Periodic Reporting for period 1 - MAST3RBoost (Maturing the production standards of ultraporous structures for high density hydrogen storage bank operating on swinging tem-peratures and low compression)
Période du rapport: 2022-06-01 au 2023-11-30
Because the transport segment makes up about one-third of all CO2 emissions in the EU (> 1,000 MILL ton), its decarbonization represents a key element in achieving the energy transition. Fuel Cells and Hydrogen (FCH), outperforming batteries in all relevant indicators, is the most promising solution for decarbonizing trucks, buses, ships, trains, large cars, with commercial vehicles being considered the early adopters. By 2030 this new industry has the potential to generate a € 130 bn market only in the EU. The market-entry goal is to fit 5 kg H2 in a gasoline equivalent tank (80 kg/90 l). However, state-of-the-art technology for H2 storage on-board, based on compression at 700 bar, is still disappointing in terms of volumetric density (25 gH2/lsys), preventing a widespread penetration of FCEVs. At least 40 gH2/Lsys is considered a significant milestone (settled by the DOE) to provide the market with an actual FECV replacement to current internal combustion engine vehicles, ICEV. MAST3RBoost will enable a disruptive path to meet these goals based on a new generation of ultraporous materials (ACs and high-density MOF) with H2 delivery capacities 33% higher than current record holders (NU-1103, SNU-70, MOF-5). KPIs already demonstrated are > 9 wt% and 44 gH2/lPS at 100 bar and below. Best candidate materials will be improved with Supervised Machine Learning and industrially produced as pellets or monoliths at a scale beyond 10 kg for the first-time. A brand-new pressure vessel technology of 20+ litres will be fully designed via Digital Twins within the consortium to operate at the optimum thermodynamic (TPS) regime for the materials. This will be the first adsorption-based demonstrator worldwide with a capacity of 1 kg H2, aiming at a system KPI as high as 33 gH2/lsys, and become a record-holder among all H2 storage technologies, with a projected system cost of 1,780 € (5.6 kg H2). MAST3RBoost’s ground-breaking additive manufacturing technology (WAAM) will create lightweight type I vessels with dedicated shapes to better fit on-board specific transportation spaces. In round numbers, the project will produce materials with a projected SAM by 2028 of € 345 MILL. A business case based on a 1,250 ton plant (CAPEX: € 14 MILL) of the new ultraporous material has the potential to support 20,000 unit passenger FCEV (or 2,000 heavy-duty FCEV), with an estimated FOB price <15 €/kg.
The application of Machine Learning, ML, to the development of adsorbent materials has been conceptualized. In the synthesis of AC, specific inputs are related to the elementary composition of the starting material (char), while for the synthesis of MOFs the presence of trace metals in the recycled raw material has been selected as one of the key parameter. The outputs include the volumetric capacity for H2 storage at 100 bar and cryogenic conditions.
A database for unsupervised ML was generated taking advantage of the existing publications by the University of Nottingham. The database is currently populated with >180 samples. The outputs include mainly textural parameters. A new database is being built that will also include outputs such as the H2 uptake at 100 and 5 bar and the determination of the packing density of the materials. Supervised Machine Learning and Active Learning strategies have also been considered and put to the test with promising results. 2D maps generated via supervised ML have shown some clustering of the more promising materials.
The generation of a new library of ACs has been actively pursued with more than 50 new samples already generated mainly based on CO2 activation. The total sum of 100 new AC samples is expected to be achieved by adding chemical activation of a number of precursors. Furthermore, several synthetic bio-based polymers have been produced varying in their O/C and N/C ratios. Polymeric precursors have been used for the first time in the production of ACs, via pre-mixing with biomass.
In parallel, preliminary scale-up of highly microporous ACs has been demonstrated at the scale of hundreds of grams. A wide typology of pore size distributions has been achieved. Most of the scaled-up ACs have high BET surface areas up to 3000 m2/g. Working capacities as high as 9.5 wt% of H2 are estimated based on direct measurements at 100 bar and 77 K and the retention modelled at 5 bar and 157 K.
The catalogue of MOFs is based on the use of recycled raw materials, e.g. Acid Mine Drainage, AMD, as the Fe source, poly-Alu from food multilayered packaging as the Al source, and PET bottles as the source for the ligand benzene dicarboxylic acid, BDC. For Fe-based MOF solvolysis is used to produce a number of bimetallic MOF, with a second, artificially added, metal. For the Al-based MOF (Al-BDC) the main variable studied is the temperature of the hydrothermal synthesis.
The scale-up efforts have focused on Al-fumarate as a model MOFs. Batches of up to 3 kg have been produced with good granulation behaviour.
Relevant progress was made also in relation to the structural materials for the vessel via Wire Arc Additive Manufacturing, WAAM. The cryogenic tensile properties of Aluminum alloy MA-5183 are excellent. In addition, the WAAM samples displayed excellent microscopic properties: equiaxed grain morphology and low pore defects. In parallel, qualification of coatings to be used as a barrier has proceed, and is particularly relevant to produce moisture-induce H2 embrittlement. Three coatings passed all tests including thermal-shock testing.
The design of the tank and the complete refuelling system has witness major progress. Simulations based on Computational Fluid Dynamics (CFD) shown large internal thermal gradients due to limitations in the efficiency of heat dissipation. This can greatly limit the efficiency of the storage under real-life refuelling conditions (e.g. 10 min). To tackle this problem, innovative heat exchanger geometries have been designed, exploiting the flexibility of additive manufacturing. All the progress made in relation to the design and the WAAM technology is expected to crystallize in the manufacturing of the actual vessel by M30 (instead of M40), enabling extra-time for the actual testing of the system.
A database for unsupervised ML was generated taking advantage of the existing publications by the University of Nottingham. The database is currently populated with >180 samples. The outputs include mainly textural parameters. A new database is being built that will also include outputs such as the H2 uptake at 100 and 5 bar and the determination of the packing density of the materials. Supervised Machine Learning and Active Learning strategies have also been considered and put to the test with promising results. 2D maps generated via supervised ML have shown some clustering of the more promising materials.
The generation of a new library of ACs has been actively pursued with more than 50 new samples already generated mainly based on CO2 activation. The total sum of 100 new AC samples is expected to be achieved by adding chemical activation of a number of precursors. Furthermore, several synthetic bio-based polymers have been produced varying in their O/C and N/C ratios. Polymeric precursors have been used for the first time in the production of ACs, via pre-mixing with biomass.
In parallel, preliminary scale-up of highly microporous ACs has been demonstrated at the scale of hundreds of grams. A wide typology of pore size distributions has been achieved. Most of the scaled-up ACs have high BET surface areas up to 3000 m2/g. Working capacities as high as 9.5 wt% of H2 are estimated based on direct measurements at 100 bar and 77 K and the retention modelled at 5 bar and 157 K.
The catalogue of MOFs is based on the use of recycled raw materials, e.g. Acid Mine Drainage, AMD, as the Fe source, poly-Alu from food multilayered packaging as the Al source, and PET bottles as the source for the ligand benzene dicarboxylic acid, BDC. For Fe-based MOF solvolysis is used to produce a number of bimetallic MOF, with a second, artificially added, metal. For the Al-based MOF (Al-BDC) the main variable studied is the temperature of the hydrothermal synthesis.
The scale-up efforts have focused on Al-fumarate as a model MOFs. Batches of up to 3 kg have been produced with good granulation behaviour.
Relevant progress was made also in relation to the structural materials for the vessel via Wire Arc Additive Manufacturing, WAAM. The cryogenic tensile properties of Aluminum alloy MA-5183 are excellent. In addition, the WAAM samples displayed excellent microscopic properties: equiaxed grain morphology and low pore defects. In parallel, qualification of coatings to be used as a barrier has proceed, and is particularly relevant to produce moisture-induce H2 embrittlement. Three coatings passed all tests including thermal-shock testing.
The design of the tank and the complete refuelling system has witness major progress. Simulations based on Computational Fluid Dynamics (CFD) shown large internal thermal gradients due to limitations in the efficiency of heat dissipation. This can greatly limit the efficiency of the storage under real-life refuelling conditions (e.g. 10 min). To tackle this problem, innovative heat exchanger geometries have been designed, exploiting the flexibility of additive manufacturing. All the progress made in relation to the design and the WAAM technology is expected to crystallize in the manufacturing of the actual vessel by M30 (instead of M40), enabling extra-time for the actual testing of the system.
The first set of significant results produced by the MAST3RBoost project potentially beyond the state of the art include:
- A pioneer Machine Learning-compatible and extensive database correlating synthesis conditions of ACs with their final textural and composition parameters.
- Activated Carbons produced from biomass at the scale of hundreds of grams under industrial-compatible conditions with excellent gravimetric working capacity for H2 (9.5 wt%), already on the verge of the best materials in the literature.
- A variety of MOF structures successfully produced for the first time in the lab with recycled sources of Aluminum, Iron and Benzene Dicarboxylic acid (BDC) ligand.
- Fast maturing of the WAAM technology for the use of Al-based alloys in the manufacturing of vessels and heat exchangers suitable for application under cryogenic conditions and totally compatible with dry H2.
- Innovations in the design of heat exchangers geometries enabled by additive manufacturing that can reduce the thermal gradients under realistic refuelling conditions.
- A pioneer Machine Learning-compatible and extensive database correlating synthesis conditions of ACs with their final textural and composition parameters.
- Activated Carbons produced from biomass at the scale of hundreds of grams under industrial-compatible conditions with excellent gravimetric working capacity for H2 (9.5 wt%), already on the verge of the best materials in the literature.
- A variety of MOF structures successfully produced for the first time in the lab with recycled sources of Aluminum, Iron and Benzene Dicarboxylic acid (BDC) ligand.
- Fast maturing of the WAAM technology for the use of Al-based alloys in the manufacturing of vessels and heat exchangers suitable for application under cryogenic conditions and totally compatible with dry H2.
- Innovations in the design of heat exchangers geometries enabled by additive manufacturing that can reduce the thermal gradients under realistic refuelling conditions.