High temperature electrolyser with novel proton ceramic tubular modules of superior efficiency, robustness, and lifetime economy

Final Report Summary - ELECTRA (High temperature electrolyser with novel proton ceramic tubular modules of superior efficiency, robustness, and lifetime economy)

Executive Summary:
ELECTRA runs 3 years (2014-2017) with 7 partners (4 industries/SMEs, 3 research institutes and universities) coordinated by University of Oslo. It develops tubular proton ceramic electrolyser cells (PCEs) and a kW-size multitubular module operating around 700°C for production of hydrogen from steam at up to 20 atm pressure.
PCEs have the potential advantage to produce directly pure dry pressurised hydrogen at the maximum total pressure of the electrolyser, as compared to polymer and solid oxide electrolysers. Furthermore, tubular modules may handle certain aspects of pressure, sealing, and lifetime issues better than planar concepts. Co-electrolysis of steam and CO2 to syngas by using the proton ceramic electrolyte in a co-ionic (proton and oxide ion) mode is also investigated. Finally, a techno-economic analysis is made for steam electrolysis and co-electrolysis with supply of heat or/and steam, CO2, and electricity from geothermal and solar-thermal sources.
Supporting H2-side Ni-cathode and doped BaZrO3-based cermets and electrolyte materials and their sintering procedures have been selected and optimised. Tubular cells are developed in three stages; Single cell tubes (Gen1), stacked tube segments (Gen2), and segments-in-series deposited on a porous support tube (Gen3). Results indicate that Gen3 is the most promising technology in long term, while Gen2 is potentially cheap and functional for certain applications, and will be developed to scaled-up production and assembly in a follow-up-project.
Perovskite-based candidates for steam+oxygen side anodes have been identified and developed. The partial anode ASR for proton ceramic operation has been brought below the milestone target of 0.2 ohm*cm2 in steam+oxygen atmospheres. Stable and sufficiently conducive current collectors and interconnects have also been identified and tested, along with various seals depending on tube generation.
Tubular single cells have been tested with high stable currents (several A) and high faradaic efficiency attributed to the use of over-atmospheric steam (e.g. 4 bar) and the novel state-of-the-art anode. However, the durability target was not met, attributed mainly to anode degradation, which need further research. Co-electrolysis of CO2 and steam to syngas by the proton ceramic in co-ionic mode was demonstrated on button-cell scale.
The multitubular module with capacity of 18 tubes for 1 kW steam electrolysis allows easy insertion of tubular cells from the cold outside, plus individual monitoring, shut-off, and replacement of tubes. It has been tested to 20 bar, and thermal management, mass flows, and electrical supply have been optimised. The supply of tubes was however not sufficient to test the module for 1 kW hydrogen production as planned within the project timeframe.
Balance-of-plant schemes have been designed for PCE steam electrolysis and co-electrolysis integrated with geothermal and solar-thermal sources of heat, steam, CO2, and electricity, and techno-economic analyses have been made, pointing at critical capital cost limits for viability under given assumptions. In particular, multitubular units based on the progress and experiences in ELECTRA are foreseen to be competitive at moderate scale in a relatively short timeframe because of modularity, ease of fabrication, and cheap materials.
Project Context and Objectives:
High temperature electrolysers (HTEs) can utilise available steam and/or heat to produce hydrogen at higher electrical efficiency and hence lower cost than low temperature (polymer or alkaline) electrolyzers. Planar solid-oxide electrolyser cells (SOECs) have reached a mature lab scale demonstration scale but the high temperature (e.g. 800°C), sealing problems, and oxygen evolution at the anode causing electrode delamination are problematic. More important is probably that the hydrogen is produced mixed with the input steam, hence at a lower pressure than the total pressure of the electrolyser. Less mentioned is the additional hazard of the pure undiluted oxygen produced at the total pressure of the unit, which is a considerable safety hazard.
Proton ceramic electrolyser cells (PCEs) have an electrolyte that transports protons, and thereby form hydrogen on the side, undiluted with steam, and at the total pressure of the electrolyser. Hence, the efficiency of the compression is higher, condensation and drying is alleviated, and the hazard of pure oxygen at the maximum pressure is avoided. (Pure compressed hydrogen is standard technology.) Moreover, the nickel usually used in the cathode is not in danger of oxidation if the current is interrupted in the PCE, unlike the SOEC where the H2/H2O ratio then gets too low and the Ni oxidises. Finally, proton conduction has a lower activation energy than oxide ion conduction, and hence the operation temperature can in principle be lowered to a more ideal range, e.g. 600°C.
However, PCE is a less mature technology. There is less experience in fabrication, and the range of compatible sealants, interconnects, current collectors, and, not least, anode materials is still limited. ELECTRA addresses these challenges. ELECTRA develops tubular proton ceramic electrolyser cells (PCEs) and a kW-size multitubular module for production of hydrogen from steam. Tubular modules may handle many aspects of pressure, sealing, and lifetime issues better than planar concepts, and may provide a faster way to commercial applications on various scales.
Co-electrolysis of steam and CO2 to syngas by using the proton ceramic electrolyte in a co-ionic (proton and oxide ion) mode is also investigated, targeting total efficiencies >85% according to the call. Finally, a techno-economic analysis is made of steam electrolysis and co-electrolysis with supply of heat or/and steam, CO2, and electricity from geothermal and solar-thermal sources.
The targets for steam electrolysis are set at 1 kW power according to the call, which corresponds to around 250 Ln H2 / h. The call furthermore requires operation at 800-1000°C while ELECTRA must necessarily aim for the better range of 600-800°C due to the nature of proton conduction in oxides. The target pressure is 20 bar pressure operation, with high faradaic and electrical efficiency. The faradaic efficiency should exceed 95%, while electrical efficiency can be defined in different ways – and different values are used in the call. Our modelling analyses show that values of >85% are achievable.
The main objective of ELECTRA is to develop scalable fabrication of tubular high temperature electrolyser cells with ceramic proton conducting electrolytes. This encompasses development of oxygen+steam side anode materials which are stable in steam and exhibit mixed proton and electron hole conductivity in order to provide low overpotentials, necessary for high electrical and faradaic efficiency. The milestone objective is to reach an area specific resistance (ASR) for the anode of 0.2 ohm*cm2 within the project, and a total ASR of 1 ohm*cm2 for the entire cell.
Current collection materials are required for both electrodes, with targeted conductivities of 1000 S/cm and 300 S/cm on the cathode and anode sides, respectively.
The tubes must be segmented in series to increase voltage and reduce current. This is done in different development stages: Single cell tubes with co-sintered electrolyte on cermet cathode support tube (Gen1), stacked tube segments (Gen2), and monolithic tubes with cell layers deposited on a single inert porous support tube (Gen3).
The tubes will be assembled in a flexible multitubular module to produce pure dry pressurised H2. The module shall allow cold-side access and easy monitoring, shut-off, and replacement of individual tubes to increase the lifetime of the module. The cells will operate at temperatures of 600-800°C in steam electrolysis mode to promote efficient integration of PCE technology in geothermal and solar heat power plants.
ELECTRA will also perform proof-of-concept test of CO2 and steam co-electrolysis using PCEs in co-ionic mode, enabling novel system concepts for economical production of DME.
To reach its objectives, the project will develop fabrication of segmented-in-series tubes of supported thin proton conducting electrolytes, novel anodes, novel interconnects and current collectors, a 1 kW multi-tube module for high pressure steam operation with the possibility to monitor and replace individual tubes, and process modelling tools. Single tube cells will be tested for efficient steam electrolysis, durability, and proof-of-concept co-electrolysis, while target hydrogen production will be tested and evaluated also in the multitubular module.

Project Results:
ELECTRA develops tubular proton ceramic electrolysers (PCEs) for steam with Y- and Ce-substituted BaZrO3 (BZCY) electrolytes, Ni-BZCY cathodes, perovskite oxide anodes, and various interconnects, current collectors, and seals. The tubes are single cell (GEN1), segmented-in-series by stacking (GEN2) or segmented-in-series by printing segments on a common inert support tube (GEN3). It furthermore manufactures and evaluates a 1 kW multitubular PCE module, tests co-electrolysis of CO2, and makes techno-economic analysis of the use of PCEs integrated with solar-thermal and geothermal sources of heat, steam, CO2, and electricity.
The work in ELECTRA has been organised in 5 technical work packages WP 1-5, in addition to WP6 on dissemination and exploitation and WP7 on management. WPs 1-3 perform steps in making tubes with electrolyte (WP1), developing and applying steam+oxygen side anodes (WP2), and testing the cells for performance and durability in steam electrolysis as well as proof-of-concept testing co-electrolysis of CO2 (WP3). WP4 deals with modelling of balance of plant and techno-economic analysis of electrolysis and co-electrolysis integrated with geothermal and solar-thermal sources of heat, steam, CO2 and electricity. WP5 designs, fabricates and tests a 1 kW size multitubular electrolyser module.

WP1: Development of supported cathode electrolyte tubes

WP1 develops and produces cold-sealed closed-end tubes containing support, cathode, electrolyte, and interconnect (where applicable), for application of anode and current collection in WP2, testing of single tubes in WP3 and in multi-tubular module in WP5.

GEN1 tubular single cells

Nominally 25 cm length cathode-supported tubes with dense electrolyte membranes were produced by extrusion and co-sintered in a solid-state reaction sintering (SSRS) route. These tubes were cut to 5 cm length segments and provided for segment-in-series and electrode development. Initial membrane defect densities of more than 5 per cm were reduced to 1 per 10 cm by continuous process improvements over the course of the project.
A sealing process was developed to join tubes to a sintered glass ceramic shoulder preform affixed to an alumina riser. The same overall process can be used to seal one end of the tube using a glass-ceramic cap. The sealing temperature is 1000°C, which is also the maximum temperature that the tube assembly should be exposed to after sealing. The gas tightness of the seals has been verified by monitoring the He leakage from the outer to the inner compartment using a gas chromatograph. No helium leakage (BDL, less than 10 ppm) was detected during 3 cycles from room temperature to 700°C.
The electrolyte coating production has been optimised to meet grain size and thickness requirements. Parameters investigated were associated with conditioning of raw materials (milling procedure), de-binding step and sintering step (temperature, rate, dwell, atmosphere, use of bung) and with coating parameters (water and organic based solvents; dip-coating versus spray-coating; thickness of the coated layer). Effects of these parameters were identified through extensive characterisation of samples using laser diffraction (particle size distribution analysis), electron microscopy, XPS, and white light interferometry. From this work, mechanisms for sintering of SSRS based electrolyte have been studied enabling to identify main parameters affecting the sinterability of the electrolyte. Results of this work were continuously fed to increase the robustness of the manufacturing process.

GEN2 stacked segmented-in-series cells

Tube segments, interconnect materials, glass preforms and sealing rings were produced and stacked. The design of stacked segments-in-series cells has been improved from a starting point with Pt-wires, glass seals and alumina spacer to a stacking where the glass seal is also the insulating spacer. The stacked cells exhibit good electrical contact between the segments from inner to outer Pt-wire (resistance < 0.1 Ω), and no electrical contact between the membrane surfaces of the segments was measured. Nevertheless, it is challenging to stack more than 3 cells in series with this design.
As an alternative design, Al2O3 ceramic interconnects with Pt paint were tested. Good electrical contacts were obtained for this new configuration. Nevertheless, gastight interconnects were not achieved, and this route was not further pursued.
Another interconnect design was based on the use of (La,Sr)(Co,Mn)O3 (LSCM) oxide and a glass seal. However, the results indicated leakages due to thermal expansion mismatches and this can remedied by adjusting compositions, this design was not pursued in ELECTRA.
All in all, the project has demonstrated one viable route for stacking in series three tubular segments. Increasing the number of cells would require more advanced engineering of jigs necessary to mount together the cells, interconnects and sealing components. It is deemed feasible, and this route is further investigated in other projects developing multi-segmented cells for a different chemical process.

GEN3 segmented-in-series cells by printing segments on a common inert support tube

GEN3 requires development of compatible porous support tube, cathode, dense electrolyte and interconnect, anode, and current collectors and contacts. A number of required materials were made at kg-scale for the development of GEN3 supported cathode electrolyte tubes. The major challenge in GEN3-type manufacturing is the co-sintering of various porous and dense layers. The development of this includes production of tubular porous BZCY supports, reliability in applying and co-sintering the NiO-BZCY electrode and BZCY electrolyte in series, developing and integrating novel interconnect materials, and applying anodes and current collectors. The work for GEN3 has in sum led to the successful first complete GEN3 3-segment tubular cells. These have however not yet been through significant testing.

WP2 Development of anodes and current collection

WP2 develops and deposits stable and efficient PCE anodes (H2O+O2 side) and current collection on tubular half-cells delivered by WP1. The requirements for PCE anodes are good transport properties, high catalytic activities and long-term chemical stability in steam electrolysis conditions. Moreover, the material for this steam electrode, specifically, must exhibit low ohmic, kinetic, and mass transport polarization. The target performance is area specific resistance (ASR) less than 0.2 Ωcm2 for the anode at 700ºC.

Development of anode

The anode materials development encompasses the optimization of the electrode materials (single phase or composite), i.e. electrical properties of the ionic and mixed conducting phases, catalytic properties, and microstructure. The experimental activities are supported by electrode modelling using empiric or phenomenological models. A number of candidate materials were identified and synthesised through a combustion process followed by a thermal treatment. They were then subjected to a deagglomeration step to reduce particles size and its distribution until reaching the targeted specification.
The first step in WP2 was screening a number of materials for their compatibility with BCZY based electrolyte and stability as PCE anode at 700ºC and 3 bar air and steam during 72 h. The screening rendered (La,Sr)MnO3 (LSM) and (Ba,Gd,La)CoO3 (BGLC) of various compositions as candidate material stable at high steam pressures.
Based on the materials screening and the manufacturing protocol, the following routes were pursued with respect to steam electrode design: Composite electrodes of LSM/BCZY; Composite electrodes of LSM/BCZY infiltrated with BGLC; Single phase electrodes of BGLC; BZCY backbones infiltrated with BGLC. In order to improve the catalytic activity, anodes are additionally infiltrated with catalytic nanoparticles, such as BGLC, Pr-O, CeO2, Pr-Ce-O, and ZrO2, resulting in significant improvements.
The development of optimised steam anodes has been successfully completed with focus on mechanical stability and low polarization resistance under running conditions. Composite BCZY-BGLC electrodes have been produced and tested, and a polarization resistance of 0.7 ohm cm2 is achieved for the combined anode and cathode at 700°C for a tube segment. The composite electrodes show high mechanical robustness and are shown to operate well also under higher current loads and high steam pressures.
The proton incorporation of the different BGLC compositions was studied using thermogravimetry. BGLC hydrates significantly at relevant operating temperatures thus extending the reaction zone from the triple phase boundary in proton ceramic cells. An aim of this project was to verify that the measured proton incorporation/hydration was indeed a bulk phenomenon, representing a structural protonic defect within BGLC. To investigate this, BGLC powders equilibrated in deuterated (D2O-wetted) air were taken to the Canadian Nuclear Laboratories to conduct neutron diffraction studies. The preliminary results represent the first unambiguous evidence of the presence of proton defects in the BGLC lattice.
The work furthermore showed that stability in steam is increased as more La is introduced into the structure. Detailed analysis and further development will be carried out in a spin-off MERA-NET project “GOPHY-MICO”, where advanced structural characterization of these double perovskites is a key part.
In trying the correlate physical and electrochemical properties of oxides, it is useful to develop a defect chemical model. Some attempts have been made in the literature for comparable double perovskites, but they have not adequately been able to describe and rationalize the electronic properties of the materials. Thus, a new defect chemical model was developed based on conductivity and oxygen stoichiometry measurements. This work has resulted in a paper recently published in the Journal of Materials Chemistry A.
BGLC steam anodes applied on tubular PCEs were tested at different steam pressures (e.g. 1.5 bar) at the anode side over a range of temperatures 400-700°C and bias currents. The electrode polarization resistance could be delineated into three different contributions; charge transfer, mass transfer, and gas transport, all of which were reduced upon increased anodic current.
Several materials for enhancing current collection of the anodes were investigated, including BGLC itself, Cu-Co-Mn spinels, and silver. These were stable and had conductivities near or above the targeted 300 S/cm.

Application of anode, current collectors and contacts

Optimization of electrodes
Single phase BGLC electrodes were prepared by mixing BGLC powder with an ink vehicle. Three tube segments were painted with three, five and seven layers with drying in heating cabinet between each coating step. The electrodes were fired in air at 1200°C for 10 h. The sintered electrode layers were tested for mechanical robustness by fastening and removing a rubber Scotch tape. All samples showed severe delamination. It was therefore concluded that the thermal expansion mismatch between BGLC and BZCY is too high to produce single-phase electrodes at this stage.
To overcome this, composite electrodes of 60 vol% BGLC and 40 vol% BZCY mixed with 0.5 wt% NiO sintering aid were prepared by ball-milling in isopropanol. The mixture was dried and dispersed in the ink vehicle. Before painting the composite layer, the tube surface was primed with NiO in isopropanol to further improve sinterability. Tube segments were painted with seven layers and fired in air at 1200°C for 10 h. After sintering, a thin layer of Pt ink was applied to all samples and fired at 1050°C for one hour. The Pt layer ensures good electrical contact and current collection during testing in WP3.
Composites of LSM (60 vol%) and BZCY (40 vol%) (LSM-BZCY) electrodes were sintered on unreduced tubes, and can be later reduced and sealed, under controlled gas atmospheres. Manufacturing parameters were studied to achieve good adhesion and proper electrode microstructure. The protocol developed to produce these electrodes enables the manufacturing of porous electrodes with homogeneous distribution of the two phases.

Optimization of firing, sealing and reduction in dual atmosphere under electrical bias
The following experiences gained in firing the electrodes on PCEs form an important part of the outcome of ELECTRA.
The tubular cells were sealed to alumina risers and capped in the open end by use of a glass-ceramic sealing material. The thermal expansion mismatch between NiO and the sealing materials impose a certain restriction on the sealing procedure: Either, the tubular cells must be pre-reduced, or they must be sealed and reduced in one step without cooling in between. In this work, both procedures were applied. In the former, the full assembly of riser, lower sealing ring, reduced tube segment and sealing cap was mounted in a custom-made sealing jig in a ProboStat sample holder with an outer tube for controlled atmosphere and heated to 1000°C at 3°/min in 5% H2 in Ar. After dwelling for 1 hour at 1000°C, the cell was sealed, and the temperature lowered to RT at 3°/min.
In the one-step sealing and reduction procedure, un-reduced full cells were mounted with riser, lower sealing ring, un-reduced full cell, upper sealing ring and ceramic cap in a custom-made sealing jig. The assembly was heated to 1000°C in air where all parts sealed. The atmosphere in the inner compartment was subsequently switched to 5% H2 in Ar for 30 min and to 10% H2 in Ar for another 30 min. After 1 hour, the temperature was lowered to 800°C at 0.5°/min and left for 40 hours for the completion of the NiO to Ni reduction.
At high temperatures, BGLC requires a slightly oxidising atmosphere to maintain phase stability. The anodes may thus be sintered to an open, un-reduced tube segment in symmetric oxidising atmosphere at 1200°C, or to a sealed, pre-reduced tube segment to be fired at ≤ 1000°C in dual atmosphere. The thermal expansion of the sealing components at high temperatures imposes this maximum temperature for sintering of electrodes to pre-reduced tubes. In the one-step sealing and reduction procedure and in dual atmosphere anode sintering there is a large chemical gradient across the cell. Due to the significant oxygen transport in BZCY at high temperatures and dry oxidising conditions, pre-reduced cells faced Ni re-oxidation and the one-step sealing and reduction procedure was stalled.
To counteract unwanted oxygen transport, electrodes were connected to the anode and cathode and 1.5 V positive bias was applied across the cell during high temperature treatment in dual atmospheres. Monitoring current during the process gave control over the reduction, and the result was fully reduced, crack-free, and gas-tight cells.
Due to the high sintering temperature and better TEC compatibility, the anode layers from one-step reduction were the most robust and displayed the best adhesion to the electrolyte layer.
In the case of LSM/BCZY (60/40 wt. %), the tube is reduced at 700°C in wet H2 and then it is sealed at 1000°C in dry Ar using an alumina riser and a jig on the top. After sealing, it is infiltrated with catalytic nanoparticles of Pr-Ce-O 50% and painted with Ag or Au as a current collector to carry out the measurement in electrolysis and co-electrolysis mode (CSIC).
Despite the strict requirements imposed by the materials of the half-cells and the sealing used to close tubes and interconnection, different solutions have been found and consolidated, and form a robust base for the development of complete cells. The restrictions are related to the manufacturing parameters of the steam electrodes, i.e. gas environment and temperature range to achieve good adhesion and percolation while preserving the integrity of the sealing and the reduced Ni- BCZY support. As final outcome, a set of complete cells was delivered to WP3 for electrochemical testing.

Electrodes on GEN3 cells
The protocol developed to produce LSM-BZCY composite electrodes on tubular segment cells was also applied to GEN3 cells. A Teflon tape was used to mask the cells, which were dip-coated in the LSM-BZCY slurry. The coated tubes were then dried in air. The procedure was repeated 4 times. After final drying, the Teflon tape was removed and the electrode coated GEN3 cells were annealed in air at 1200°C for 5 h.

To conclude, WP2 developed new materials and architectures for steam electrode with promising electrochemical performance and stability under oxidizing and steam-rich conditions. State-of-the-art performance and stability was achieved with both single-phase and infiltrated novel electrodes. Current collectors were investigated for deposition on the anode on half-cell tubes. Successful processing of steam anodes on tubular cells was developed yielding ASR below 0.2 ohm cm2 at 700°C and high steam pressures.

WP3 Single tube module development and testing

WP3 designs and installs a rig for testing single tubes of all three generations at high steam pressures, tests all three generation tubes for performance and stability, and performs and evaluate test of co-electrolysis and for syngas production.

Single tube module design and construction

Common testing protocols in electrolysis and co-electrolysis modes were defined through meetings across the partner institutes. These protocols define and set parameters required for start-up, operation, and shut-down of the tubular cells.
Measurement setups with controlled gas supply, online gas analysis and potential for high steam pressures were designed, commissioned and installed for testing in electrolysis and co-electrolysis mode at high pressure. Upgraded ProboStat and custom sample holders with high current feedthroughs and high steam and temperature tolerance were designed, built and tested.

Testing of single tubes

Electrolysis measurements were performed in a special-made experimental reactor designed for measurements with high steam content (up to 15 bar H2O) and high total pressures (up to 50 bar total). Mass flow controllers and back pressure regulators control and adjust the gas flow rates and partial pressures on either compartment. The gas (and liquid water) is fed through an evaporator before it enters a hot-box which can be heated up to 200°C. A three-zone furnace and gas distribution lines are placed inside the hot-box, to ensure proper and uniform heat throughout the measurement system – avoiding condensation of steam due to cold spots. Exhaust gas is passed through a condenser to extract excess water before the gas is further passed either into a GC gas analyser or vented.
Two different sample holders were used to measure the single cells: A modified ProboStat with a base unit made to tolerate temperatures up to 165°C and pressures up to 10 bar was employed for most of the experiments. An in-house built measurement system based on steel tubing and Swagelok fittings was also employed to enable higher steam pressures.
The tests comprised single-segment cells, capped and sealed towards an alumina riser using in-house glass ceramic sealant. BGLC was brush-painted onto a reduced and sealed segment, and subsequently fired in dual atmosphere of 2% O2 on the anode and 5% H2 on the cathode side, with Ar as the balance gas in both cases. Moreover, a constant potential of 1.4 V was kept over the cell during the firing procedure to avoid any oxygen permeation during the annealing step at 1000°C, as had been observed previously when fired in open-circuit conditions.
The cell was characterized using gold paint and mesh as current collector. Polarization curves were taken at different temperatures with 1.5 bar H2O and 80 mbar O2 on the anode side, and 0.3 bar H2 on the cathode side. The total pressure was 3.5 bar with Ar as balance in both mixtures. As expected, the performance was enhanced with increasing temperature, with lower activation overpotentials. It also noteworthy that the total resistance over the cell – calculated from the slope of the I-V curve – continues to decrease also after the initial electrode activation. This can – in part – be attributed to a decreased resistance of the parallel electronic rail with higher applied potentials over the electrolyte. It may seem that parasitic ohmic resistances related to current collection or contact resistances through the measurement setup may still contribute significantly to the overall single cell resistance.
The faradaic efficiencies of the cell, calculated from the measured and theoretical hydrogen flow for a given current, decreases with increasing overpotential, in accordance with an increased electronic leakage current at higher potentials.
Following the study on single-phase BGLC electrodes coated on sealed and reduced half-cells of Ni-BZCY a set of cells were made by depositing composite BZCY-BGLC electrodes unreduced half-cells and annealing at 1200°C in air. Both Pt and Ag were used as current collectors to investigate the effect of different metallic current collectors on the cell performance. The composite electrodes display considerably lower electrode impedance than the single phase BGLC electrodes. Post-measurement inspections showed delamination of the single phase electrode layer, and the larger impedance for this cell is thus attributed to poor adhesion and limited reaction zone (triple-phase boundary length) between the electrode and electrolyte resulting from TEC-mismatch between BGLC and the Ni/BZCY support tube.
Hence, implementing a steam electrode like BGLC with activation energy of only 0.5 eV for water oxidation enables a lower total activation energy for the protonic rail as compared to the electronic, and allows higher electrical efficiencies at lower temperatures. I-V-characteristics and hydrogen production rates at 500, 600 and 700°C showed high faradaic efficiencies (> 85%) and the highest reported hydrogen production rate for a PCE (17 mL∙min-1), highlighting the importance of a good electrode to promote ionic transport and minimize electronic leakage currents in steam electrolysers.
PCEs are beneficially affected by increased steam pressure, as seen by increasing it from 1.5 bar to 4 bar steam on the anode side.
Impedance sweeps at a slight electrolytic bias before and after three days of operation shows good stability at 700°C and 1.5 bars of steam in the anode compartment. After increasing the steam pressure to 4 bars, electrolyte resistance decreases from 2.0 to 1.8 Ω∙cm2, indicating that the electrolyte is not fully hydrated at 1.5 bars of steam. This also highlights the benefit of PCE operation, where operation at higher total pressures without compromising cell performance – as can be observed in conventional electrolysers – enables production of pressurised hydrogen directly in the cathode compartment without dilution by unreacted steam (SOE) or water drag (PEM).
Continued durability tests at 600°C with 1.5 bar steam were performed with both 2 wires and 4 points measurements, the latter giving better results as expected. After one day of operation, we observed a significant increase in the polarization resistance and a decreased hydrogen evolution. Characterization of the cell after this test indicated a phase separation of the BGLC anode, which may be responsible for the degradation of the cell. We may conclude that the most vulnerable component for PCE durability is the anode, and the BGLC composition and processing used in the cells tested may need further optimisation in future research projects. This may comprise especially the La content and the interdiffusion during sintering.

Co-electrolysis

Co-electrolysis of steam and CO2 was performed with a thick dense BZCY electrolyte. An LSM-BZCY composite infiltrated with Pr and Ce oxide catalytic nanoparticles was used as anode. A porous Pt layer was used as cathode. Cell performance was evaluated by means of voltammetry and impedance spectroscopy. The performance of the cell was studied as a function of temperature ranging from 700 to 600ºC, Steam concentration in the anode chamber from 3% to 7.5 % H2O, H2 concentration in the cathode chamber, and CO2 concentration in the cathode chamber. Performance of the cell in both fuel cell and electrolysis mode increases when CO2 is added to the cathode stream. Thus, the co-electrolysis process gives rise to lower overpotentials that allow working at higher current densities and subsequently, an improvement of the Faradaic efficiency is expected. The products obtained in the co-electrolysis were quantified by gas chromatography, confirming the expected reduction to CO and H2.

WP4 Process integration, system design, and evaluation

The objectives of WP4 are to evaluate efficiency and Balance of Plant (BoP) of a PCE integrated in various scenarios involving supply of electricity, heat, and steam, CO2, and uses of hydrogen, to assess system design parameters for operation and HSE; and to design a PCE system based on inputs from tube characteristics (WP3), process, and HSE analyses.

System Design

A 100 MW molten salt tower-type solar-thermal plant fully dedicated to hydrogen production and integrated with a PCE unit has been analysed. In the endothermic mode (inlet temperature is higher than outlet temperature), heat is required for the reaction to take place and thus the operating temperature of the electrolyser will be reduced. For instance, an outlet temperature of 700°C may require an inlet temperature of 800°C. Thus, more electricity is required to achieve this inlet temperature if heat is not available. The heat available from the molten salt plant is at 565°C and this temperature is not sufficient to fulfil the heat requirement to obtain 800°C. Electric heating is to be used for superheating the inlet steam up to the required temperature.
If the operating temperature is set at 450°C, then an external stream at 565°C can be utilized to operate in an endothermic mode. Heat is required for superheating the inlet steam to 545°C, available from 565°C heat stream from the molten salt plant. The electrolyser utilises the heat from a superheated stream (545°C) to maintain the operating temperature at 450°C as well as to operate below thermoneutral voltage. Thus, this reduces the electricity requirement by the electrolyser. Compared with high temperature operation, the heating element can be eliminated from the balance of plant (BoP) for superheating as well as reducing the electricity demand by the electrolyser.
Operation in exothermic mode is desirable when there is less heat available from external sources to utilize, but as a result the electricity demand is higher. Effect of steam utilization: Steam partial pressure will decrease with increase in steam utilization and this will result in an increase in the open circuit potential (according to the Nernst equation) and therefore resulting in a decrease in current density. Thus, the hydrogen production will decrease if the electric supply is constant.

Process operability and HSE analysis

Piping and instrument diagram (P&ID) was analysed further for its reliability and availability. Possible failures were identified and recommendations were made for a durable and reliable system. Based on those recommendations, P&ID diagram was updated. In order to achieve the highest availability, reliability and safety of the system, the following improvements have been implemented in the control system: Industrial network with ring topology; Safety PLC modules for classified areas; Steam flow rate measurement redundancy; Temperature interlocks avoiding steam condensation; Water pumps redundancy; Continuous monitoring of the hydrogen quality; Redundancy in critical parameters of the electrolyser.
Recommendations of system improvements (electrolyser and remaining equipment separately) were made based on HSE investigations.

Techno-economic evaluation of integrated processes

Techno-economic evaluation of 1,35 MW and 100 MW electrolysers integrated with solar-thermal and geothermal sources of heat, steam, and electricity were made are under publication. The relatively low capital cost and high electrical efficiency allow the PCEs to be competitive to other electrolyser technologies under certain conditions and assumptions.
The cost for co-electrolysis of CO2 and steam to form syngas using PCEs is calculated at different spot electricity prices in the range 21-53 € / MWh. The co-electrolysis cannot currently compete with other production methods based on fossil fuels on price. The technology does however have a competitive advantage in terms of being able to generate syngas from water and CO2 without utilizing any fossil fuels. The CO2 can potentially come from industrial flue gas and if the electricity were to be generated by renewable energy the process would be considerably more environment friendly than production using steam methane reforming.
Life cycle analyses (LCAs) of the processes’ carbon footprint indicate that a significant reduction can be achieved compared to non-renewable hydrogen production. The emission of CO2 is primarily due to the production of electricity required for the process. The emission values range from 4.3-28.7 kg CO2-e/t product, which are similar to comparable electrolysis processes and much smaller than fossil based hydrogen production. The co-electrolysis process has two main advantages. Firstly, it has lower emission than steam reforming, which is the main syngas production method. Secondly, the CO2 is reused rather than emitting carbon bound in fossil fuels. The emission efficiency is also increased significantly if oxygen produced is taken as a product rather than being vented.

WP5 Operational performance of multi-tube module

The objectives of WP5 are to define a common testing and operation protocol for ProboStat and multi-tubular (Tubinder) modules, to construct and commission a 1 kW-size multitubular module for steam electrolysis, and to test, optimise, and evaluate it.

Definition of testing protocols and durability tests

Based on standards and design, the following protocols were defined:
Protocol for pre-commissioning includes mechanical equipment, pipelines, pressure vessels, tanks and instrumentation. The protocol describes how number of detailed issues will be verified for all the systems and sub-systems. If any of these aspects are not achieved, this will be included in the parts list. After that, all the faults will be solved and the protocol will be repeated. The pre-commissioning protocol will be iterative until all the points included in the protocols can be successfully confirmed.
Protocol for commissioning describes how it will be checked that each system works properly through several testing protocols. After that, all the faults will be solved and the protocol will be iterative. The commissioning will be repeated until all the points included in the protocols can be successfully confirmed. The commissioning protocol for the pressurised module itself involves: Leak test (Air – He); Pressure (1 – 20 bar): Vessel – individual tubes and fittings; Cold and hot operation; Plugging checks; Evaporator operation combined with multi-tube module checks; Leak test; Check corrosion including fittings; Heating the module (without current); Control T; Temperature gradient; Optimisation, Electrochemistry; Electrical connection; Shortcuts between tubes and vessel; First operation with 1, 2, 5 tubes increasing the current stepwise.
Protocol for start-up, shutdown and restart the test bench: Start-up involves warm-up procedure and sweep gas and water feeding procedure. Summarizing, the tests proposed are the following: start-up sequence, normal shut-down sequence and emergency shut-down sequence.
Protocol for operation: In this protocol all the tests required to validate the proper operation of the system are defined, in order to evaluate the behaviour of the system considering the established specifications. These functional tests will be able to establish the basic recommendations related to operational strategy.
Protocol for durability test: The objective of durability tests is to operate the system so as to record a significant degradation. Several operational parameters and cycles have to be tested in order to observe their impact on the cell behaviour. Steady state current should apply for more hours before and after transient tests in order to measure the cell degradation at reference stable operation. Transient tests included rapid switches between very low current and operating current in order to avoid temperature variations and analyse specifically electrical transients.
For the elaboration of all of these test protocols, extensive research has been carried out, including analysis of information regarding more conventional solid oxide electrolysers.

Multi-tube module construction and commissioning

A compact size electrolyser unit with power density over 20 Wdm-3 has been designed, assembled and tested up to 20 bar of pressure. It is suitable for operation up to 700 ºC. The electrolyser unit is a cylindrical vessel that contains 18 sockets for tubular cells. The multi-tube module is a concept capable of converting electric energy to fuels by producing dry, pressurised hydrogen during steam electrolysis reaction. The system enables quick insertion of up to 18 tubular electrolysis cells (<30 cm length) from the top socket-board. Manifolding and gas connections enable, when required, the sampling and/or shutdown of individual cells. The design and construction of the electrolyser provides a flexible system suitable for testing many types of tubular electrochemical cells. It is considered a major achievement of the project, and involved partners are engaged in patenting it.
In order to the keep socket plate of the module at temperature below 200°C a liquid cooling system has been used. Based on CFD simulation 100 W of heat has to be dissipated.
In the electrolysis process, a mixture of steam and air is utilized. For safety reason inert gas (nitrogen, argon) is pumping during start and stop procedures in order to remove reactive gases (hydrogen, oxygen). In order to investigate electrochemical properties of proton conducting ceramic argon as a sweeping gas can be used as well.
A power supply is used to supply the 18 tubes containing up to 5 segments in series. This power supply provides maximum 8 kW of electric power, with voltage level 20 V and up to 440 A of current.
In order to evaluate faradaic efficiency during steam electrolysis as well as co electrolysis of CO2 reactions and detect possible gas leakages, a micro gas chromatograph is employed to analyse the composition of the gas produced in the cathode.
The first stage of the hydraulic test was leak examination and it was carried out by supplying nitrogen at 50 bar to each part of the multi-tube module individually at room temperature. After positive result the module has been assembled, sealed with suitable custom gaskets and tested up to 20 bar of nitrogen at room temperature. Note that individual welded components passed a test up to 50 bar of nitrogen at room temperature The fully-assembled module (recipient + tube cells) has passed a series of quality tests: pressure resistance (20 bar) of the recipient, leakage test to ensure appropriate joining of the tube cells in the module as well at room temperature as up to 700°C.
Thermal studies were carried out in the fully-assembled electrolyser module in order to evaluate thermal profile along all ceramic tubes as well as comparison isothermal zone with CFD simulation results. The temperature profiles were recorded using closed-end alumina tubes and therefore there was no generation of heat by Joule effect during the electrolysis operation. Heat is provided externally using ceramic electrical heater located around the cylindrical chamber. The temperature profile along the cell axis in these conditions is reasonable while the different between cell positions is within 10 ºC.
The module and periphericals were commissioned after verifying and checking electrical system, instruments, pipes and containers, local C/I system, and mechanical systems, according to the protocol defined above, and hence comprised Electricity (Verification of connections, actuation of switches and protections, energizing boards, powering cables); Instruments (Testing of control loops and safety valves, verification of signals, verification of interlocks and associated alarms); Pipelines and containers (checking connections and direction of flows, inertization, leak test, local C/I system, operability, behaviour of different systems); Mechanical equipment (Functional test, performance, leaks, vibrations).

Testing and optimization of multi-tube module

A detailed test protocol was developed and optimised. Following this, the multi-tube electrolysis module has been tested and optimized showing 1) proper thermal behavior and heat/steam management and integration; 2) good accessibility for mounting and dismounting electrolysis cells and other components from the top socket-board; 3) quick operation for replacing cells, during hot operation, e.g. it is possible to shut-down individual cells (closing gases); 4) good H2 gas tightness in the overall H2 ducts and cathode chambers. Although the system was ready for complete electrolysis operation, the test with several electrolysis units to produce hydrogen was not possible to the lack of fully-assembled tubular cells and severe delays during commissioning and optimization. These circumstances were caused by unexpected problems in welded components (800HT special alloy), malfunction of power controllers, accidental failure of cells and heating furnaces. Nevertheless, the system is fully operative and the 1 kW-electrolysis demonstration is being demonstrated beyond the timeframe of ELECTRA.
Potential Impact:
Impact

High temperature electrolysers (HTEs) offer an ideal solution for medium- and large-scale centralised hydrogen production for transport, grid-balancing, or industrial use when combined with renewable electricity and sources of heat and/or steam. Further it is possible to utilise HTEs to electrolyze CO2 and steam into syngas mixture used to synthesize valuable green synthetic fuels and synthetic petrochemical products. Proton ceramic electrolysers (PCEs) can do these with certain advantages in terms of simplicity, efficiency, and safety.
ELECTRA has demonstrated a series of advances towards realisation of these possibilities, namely fabrication of segmented-in-series tubular cells and multitubular module technology, improved steam+oxygen-side electrode materials, and efficient operation at over-atmospheric steam pressures. Techno-economic analyses performed suggest that proton ceramics can be competitive for industrial production of hydrogen based on supplies of waste or renewable heat or steam and low electricity prices provided the tubular cells can enter mass production.
In particular, novel tubular units based on the progress and experiences in ELECTRA are foreseen to be competitive at small and medium scale in a relatively short timeframe because of modularity, ease of fabrication, and cheap materials. This is pursued in a follow-up project for 10 kW electrolyser at 30 bar hydrogen for industrial use under the FCH scheme by some of the ELECTRA partners and additional new industry partners.
It may hence be expected that an operating tubular PCE system will materialize in the next phase (3-5 years) on a small, e.g. 10 kW scale. In the following years, this may be scaled up to mass production for small and medium scale and serve certain markets for hydrogen production competitively. If these efforts prove successful, the inherent advantages of PCEs may form basis for large scale THE hydrogen production based on availability of steam, and heat, and electricity. The techno-economic analyses of ELECTRA exemplify the technical and economic aspects of integration of PCEs with solar-thermal and geothermal plants.

Dissemination

ELECTRA dissemination activities include organisation of a summer school, organisation and participation in international workshops on electrolysers, and regular talks and posters at national and international conferences as well as peer reviewed scientific papers and various outreach activities.
A Summer School with ELECTRA scientists and invited speakers with international leading expertise was organised in Valencia on Sept. 23-25th 2015 in collaboration with 2 other EU-projects HETMOC and GREENCC. It lasted 3 days with scientific lectures and a social program. It gathered 120 researchers coming from 5 continents and 16 different countries (6 EU member states) and the 5 ELECTRA partners to share knowledge and exchange ideas across the multiple scientific fields involved in the world of ionic and protonic ceramic membranes. A total of 31 oral presentations and 47 posters including both industries and academia offered a wide area for fruitful and interactive discussions. ELECTRA project was presented through 3 oral and 3 posters. ELECTRA partners CMS and CRI took part in the industry panel discussion.
ELECTRA organised a HTE session in the international workshop IDHEA 2016 in Nantes on Nov. 02-04th 2016 dedicated to "High Temperature Electrolysers". The ELECTRA project was presented through 1 keynote and several oral presentations.
ELECTRA has otherwise presented the project concept, objectives, and results at several European and International scientific conferences and workshops with contributed oral presentations, keynote lecture, and posters. It has also been presented at cutting edge innovation events and in newspapers, in those connections advertised through social media, e.g. Facebook. Altogether the project has at the end of the financing period published 3 peer-reviewed scientific articles in international journals and have 45 presentations of various kind, from invited scientific lectures to industry- and society-oriented outreach activities. In addition we have publicly available videos and other media. More papers are in the pipeline, and more presentations will be given onwards.

Exploitation
The innovation and exploitation (I&E) board of ELECTRA has met during project meetings throughout the project. In the last part the work of the I&E board changed into working mainly on the Exploitation Plan deliverable of the project. Main exploitation activities have comprised the items below.
The project has prepared a template to collect the relevant information related with exploitable results. This has been the basis for preparation of the Exploitation Plan, which is a deliverable of the project.
Two of the partners are working on IPR (patent application) for the multitubular module design.
The sealing technologies gained experience with in ELECTRA is taken further in other projects.
Follow-up project application. Some of the partners – together with additional industries - have submitted and been awarded what can be considered a follow-up project of ELECTRA. The new “GAMER” project takes the progress of ELECTRA, combined with the identified needs for technical progress and the techno-economic analysis and provides a concept that meets the call in FCH for 10 kW 30 bar HTEs for industry with supply of heat or steam and need for hydrogen. The concept is feasible also for certain other applications of hydrogen under given assumptions.
List of Websites:
Project public website: http://www.mn.uio.no/smn/english/research/projects/chemistry/electra/index.html

Coordinator: Prof. Truls Norby, University of Oslo. Phone: +47-22840654. Mobile: +47-99257611.

Final Report Summary - ELECTRA (High temperature electrolyser with novel proton ceramic tubular modules of superior efficiency, robustness, and lifetime economy)

Related documents

Share this page

Download