastructure. This study presents a techno-economic assessment of retrofit measures for the Austrian biomass district heating sector. Plants are grouped into representative clusters and evaluated for combustion optimization, passive and active flue gas condensation, and waste-heat heat pump integration under varying energy price scenarios. Results are projected into sector-wide rollout pathways to 2050. Active flue gas condensation emerges as the most broadly viable measure, adopted by 36%–84% of plants across scenarios. Heat pumps deliver the largest per-plant biomass savings but depend on waste heat availability and electricity prices. Plants near hydropower facilities achieve levelized costs approximately 13 EUR/MWh lower than grid-connected configurations. Waste-heat heat pumps also provide operational flexibility, with 14.4% of their annual electricity consumption shiftable in response to hourly price signals. The findings provide a basis for prioritizing retrofit investments and designing targeted policy support across plant types and market conditions.
The increasing adoption of electric vehicles (EVs) presents significant operational challenges for charging infrastructure, particularly in managing charging demand under feeder capacity and electricity tariff constraints. This paper proposes a practical data-driven framework that integrates short-term charging load forecasting, anomaly detection, and tariff-aware scheduling using real Open Charge Point Protocol (OCPP) telemetry collected from a multi-station residential charging network. The forecasting module employs lightweight statistical models with Fourier seasonality, while an Isolation Forest identifies abnormal charging behavior and a heuristic scheduling strategy coordinates charging according to feeder limits and time-of-use electricity prices. Experimental results demonstrate that the proposed framework achieves up to 35.0% peak demand reduction and 28.0% electricity cost savings while maintaining 100% energy delivery. The forecasting component requires only 0.0247 s for model training and 0.0008 s for prediction, demonstrating its suitability for real-time deployment. The results show that practical and computationally efficient methods, when integrated using real OCPP operational data, can effectively improve EV charging management under realistic operating conditions.
District heating and cooling (DHC) networks, particularly Fifth Generation (5GDHC), effectively reduce building energy use and greenhouse gas emissions by integrating low-grade thermal energy sources at neutral temperature levels. However, the operation of these networks is not yet optimised. A key challenge is the integration and control of multiple distributed heat and cold sources, with pumping energy being crucial at neutral temperatures. Rule-based control sequences are conventionally used to manage these networks, whereas more advanced strategies like optimal control can act as a system integrator, facilitating the transition to a cost-effective decarbonised heating and cooling sector. This paper delves deeper into control strategies for a virtual 5GDHC network, comparing current-practice rule-based control with white-box optimal control approaches through dynamic simulations. A sensitivity analysis evaluates system sizing strategies and their impact on control and performance. Results show that optimal control significantly improves thermal comfort, particularly during transitional and cooling seasons, while reducing energy use by almost 50%. These benefits are achieved by optimising network temperatures, utilising anticipatory control, and leveraging the buildings’ thermal inertia and the different building loads. Additionally, optimal control enables substantial component size reductions by exploiting system flexibility during operation and thus acting as an effective system integrator.
Decarbonisation of energy systems increasingly relies on integrating various energy carriers as part of a broader sector-coupling strategy. Especially in the industrial and transport sector, hydrogen has become increasingly important, offering flexibility across several domains. As energy systems grow more interconnected, the need for smart supervisory-level control increases. However, hydrogen components introduce unique operational challenges due to their intrinsic non-linear behaviour. This work contributes to developing such supervisory control strategies, supporting the transition to climate-neutral, smart multi-energy systems.Therefore, an existing model predictive control framework for multi-energy systems is extended to hydrogen subsystems. This expanded framework enables coordinated control of hydrogen flows alongside other energy carriers, focusing on operational optimisation and strategic component use. Various modelling strategies for pressure-mass flow relationships are presented, ranging from linear to piecewise-affine (PWA) linearised and non-linear formulations. These are compared in simulation scenarios for load profiles and system constraints, with quantitative accuracy metrics against reference data from the National Institute of Standards and Technology (NIST).The paper shows that non-linear models require careful reformulation to achieve computational tractability, yet they do not provide improved accuracy over PWA models: the Redlich–Kwong equation of state introduces a systematic bias exceeding that of a 4-segment PWA approximation of NIST data. Linear models such as the ideal gas law severely underestimate pressure levels, leading to over-production and component damage risks. Thus, PWA models offer the best compromise between speed, accuracy, and safety, making them the recommended modelling approach for real-time supervisory control.
The increasing complexity of energy systems transitioning towards decarbonization requires advances in planning tools, particularly in terms of spatial and temporal resolution. This study applies EnergyPLAN, together with its MULTINODE add-on, to model energy systems at high spatial resolution. We revise the original MULTINODE equations to better represent cross–border flows and introduce a post–processing procedure that improves the dispatch of thermal generation, hydro reservoirs and pumped storage. We also use PyPSA to cluster the national transmission grid and estimate representative internal inter-regional transfer capacities.The approach is applied to mainland Portugal’s 2023 electrical system, comparing results at the national, 2-region, and 5-region (NUTS 2) scales and performing sensitivity analyses on the most critical methodological assumptions. Results show that the NUTS 2 configuration significantly improves model accuracy by more faithfully reproducing the dispatch of all technologies, reducing deviations from real-world data to below 16%. Higher spatial resolution also reveals important regional dynamics, such as Lisbon’s structural import dependency and Alentejo’s suitability for developing green-hydrogen hubs. However, challenges remain regarding the availability of granular data and the estimation of internal transmission limits.
The global Arctic and sub-Arctic region remains underrepresented in high-resolution techno-economic analyses of east-west bifacial vertical photovoltaics systems. This study delivers detailed assessments of electricity yield and levelised cost of electricity for these regions, evaluating fixed-tilted, horizontal single-axis tracking, and east-west bifacial vertical photovoltaics technologies, as well as their combined portfolios. A novel methodology is introduced to quantify the diurnal performance similarity between non-tracked and horizontal single-axis tracking systems, with results mapped on a high resolution 0.45° × 0.45° gridded map of the region. In the Arctic and sub-Arctic region above 66° N, horizontal single-axis tracking achieves 767-2171 full load hours annually, while the east-west vertical portfolio demonstrates a relative yield deficit of only 0-13% and strong diurnal correlation (0.88-0.95) above 75° N. A 50/50 mix of fixed-tilted and east-west vertical systems achieve similarly high correlation (0.86-0.96) between 60° N and 75° N. For fixed-tilted portfolios, extreme electricity yield deficit is observed compared to horizontal single-axis tracking ranging between 3.9% and 37.3%. The reference portfolio's levelised cost of electricity ranges from 18.2 €/MWh to 51.7 €/MWh, with the east-west vertical configuration exhibiting a 1.4-14.0% reduction in all northern locations above 75° N. These findings highlight the technical and economic viability of east-west bifacial vertical solar photovoltaics systems for low-cost electricity generation in the uppermost Arctic and the essential role of conventional tilted setup in sub-Arctics.
District heating networks require well-adjusted heating curves to ensure efficient heating while maintaining indoor comfort. Misconfigured heating curves can lead to insufficient heating or energy waste, yet determining when an adjustment is needed remains challenging. This study evaluates two data-driven methods for identifying substations that require heating curve updates and demonstrates why, under current data and labeling conditions, they largely fail to provide reliable signals for heating curve misalignment. First, we use the Prophet model to analyze return temperature trends, assuming that persistent deviations indicate a mismatch between supply temperature and building demand. Second, we train a conditional autoencoder (CAE) as a normal behavior model to detect anomalies before customers report insufficient heat.The results show that the Prophet model is only reliable for certain substations. Its performance varies widely and often fails to capture complex patterns. The CAE successfully detects technical hardware faults, such as safety valve malfunctions that cause abrupt physical inconsistencies. However, incorrect heating curve settings usually result in gradual drifts within normal system temperature ranges rather than sudden anomalies. Consequently, the CAE interprets the suboptimal settings as normal behavior.Key challenges include a lack of preventive maintenance data, since models are trained primarily on customer reports rather than operator interventions. Furthermore, data gaps and substation-specific models that overfit reduce the robustness of network-wide detection. These findings suggest that identifying misaligned heating curves requires more than just detecting abrupt sensor changes and highlighting the need for datasets that incorporate operator-led interventions and richer labels.
The increasing penetration of photovoltaic (PV) generation has intensified the temporal mismatch between energy supply and demand in park-level energy systems. However, the coordinated planning value of energy storage and electric vehicles (EVs) in high-PV building clusters, especially the trade-off between system and user costs, remains insufficiently quantified. This study develops an optimization model for an energy storage-EV coordinated multi-energy complementary modular energy supply system considering PV output uncertainty. Typical days and multiple weather scenarios are used to describe PV variability, and a bi-objective framework is formulated to minimize the system net present cost (NPC) and EV user NPC while jointly optimizing equipment capacities and hourly operation. EVs are aggregated as a bidirectional virtual storage unit through vehicle-to-grid to enhance flexibility against PV fluctuations. A case study of a building cluster in western Inner Mongolia is conducted. The results show that the coordinated scenario improves regulation capability and PV utilization. The PV utilization rate reaches 99.51%, compared with 93.69% and 91.90% in two non-coordinated benchmarks. The system NPC is 33,856.71 & times;104 & YEN;, which is 279.62 & times;104 & YEN; and 617.08 & times;104 & YEN; lower than the benchmarks, respectively. Carbon emissions are reduced by 14,903.05 t and 13,946.70 t.
Industrial electrification efforts and the expansion of renewable generation, combined with lengthy grid infrastructure permitting processes, have created extensive queues for connections across European electricity systems, limiting the ability to connect new customers under existing market structures. This paper develops a coordinated flexibility trading mechanism, operated by an independent aggregator, that mediates trades between customers and the grid operator. The mechanism is formulated as a two-stage optimisation model, including a day-ahead scheduling stage followed by a flexibility market that coordinates capacity and energy trades under network constraints. Using a case study of an industrial distribution network with six customers, we show that coordinated flexibility trading can reduce load shedding and allow for additional non-firm connections without grid reinforcements by reallocating capacity rights and sourcing of flexibility. In network-wide coordination scenarios, the mechanism reduces total system costs by up to 15% compared to local-only coordination, with outcomes depending on the congestion type and magnitude. Our framework demonstrates that existing grid infrastructure can accommodate substantially higher connection capacity through coordinated flexibility mechanisms, offering a practical intermediate solution between costly grid expansion and nodal market efficiency.
This paper provides an up-to-date overview of legislation focused on the development of markets for hydrogen and other low-carbon fuels in the European shipping sector. We discuss the physical properties of a selection of these fuels and the current technical and economic hurdles they face for wider adoption in the shipping sector. These attributes, which strongly influence production, transport, and storage costs for each fuel, will engender different market architectures depending on which fuels prevail. Drawing upon lessons from the natural gas industry, we offer insights about economic regulations that will encourage competition and the efficient utilization of infrastructure across these potential market architectures. Given the uncertain future of markets for low-carbon shipping fuels, we encourage policymakers to think proactively about economic regulations and consider adjustments as these markets come into sharper focus in the coming years.
5th generation district heating and cooling (5GDHC) networks have experienced a vast boost in newly built residential districts. High shares of cooling demands as waste heat sources to the network and correspondingly high demand overlaps are often reported as key factors for their feasibility. For a roll out of the concept of 5GDHC networks into heating-dominated existing building stock, technical and non-technical requirements apply which may not be valid in newly built districts. In addition, shares of cooling demands as waste heat sources to the network and correspondingly high demand overlaps are most often not present in heating-dominated climates and residential areas. (i) A summary of technical requirements for 5GDHC networks in existing districts is given which need to be met or at least considered in its planning. (ii) In a simulation-based techno-economic analysis a 5GDHC system is compared to an insulated 4th generation district heating (4GDH) network and individual, decentralized air-source heat pumps at varied linear heat density. All examined systems rely on ambient air as the main heat source. Besides the line density, varied input parameters comprise the proportion of cooling demand and the installed PV capacity in rural and municipal districts. 5GDHC networks attain lower electricity imports than 4GDH networks even in absence of continuously available environmental heat sources other than ambient air, PV generation or elevated space cooling demands. Economic advantages over 4GDH networks in the defined setup are observed for low line densities in both district characteristics under the specified cost structure and modelling assumptions.
The expected ramp-up of heat pumps and electromobility as part of the energy transition will result in an increasing overall load on the electricity grid. To take this into account for grid planning, grid operators typically use simultaneity factors or simultaneity functions to determine the expected peak load. In case of decentralized heat pumps, the electricity load correlates strongly with the thermal demand for heating and domestic hot water. Reduced efficiency at low temperatures affects the electricity load additionally, the operation of heating elements at low outside temperatures implies a further challenge for the grid. In addition to the peak load, the daily profile of the electricity consumption of heat pumps is of major relevance for electricity balancing management. Typically, electricity providers use temperature-dependent standard load profiles for electricity consumers. However, the number of active customers, who will adjust their consumption depending on electricity prices or the availability of (renewable) generation capacity and grid capacity, will significantly increase in the future. Hereby, flexible operation will affect both the maximum peak load and the distribution of electricity consumption over the days.To assess all these aspects, a comprehensive study has been performed using a large set of operational data derived from more than 6000 installed heat pumps over up to 2.5 years as well as from scientific monitoring projects in Germany. To support the correlations derived from the operational data e.g. at very low temperatures, simulation models have been developed according to respective buildings, system designs and operational behaviours. To evaluate the impact of flexible heat pumps, the operation has been simulated and optimized with respect to flexible tariffs for future heat pumps, considering electro-mobility as well as renewable energy market penetration.
Waste heat is widely recognized as a key lever for advancing climate neutrality in urban heating systems. While recent policy regulations have significantly improved the transparency of waste heat potentials, the gap between reported theoretical potentials and their practical usability remains insufficiently explored. This study develops a novel residual-aware, guided source-demand allocation approach, leveraging officially reported, technically detailed waste heat source information and high-resolution building typology data to reveal key insights into spatial and technical utilization potentials and limitations. The results also derive key distributional and operational source characteristics, combined with spatial and typological building patterns, considering three refurbishment scenarios. The proposed methodology provides a transparent and policy-relevant tool for translating reported waste heat potentials into actionable planning insights. The results indicate that the utilizable share of the reported waste heat potential is reduced to approximately 35% due to techno-spatial constraints, decreasing further to approximately 24% when temporal source-demand coincidence is additionally considered. Depending on building refurbishment progress, the number of potentially supplied buildings ranges from 1.8 to 5.7 million. Overall, these findings demonstrate that coordinated spatio-temporal planning, combined with efficiency measures, is essential to unlock the significant, yet primarily spatially constrained, waste heat potential.
This paper proposes and experimentally validates a two-stage scheduling and control strategy for a behind-the-meter battery energy storage system (BESS) delivering both local and grid services. Considered services are the maximization of PV self-consumption, peak-load reduction, and secondary frequency control (aFRR).The day-ahead stage allocates battery capacity across local and balancing services using a scenario based approach, reflecting potential remuneration from aFRR participation without committing to fixed power availability; in the real-time stage, BESS set-points are computed in a periodic fashion at a high time resolution based on updated information on balancing prices, net load realization and BESS state of charge. The strategy is experimentally validated on a building at the Energypolis Campus of HES-SO Valais (Sion, Switzerland), which exhibits a peak power demand of 300 kW and is equipped with a 264 kWh / 140 kW lithium-ion BESS. The experimental results demonstrate the effectiveness of the proposed framework in scheduling and actuating the provision of both behind-the-meter and front-of-the-meter services.
The role of hydrogen in future energy systems is uncertain due to a wide range of factors. This paper lays out the current trends, state of the art, and future perspectives for flexible production of an energy carrier that is crucial to the decarbonisation of hard-to-abate sectors. The approach of the paper has been to compare current projects and studies of the technological progress of electrolysis, investment trends and energy system scenarios. This is discussed in a holistic comparison between two large regions: the EU and the US-with focus on hydrogen supply and policies. We find large discrepancies between these regions, especially in expectations of renewable energy and infrastructure build-out. However, in both regions, a significant capacity of electrolysers will be required by 2050 to reach net-zero emissions and in both regions, flexible operation of electrolysers with 3-5000 h of operation is required to facilitate integration of renewable energy. Thus, a large discrepancy between announced projects and future needs is evident, highlighting the demand for policies to encourage investments. Here, a fixed subsidy as proposed in the US may promote fast investments, while auctions as implemented in the EU may provide cost-efficient, but potentially slower.
The increasing share of renewable energy in the power sector has led to increasingly volatile electricity prices in the Northern European power markets. The district heating systems in Finland will see a rapid expansion of heat pumps and electric boilers. Thermal energy storages enable flexible heat production, when it is cheap. Thermal energy storages can also help combined heat-and-power production by decoupling the electricity production from the heat demand. In this study, North European energy system model was used to study the impact of increasing heat storage and demand response capacity in Finland’s highly electrified district heating systems on the electricity system flexibility.The results show that the impact of the additional heat storages on the annual production mix is only marginal with relative changes of less than 0.6%. The most significant impact was on the production of combined heat-and-power and electric boilers. The use of electric boilers increased especially in the climate year scenario of 2016, where the electricity production costs were highest. The increased flexibility can be seen in the terms of decreased ramping of hydropower and electricity imports. At the same time, the ramping of combined heat-and-power and electric boilers increased. Tank thermal energy storages contributed to this more. The use of demand response differed from the tank thermal energy storages as it was used more for short term adjustments of power demand with lower charging and discharging rates. In the end, the impact on the entire energy system remained small.
This contribution addresses a methodological gap in energy system modelling by developing an optimisation approach that endogenously integrates battery waste heat into sector-coupled electricity and district heating systems. With increasing shares of variable renewable electricity, battery storage plays a growing role in providing flexibility, while simultaneously generating low-temperature waste heat that is typically neglected in planning models. The proposed framework extends existing optimisation approaches by representing batteries not only as electricity storage units but also as potential low-temperature heat sources. In contrast to predefined waste heat profiles, thermal energy is generated endogenously as a direct byproduct of battery charging and discharging operations. Battery dispatch is optimised in response to electricity price signals and system constraints, such that waste heat availability is directly linked to operational behaviour and capacity sizing decisions. Recoverable waste heat can be utilised directly, upgraded via heat pumps, or stored based on techno-economic trade-offs. The approach is implemented within the open-source energy system model Calliope and applied to a realworld district heating planning project in eastern Germany. Results show that endogenous waste heat integration substantially alters optimal system design, increasing battery and photovoltaic capacities by 54.22 % and 39.53 %, respectively, while reducing cumulative heat pump capacity by 1.9 %. In contrast, the impact on net heat generation costs remains moderate, with maximum savings of up to 0.45 ct/kWh in the best-case scenario, which corresponds to annual cost savings of approximately 43.800 & euro;.
District heating network simulation faces a fundamental computational challenge: traditional nonlinear models become intractable at large scales due to the curse of dimensionality, while linear models cannot accurately represent the nonlinear dynamics essential for district heating systems. Tensor-based methods have demonstrated effectiveness in modeling heating, ventilation, and air conditioning (HVAC) as well as local heating systems by providing a scalable compromise between accuracy and computational efficiency, yet their application to district heating networks is first described in this paper. This work applies multilinear time-invariant (MTI) modeling using a tensor-based framework for scalable representations of district heating networks.Tensor and multilinear functions efficiently represent the governing equations and their nonlinear relationships, especially the quadratic pressure-loss relationships defined by the Darcy-Weisbach equation and nonlinear friction factors across flow regimes without causing an exponential growth in model complexity. Binary variables model discontinuous transitions between laminar and turbulent flow, maintaining computational tractability while preserving physical accuracy. The tensor structure inherently avoids the curse of dimensionality that constrains conventional approaches by factorization. Benchmarking against established models on a small network shows minimal deviations alongside considerable memory reductions, demonstrating the potential of tensor-based methods for efficient simulation and optimization of large-scale district heating networks and supporting the integration of renewable energy sources and advanced control strategies essential for modern energy-efficient systems.
