The Green Village (TGV), located at TU Delft, is a regulatory-free field lab where innovative projects relevant to the energy transition are tested in a built environment. In this context, seasonal storage is considered a promising research pathway for TGV, as increasing renewable energy penetration and resulting grid congestion are expected to make seasonal storage a key factor in the energy transition. The objective of this thesis is to assess whether liquid organic hydrogen carriers can provide a feasible seasonal storage solution for the TGV community and to compare it to a compressed hydrogen storage pathway.
A literature review was conducted to shape the research angle of this thesis by gaining insights from relevant studies, identifying the principles and components necessary for the LOHC process, and examining existing safety regulations and measures applicable to this storage pathway. The review identified that economic viability was hindered by the heat supply, carrier selection, and system integration. At the same time, an evident gap was found for community-scaled systems and for studies that model the entire energy system under varying loads. Currently, there is no clear regulatory framework for hydrogen storage, as it is still in development. However, several studies focused on the safety of hydrogen-related components suggest using PGS 35 as a technical guideline. This document outlines the necessary permits, safety measures, and distance requirements for such installations. For the LOHC, there is currently no specific regulatory framework; however, some studies suggest that PGS 29 may be applicable to certain aspects of the process. The system of focus for this study is a community-scale hybrid energy system connected to the grid for TGV. This system comprises a PV system, a LiFePO4 battery for short-term storage, an electrolyser for hydrogen production, seasonal hydrogen storage, and a PEM fuel cell for electricity generation during periods of low PV output. For this system, two separate hydrogen storage pathways were developed. The primary scenario involves an LOHC-based hydrogen storage system (a quick overview of this scenario can be found in Figure 1.2). In this scenario, hydrogen is bonded to the LOHC through hydrogenation during summer for seasonal storage. Then, during winter, when PV availability is scarce, it is converted back through dehydrogenation. In the second scenario, hydrogen is stored in compressed gas cylinders (an overview of this system can be found in 3.4). Both pathways’ operation was modelled to follow the energy management system currently applied in TGV Energy Hub 24/7. This prioritises direct PV use, daily battery cycling, and extended operation of the electrolyser during the summer (with one start per day) to produce the seasonally stored hydrogen, as well as dispatch of the Fuel Cell during winter utilising the stored hydrogen. This way, grid exchanges are minimised as they occur only after all internal energy sources have been fully utilised. To assess the feasibility of these scenarios, a techno-economic model was developed in MATLAB. The model employs a modular approach, creating a module for each component of the system. The overall sizing and cost estimation are conducted by a main script that simulates the entire calendar year 2023 at an hourly resolution. The model preprocesses load and PV inputs, applies three hard constraints that limit annual PV export to 10%, annual grid import to 5% of load, and annual LOHC mass balance deviation to 0.5%, and then runs the EMS, battery, electrolyser, hydrogenation, dehydrogenation, and fuel-cell modules. Peak duties and temperatures are extracted for ASPEN EDR heat-exchanger sizing and costing. Economic results are reported as CAPEX, OPEX, LCOE, NPV, and IRR. The validation of the model is conducted through internal checks of the EMS logic, evaluation of pitch diagrams to ensure thermodynamic consistency, and comparison of results with established manufacturer and literature values. For the LOHC scenario, the optimal configuration is found for 552 PV modules with a total capacity of 165.6 kWp, a 261 kWh battery, a 116 kW electrolyser, and a 34 kW PEM fuel cell. Annual grid import is 0.88 % of load, and annual PV export is 8.79 %. Summer electrolysis produces about 1,813 kg of hydrogen that is stored in the NEC. Annual hydrogenation produces 30,986 kg of hydrogenated carrier. Dehydrogenation releases 1,610 kg of hydrogen, of which 898 kg is used by the PEM fuel cell and 710 kg is combusted for process heat. The annual requirement of hydrogenated NEC is 30,591 kg, which results in a mass-balance deviation of 0.41 %. Waste-heat recovery covers 88 % of hydrogenation preheat duty and 11.4 % of dehydrogenation duty. An overall system efficiency of 41.4 % is found, with 49.5 % for the LOHC chain and an electrical round-trip efficiency of 23.4 %. The capital expenditure is €2.36 million, of which €1.64 million is attributed to the LOHC process and about €1.16 million to the initial NEC fill. The operational expenditure is €12,901 per year. The levelised cost of electricity is €1.695 per kWh with an NPV of −€2.67 million and an IRR of −29.5 %. The total footprint of the LOHC system was found at 1,020 m2 while the space required for the process tanks and reactors is estimated at 66 m3 when two tanks are used and 36 when a single tank is used. For the alternate scenario with compressed hydrogen, the same EMS and component models are used, except for hydrogenation and dehydrogenation. This scenario resulted in a PV system consisting of 379 PV modules and totalling 113.70 kWp, the same battery and PEM fuel-cell sizes, and an 80 kW electrolyser. Annual PV export is 10.38 % and annual import is 1 % of load, with an annual hydrogen balance error of 2.09 %. The capital expenditure is €565,306, and the operational expenditure is €6,004 per year. The levelised cost of electricity is €0.79 per kWh with an NPV of −€0.732 million and an IRR of −18.21 %. The total area required for the system is approximately 699 m2 while the space used for the 1,039 50L tanks required a total of 52 m3.
Sensitivity analysis indicates that the discount rate is the principal driver of LCOE variation within a range of about −9.8 to +10.3 %. Project lifetime produces a range of about −5.8 to +9.5 %. LOHC process costs, in particular the carrier, produce changes of about ±8 % in LCOE and have notable effects on NPV.
Concluding, the case study shows that a PV–battery–electrolyser–fuel-cell system using LOHCs can deliver near year-round autonomy at The Green Village, but its energy efficiency is constrained by dehydrogenation heat demand (only 12% high-grade heat recovered with PEMFC) and consequent hydrogen combustion, which limits the electrical round-trip efficiency to 24% despite strong waste-heat integration on the hydrogenation side. Economically, LOHCs remain impractical at current prices as CAPEX is dominated by the LOHC chain (including the NEC inventory) and yields an LCOE around €1.70/kWh with strongly negative NPV/IRR, whereas a compressed-hydrogen baseline achieves similar operational adequacy at far lower cost but still fails to reach profitability under base assumptions. The feasibility is primarily influenced by financing conditions and the lifespan of components, with LOHC specific costs (particularly the price of the carrier) being the next most significant factors. In contrast, the costs associated with PV/batteries are less critical. This suggests that improvement efforts should focus on enhancing high-grade heat integration (such as SOFC fuel cells), reducing capital and operational expenses in the LOHC chain, and implementing financing or policy strategies to lower capital costs. Finally, safety and siting considerations favour LOHC’s ambient-condition liquid storage for built environment applications, while the compressed-gas alternative required as high amount of pressure vessels and safety setbacks; taken together, these results suggest LOHCs are technically feasible with a favourable safety profile, but deployment hinges on cost and heat-management breakthroughs or supportive market frameworks before community-scale adoption becomes realistic
... Read more