Final Report Summary - SOCTESQA (Solid Oxide Cell and Stack Testing, Safety and Quality Assurance)
Executive Summary:
The aim of the project is to develop uniform and industry wide test modules and programs for solid oxide cell and stack (SOC) assembly units. New application fields which are based on the operation of the SOC cell/stack assembly in fuel cell (SOFC), in electrolysis (SOEC) and in combined SOFC/SOEC mode are addressed. This covers the wide field of energy conversion systems, e.g. stationary SOFC μ-CHP, mobile SOFC APU, stationary SOEC systems (power-to-gas) and combined SOFC/SOEC (power-to-gas-to-power) systems. The project partners are DLR (coordinator, Germany), CEA (France), DTU (Denmark), ENEA (Italy), JRC (Belgium), EIFER (Germany) and NTU (Singapore). During the project a close interaction with an industrial advisory board (IAB) ensured to achieve industrial relevant outcome of the project. Additionally, a continuous liaison with standards developing organizations (SDOs) was established in order to implement the outcome of the project successfully into international standards.
Altogether 11 test modules were developed in the project, which are TM 00: General SOC testing guidelines, TM 02: Start-up, TM 03: Current-voltage characteristics, TM 04: Electrochemical impedance spectroscopy, TM 07: Reactant utilization, TM 08: Reactant gas composition, TM 09: Temperature sensitivity, TM 12: Operation under constant current, TM 13: Operation under varying current, TM 14: Thermal cycling and TM 16: Shut-down. These test modules were experimentally validated by the project partners in several testing campaigns with 5 test programs (TPs) for the different applications.
For the development of the test modules several steps were performed. At the beginning of the project all important specifications, especially of the interfaces between the test station and test object and the nomenclatures, were defined. For the validation of the test modules short stacks with 5 repeat units from ElringKlinger were used. In the first testing campaign the test stations were modified and harmonized in order to provide reliable and reproducible interfaces between the stacks and the test stations, to achieve operating conditions as similar as possible among all partners and to follow dynamic profiles. In the following two testing campaigns the test modules were validated and optimized with application-specific test programmes for SOFC, SOEC and combined SOFC/SOEC applications. In this context, technical input for the operating conditions of SOC systems from relevant industrial stakeholders were integrated in the test programmes. The validation process mainly addressed the quality, reproducibility and repeatability of the results between the partners and between the different test modules. The optimization of the test modules included thermal aspects like stack temperature, gas temperatures, temperature homogeneity inside the furnace, instabilities of the fuel gas humidification unit and electrical wiring setup for electrochemical impedance spectroscopy. Additionally, aspects for the reproducible determination of degradation rates were implemented in the long term operation test modules. Moreover, recommendations for the reliable calculation of derived quantities, e.g. area-specific resistance (ASR), reactant utilizations and low and high frequency resistances, were included in the test modules. At the end of the project the final round robin test was performed, which showed a high quality, reproducibility and repeatability of the results of the different test laboratories. Hence, the robustness of the optimized test modules was confirmed.
The exploitation of the project mainly focused on the transfer of the results to standards developing organizations, e.g. the International Electrotechnical Commission (IEC) and CEN/CENELEC, the main European SDO in the field. A formal liaison of the SOCTESQA consortium with both entities was established. In case of IEC a new working group (WG 13) was initiated within the technical committee TC105. The entire SOCTESQA consortium participated in and guided the activities, leading to the publication of Standard 62282-8-101: “Energy storage systems using fuel cell modules in reverse mode - Test procedures for solid oxide single cell and stack performance including operation in reverse mode”. Moreover, the project results have been broadly disseminated to general and scientific public, e.g. by scientific papers, conference contributions, workshops, fairs and the SOCTESQA website (http://www.soctesqa.eu). The results and knowledge generated in the SOCTESQA project have led to a straightforward and open approach for the universal adoption of harmonized, quality-assuring procedures, which is the key for the successful technological progress and market implementation of the ceramic solid oxide technology.
Project Context and Objectives:
The aim of the project is to develop uniform and industry wide test modules and programs for solid oxide cell and stack (SOC) assembly units. New application fields which are based on the operation of the SOC cell/stack assembly in fuel cell (SOFC), in electrolysis (SOEC) and in combined SOFC/SOEC mode are addressed. This covers the wide field of power conversion systems, e.g. stationary SOFC μ-CHP, mobile SOFC APU, stationary SOEC systems (power-to-gas) and combined SOFC/SOEC (power-to-gas-to-power) systems.
The project builds on the experience and the methodology gained in previous European projects, e.g. “FCTESTNET” and “FCTESQA”. However, these projects focused mainly on single cell tests and system level tests under steady state conditions. On the other hand other relevant stack projects, e.g. the “STACKTEST” project, do not address the high temperature fuel cell technology. None of these projects focuses on the development of test procedures of solid oxide cell/stack systems under dynamic operating conditions or in the electrolysis mode. Moreover, established advanced characterization techniques, e.g. the electrochemical impedance spectroscopy (EIS), have not yet been integrated in the test protocols of these previous projects. Only few results were validated by round robin tests.
Figure 1 shows a complete solid oxide cell/stack testing system with the corresponding control subsystems, which is very complex. Numerous test input parameters have a high impact on the test object output results. Moreover, a clear definition and specification of the interfaces between test object and test station is obligatory. Especially for high temperature solid oxide systems the temperature control subsystem and the gas control subsystems play an important role for the test output results. Moreover, the complexity of the test system is significantly increased when the operation mode is transferred from steady state operation to dynamic operating conditions. It is clear that detailed test schemes, procedures and protocols are essentially necessary for the development of the SOC cell/stack assembly unit.
Figure 1: Schematic graph of a test system for high temperature solid oxide assembly unit
The objective of the present project therefore concentrates on the development of uniform test procedures for high temperature ceramic solid oxide cell/stack assembly units both in the fuel cell and in the electrolysis mode. It is intended to develop a full set of application specific test modules and programs addressing function, performance, durability and degradation. These test procedures and programs usually consist of a combination of different test modules. The test procedures address three different operation modes, which are solid oxide fuel cell (SOFC), solid oxide electrolysis cell (SOEC) and combined SOFC/SOEC operation. Each of these cases includes both steady state and dynamic operations.
Within the project, the test modules and test programs are established and experimentally validated. The test procedures are validated on stack relevant test specimens, e.g. on short stacks with 5 cells. The test procedures include and specify the complete testing system, the different operating modes, the test conditions and the electrochemical characterisation methods, e.g. current-voltage curves, long-term operation and electrochemical impedance spectroscopy. The test modules and programs consider all relevant operating conditions of the complete SOC test system, e.g. input gases, pressure levels, temperatures, current loads, voltage level, mechanical loads, dynamic transients etc. In order to systematically study these effects, a sound understanding of the interaction of the test object with the test station is developed.
For the SOCTESQA project an organization and management structure has been built that suits the multitude of activities. In this project six European partners and one non-European partner are involved. These are DLR (coordinator, Germany), CEA (France), DTU (Denmark), ENEA (Italy), JRC (Belgium), EIFER (Germany) and NTU (Singapore). During the project a close interaction with an industrial advisory board (IAB) ensures to achieve industrial relevant outcome of the project. Additionally, a continuous liaison with standards developing organizations (SDOs) is aspired with the aim to implement the outcome of the project successfully into international standards. The activities are organized in work packages (WP), which are coordinated by work package leaders. These are WP 1: Coordination (DLR), WP 2: Specifications and Procurement (JRC), WP 3: Testing Procedures (EIFER), WP 4: Solid Oxide Fuel Cell (DLR), WP 5: Solid Oxide Electrolysis Cell (CEA), WP 6: Combined SOFC/SOEC (DTU), WP 7: Dissemination and Liaison (ENEA). Each work package contains several tasks, which are coordinated by task leaders. In this project the work package leaders are also the leaders of the corresponding tasks in their WP.
For the achievement of the project objectives, two workflow paths have to be taken into account. The first path is the development of the test modules and programs and the second path refers to the liaison activities to industry and to standard development organizations (SDO). Figure 2 show the workflow diagram for the project.
Figure 2: Workflow diagram for the project
The first step of the project concerns the specification of all relevant terms, definitions and terminology. After that a test matrix with all test modules for the SOFC, the SOEC and the combined SOFC/SOEC operation modes is developed. It is important that the test matrix reveals and confirms the industry needs from the very beginning of the project. This test matrix and the available test facility characteristics are the basis for the drafting and definition of the generic test modules and programs. In parallel the test specimens are defined and procured from solid oxide cell/stack manufacturing stakeholders. The specific test programs are defined by integration of the application specific conditions into the generic test modules. After this initial definition stage altogether 4 testing campaigns were performed. The first test bench validation campaign among the testing project partners figures out possible test station differences. For this purpose, the most critical test modules are selected. The review of this test bench validation focuses on the harmonization of the test benches by minimization of the possible differences, and the subsequent integration of any minor differences in equipment within the text of the test-specific procedures.
In the second testing campaign the specific test programs are applied to the SOC cell/stack assembly. The results of the specific test programs are the basis for the validation of the generic test modules. The test modules are reviewed and modified based on experimental results. In this context the repeatability of several measurements, the reproducibility of the different partners and methods are mainly addressed. Other issues are a sensitivity analysis of the test input parameters and the calculation of derived quantities. Establishment and validation of the generic test modules shall observe compliance with the applied specific test programs and consider any relevant feedback from these activities. With the modified test modules a third validation and review campaign takes place. After that a final round robin test is performed and reviewed to ensure comparability of the results obtained in different test laboratories. It is planned that each partner tests about 6 cell/stack assembly units during the 4 testing campaigns.
The second path refers to the liaison activities to industry and to standard development organizations (SDO). It is intended to integrate several independent SOFC and SOEC relevant industrial companies forming an industrial advisory board (IAB) in the overall project consortium instead to involve individual companies as project partners. Several important companies of the solid oxide cell technology have agreed and committed to support this project. This ensures a broad industrial advisory impact from different SOFC and SOEC stakeholders, e.g. material suppliers, cell and stack manufacturers or system developers on the project. Moreover, the test procedures can be developed independently from any possible individual industrial stakeholder`s interests thus enabling their broad international acceptance after the project. Since fuel cell testing is an integral part of international standardization activities, liaison with international standardization activities is integrated in the project and established in a separate work package. Contacts with the IEC TC105 responsible for the IEC 62282 family of standards as well as relations with the ISO TC 197 and other relevant working groups are considered appropriate in this context and the existing reciprocal influence will be enhanced. The output of the project are standardized and industry wide test procedures and protocols according to the different system applications.
Project Results:
The present section summarizes the results, which have been achieved during the three years of SOCTESQA project. For each work package (WP) the work progress and the main achievements are summarized according to the different tasks of the description of work (DoW). The relevant deliverables are referred.
1 Specifications and Procurement (WP 2)
This work package focuses on the identification and definition of all important specifications at the beginning of the project, the review of already existing test procedures and the procurement of testing specimen.
1.1 Definition of SOC specifications
At the beginning of the project all important specifications were identified and defined in work package 2. This includes the specification of the SOC test station, of the interfaces between the test station and the cell/stack assembly unit, of the testing system environment and of the nomenclatures. Figure 3 shows the corresponding interfaces between test station and short stack with the relevant test input and output parameters. Moreover, all the specifications for the solid oxide cell/stack assembly unit (SOC short stack) have been defined in cooperation with the potential stack suppliers. These specifications are compiled in the deliverable D 2.1 (List of SOC Specifications).
Figure 3: Interfaces between test station and short stack
1.2 Compilation of existent SOC test procedures
In the second task all existent SOC test procedures for single cells and stacks including those from previous projects and from the open literature were surveyed and compiled. The results were compiled in the deliverable D 2.2 (List of SOC test procedures). The definitions & terminology for use in the test procedures are taken from the IEC document (IEC 62282-1 TS: “Fuel Cell technologies Part 1: Terminology”) where needed supplemented by the FCTESTNET Glossary to avoid repetition and to preserve consistency with international standards. Definitions missing in these documents were defined by the consortium during the drafting of the test procedures. All definitions & terminology are listed in the SOCTESQA master document (Test module TM 00: General testing guidelines). The investigated literature already offers a number of tests modules for performance, reactant gas compositions, lambda sensitivity and cell/stack endurance whereas other modules especially for dynamic operation can hardly be found. Figure 4 shows the number of literature references for the different test modules. Most of the modules refer only to SOFC testing. The lack of SOEC test modules showed that the development of test programs for electrolysis mode needed the most effort within the present project.
Figure 4: Number of literature documents for the different test modules
1.3 Procurement of SOC materials and test samples
In this task the SOC short stacks intended for the validation of the test procedures and the round robin test campaign were procured. Several technical discussions with potential stack suppliers (ElringKlinger, Sunfire and SolidPower) were necessary to select the most suitable one. By weighing all pros and cons the partners decided to use short stacks with 5 repeat units from ElringKlinger for the validation of the test modules. The corresponding stack design as shown in Figure 5 required the lowest effort to be integrated in the test stations of the different partners.
Figure 5: Short stack with 5 repeat units from ElringKlinger used in the SOCTESQA project
After having organized all administrative issues the stacks for the first three testing campaigns were procured from ElringKlinger. The details of the procurements and deliveries are summarized in deliverables D 2.3 D 2.4 D 2.5 and D 2.6. For the round robin test no further stack procurement was necessary because most partners still had stacks from previous campaigns in stock, which were not yet tested. Stack shipment between different partners was organized. In this way, all partners managed to complete the full test schedule without stack redundancy.
2 Test Procedures (WP 3)
In work package 3 generic harmonized test modules and programs have been developed and elaborated for the cell/stack assembly unit. The test procedures are defined for both steady-state and dynamic operation conditions for SOFC, SOEC and combined SOFC/SOEC systems.
2.1 Test matrix definition
A test matrix (see Table 1) containing altogether 18 test modules which are relevant for four major applications has been defined. The four applications identified are:
• SOFC for micro combined heat and power (μ-CHP) and distributed power generation
• SOFC for auxiliary power unit (APU)
• SOEC for hydrogen production (power-to-gas)
• Combined SOFC/SOEC for electricity storage (power-to-gas-to-power)
In the SOCTESQA project altogether 10 test modules (bold in Table 1) and the general test module TM 00 (General SOC testing Guidelines) were developed. In order to make sure that the test matrix reflects the needs for the final applications, industrial stakeholders who are developing SOFC/SOEC products have been contacted to collect information regarding the required operation modes during the lifecycle of the product for each application. Feedbacks from 7 industrial stake holders have been received and taken into account.
Table 1: Test matrix for application-specific testing
2.2 Test procedures for initial test bench validation
In order to validate the applicability of test procedures defined in the present project, tests of solid oxide cells/stacks have to be performed in different laboratories and results should then be compared. It is therefore necessary to harmonize the specifications of test stations among different institutions. The purpose was to make sure that the facilities and the test object interfaces in all testing partners’ laboratories are suitable for the test procedure evaluation. All project partners have adapted their test stations accordingly in order to realize nearly identical operating conditions of stacks. For this purpose a test program has been developed.
In this test program, all three operation modes of the SOC stack (SOFC, SOEC and combined SOFC/SOEC) have been considered. Operating parameters have been defined by taking into account the stack specifications supplied by the stack manufacturer as well as constraints coming from test stations. Five generic test modules have been created and embedded in the test program. These are:
• TM 02: Start-up
• TM 03: Current-voltage characteristics
• TM 04: Electrochemical impedance spectroscopy
• TM 12: Operation under constant current
• TM 16: Shut-down
Figure 6 shows exemplarily the structure and the content of the test module TM 03 (Current voltage characteristics). A test module contains all relevant information and guidelines for the reliable testing of the SOC test object, e.g. the objective and scope of the test, all test input and output parameters, the formula for the derived quantities. Moreover, the test procedure and the data processing and presentation are described in detail.
Figure 6: Simplified example of test module: TM03 (Current-voltage characteristics)
In addition, a lot of effort has been dedicated to the creation of the master document TM 00 (General SOC Testing Guidelines), which contains common issues applicable for all test modules defined in the test matrix. Among others the following issues are addressed:
• General safety aspects
• Description of test object (cell and stack)
• Description of test system
• Interface between test object and test system
• List of quantities which are important to be controlled / measured
• Reporting of testing results
2.3 Test procedures for first validation
In this task the 5 test modules of the test station validation (TM 00: General SOC testing guidelines, TM 02: Start-up, TM 03: Current-voltage characteristics, TM 04: Electrochemical impedance spectroscopy, TM 12: Operation under constant current and TM 16: Shut-down) were optimized. Moreover, another 4 test modules have been drafted. They are TM 07: Reactant utilization, TM 08: Reactant gas composition, TM 09: Temperature sensitivity and TM 13: Operation under varying current.
In addition, three application-specific test programs (TP01, TP03 and TP04) have been defined, which have been used by the project consortium in the first test campaign. These are:
• TP01: Test program for the stationary application of SOFC in a µ-CHP system
• TP03: Test program for the application of SOEC in a power-to-gas system
• TP04: Test program for the application of combined SOFC/SOEC in a power-to-gas-to-power system
In these three test programs, all 9 test modules have been integrated. Figure 7 shows as an example the scheme of test program TP01 (Test program for the stationary application of SOFC in a micro-CHP system). This test program is a sequence of all important test modules for this system application. Moreover, some of the test modules were also run in SOEC mode in order to investigate their reproducibility in electrolysis mode (see below section 3: SOEC). All test procedures are documented in deliverable D 3.3 (Test procedure document for the first validation).
Figure 7: Test program for the stationary application of SOFC in a µ-CHP system
2.4 Test procedures for second validation
According to the results obtained during the first validation test campaign, test modules have been revised and optimized to a new version. In addition, a new test program and a new test module have been developed. They are:
• TP02: Test program for the mobile application of SOFC in an APU
• TM 14: Thermal cycling
TM 14 has been integrated into TP02. During the second validation test campaign, TP02, TP03 and TP04 have been applied by the project consortium. Deliverable D 3.4 (Test procedure document for the second validation) compiles all 11 test modules and 4 test programs.
2.5 Test procedures for final round robin test
Based on test results achieved during the second validation test campaign, all test modules and test programs have been reviewed and further optimised. In order to check the robustness of these test procedures, another test program (TP05) has been defined for the final round robin test, see deliverable D 3.5 (Test procedure document for the final Round Robin Test).
2.6 Final test protocols
The final round robin test results supplied the last feedbacks for optimising all 11 test modules developed in the project. After harmonising the format of all test modules, the final versions of test protocols have been worked out. In addition, a one-page summary has been created for each test module in order to provide a quick overview. Final test protocols have been compiled in the deliverable D 3.6 (Final document of test protocols) and are also available for the public on the SOCTESQA website.
3 SOFC (WP 4)
This work package aims at testing and validation of the cell/stack assembly unit in SOFC operation with the test procedures specified in WP 3. Recommendations based on the experimental validations are directly integrated into the improvement of the developed test modules.
3.1 Adaptation of test stations for SOFC testing
As the test facilities differ among the project partners, one of the first steps in WP 4 was a compilation of the test station specifications and a comparison with the specifications of different stacks and the test program requirements. As most of the test station furnaces of the partners are limited in geometrical dimensions it has been shown that the size of the stack will be a critical factor. In cooperation with WP 2 a suitable stack supplier (ElringKlinger) was chosen which fits the test stations of all partners. The results of this survey and the recommended modifications for each test station are documented in the deliverable D 4.1 (Specification list of test benches for SOFC operation).
According to the stack specifications and the aims of the project the test stations were adapted, particularly with regard to uniform test input parameters at the stack/test station interfaces. For the homogeneous gas supply to the stack an additional gas distribution plate had to be integrated in the test stations. This component was produced and customized specifically for each test station by an external supplier. In order to harmonize the adaptation and the integration of the stacks among the partners a manual called “Mounting of SOC-stack (EK) on gas distribution plate (GDP) in test station” was developed.
The adaptation and harmonization of the test stations among the partners have been continued during the testing campaigns. For instance two partners have been able to improve their gas preheating systems significantly after identifying demands during the initial test station validation. Moreover, the set-up for measuring electrochemical impedance spectra was modified by all project partners in order to minimize high frequency artefacts. The details of these test station optimizations are described in deliverable D 4.2 (Report of the initial SOFC test bench validation). Finally all test stations have been adapted successfully and have been able to perform the required test programs.
3.2 Results in SOFC mode
The following sections highlight the most important results for the validation of the test modules which were mainly obtained during the initial test station validation and in the first and the second testing campaign. The overall, detailed results are reported in deliverable D 4.3 (Report of the first SOFC testing campaign) and D 4.4 (Report of the second SOFC testing campaign).
Validation of TM 03: Current-voltage characteristics
The SOFC current-voltage characteristics of the stacks of the partners at 750°C, 50 % H2+50 % N2 // air are shown in Figure 8. The test results show a very high reproducibility between the different test laboratories. The OCV of all partners is almost the same and the voltage curves especially at lower current densities are very similar. However, at higher current densities small differences in stack performance among the partners exist.
Figure 8: Current voltage behavior of tested stacks of the different partners in SOFC mode at 750°C, 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU)
In the next step of development a complete analysis of the SOFC current-voltage results on the test input parameters was performed. This so-called sensitivity analysis qualifies the influence of the test input parameters on the performance and helps to better understand the differences in performance. Figure 9 shows as an example the stack power density at 500 mA/cm2 as a function of the average temperature of the fuel and oxidant inlet gases. A clear increase in performance with increase in gas inlet temperature can be observed. In contrast to that, nearly no influence of the performance from the measured temperatures at the top and bottom plate of the stack was observed. Hence the gas inlet temperature could be identified as one of the most important influences. The results obtained by this sensitivity analysis were used to optimize the test module TM 03 (Current-voltage characteristics).
Figure 9: Stack power density at 500 mA/cm2 in SOFC mode as a function of average temperature of the fuel and oxidant inlet gases
Validation of TM 04: Electrochemical impedance spectroscopy
Figure 10 shows the electrochemical impedance spectra of repeat unit 3 of the stacks of the different partners at 750°C, 12 mA/cm2 and 50 % H2+50 % N2 // air. The spectra show uniform results among the partners. Especially in the high-frequency range the spectra are almost similar which indicates a high reproducibility for the determination of the ohmic resistance of a repeat unit. However, often high-frequency disturbances make it difficult to determine the ohmic resistance of a repeat unit. This can be observed in the spectra of CEA. This issue was integrated in the optimized test module and solved by twisting the voltage probes to each other in order to minimize magnetic interference to each other.
Altogether three frequency dependent processes can be identified in the spectra, which are the polarization impedances of the anode and the cathode and the gas concentration impedance at the fuel gas side. In the low frequency range variations of the overall impedances among the partners can be observed, which is due to the high sensitivity of the gas concentration arc on the fuel gas composition. However, the total resistance of the electrochemical impedance spectrum correlates very well with the ASR calculated with the jV-curve, which proves a good reproducibility between both methods.
Figure 10: Electrochemical impedance spectra of repeat unit 3 of the stacks of the different partners in SOFC mode at 750°C, 12 mA/cm2 and 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU)
Validation of TM 07: Reactant utilization
The influence of the reactant utilization on the stack power density can be investigated in different ways which are described in TM 07 (Reactant utilisation). Figure 11 shows exemplarily the stack power densities as a function of the fuel gas utilization where the gas flow rates were kept constant at 0.5 H2 + 0.5 N2 NLPM/ RU on the negative electrode side and 4 NLPM/RU of air on the positive electrode side at 750°C. The current was increased stepwise from 0.3 A/cm² to 0.5 A/cm² to 0.6 A/cm² and to 0.7 A/cm².
Figure 11: Stack power density at different current densities as a function of fuel gas utilization at constant gas flows of 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU), 750°C
The test results of the different test methods show a good homogeneity among the tested stacks of both testing campaigns and also for both fuel and oxygen utilization. According to the theory the stack power density increases almost linearly with increasing fuel gas utilization and oxygen utilization. Minor deviations can be explained by temperature differences, stack differences and rounding errors. Compared to the negative electrode side the influence of decreasing O2 supply on the stack power density is less pronounced due to the applied excess amount of air. It has been found that reproducible and accurate settings for the operating value and the step size of the electrical current are essential in order to identify inhomogeneity among the tested stacks.
Validation of TM 08: Temperature Sensitivity
The temperature sensitivity has been investigated in both testing campaigns. Figure 12 shows representatively the power densities of the tested stacks of the second validation campaign at different electrical current densities as a function of the average stack temperature. The stacks were operated at furnace set points of 720°C, 750°C and 780°C with 0.5 H2 + 0.5 N2 NLPM/RU on the negative electrode side and 4 air NLPM/RU on the positive electrode side. After temperature stabilization the current density was increased from 0 A/cm² to 0.3 A/cm² and 0.5 A/cm² up to 0.7 A/cm².
Figure 12: Stack power densities at varying current densities and different operating temperatures with operating gases of 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU)
The test results among the partners are very homogenous and according to the theory. Hence, only minor improvements in the test module have been necessary. In the first testing campaign the stack power densities have been represented as a function of the specified operating temperature. But due to the heat generation of the proceeding exothermal reactions the average stack temperatures at given current densities differ strongly from the specified operating temperature. This temperature increase becomes more obvious with increasing current densities (up to 29°C at 700 mA/cm²). However, it can be seen that the temperature dependency among the partners is nearly the same which is indicated in Figure 12 as dotted lines.
Validation of TM 12: Operation under constant current
The operation under constant current in SOFC mode was performed at 300 mA/cm², 750°C and 0.5 H2 + 0.5 N2 / 4 air (NLPM/RU). These test conditions were set stable during the testing period of about 1000 h. The voltage behavior of the stacks of two partners (DLR, DTU) is shown in Figure 13. The OCVs of both stacks are very stable and slightly increase with time. This can be explained by an increase of the gas tightness of the sealing between stack and gas distribution plate. Moreover, nearly similar degradation rates of the stack voltages were obtained by both partners. This is due to the reliable control and the accurate monitoring of all interface conditions between stack and test station. All this knowledge for obtaining reproducible degradation values was used to optimize test module TM 12.
Figure 13: Long term behaviour of two stacks (DLR, DTU) at 750°C, constant current density of 300 mA/cm² and 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU)
Validation of TM 14: Thermal Cycling
TM 14 was investigated during the second testing campaign. Detailed information concerning the operating procedure is given in test module TM 14 and deliverable D 4.4 (Report of the second SOFC testing campaign). During thermal cycling two types of degradation can be distinguished: the degradation of OCV and the voltage degradation under constant current. Figure 14 shows the OCVs and stack voltages of three different stacks during 30 thermal cycles. For each cycle the stacks have been operated at 750°C for approximately 90 minutes at 500 mA/cm² with 0.5 H2 + 0.5 N2 NLPM/RU on the negative electrode side and 4 NLPM/RU air on the positive electrode side
Figure 14: Long term behaviour of different stacks (DTU, JRC, ENEA) during 30 thermal cycles (200°C - 750°C) at 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU) and 750°C
The OCVs of the stacks remain almost constant during thermal cycles which prove a very good gas tightness of the stacks with low thermo-mechanical stresses. The voltages of the stacks at 500 mA/cm² differ in a range from 3.9 V to 4.2 V which is in good accordance with the initial jV-curves of these stacks. Regarding the total degradation rates after 30 thermal cycles stacks with similar performance seem to degrade in a same range. There is also a tendency that stacks with lower performance show more stable behaviour thanks to the lower degradation rates. Similar to TM 12 the very accurate control of all stack input parameters is necessary in order to achieve reproducible degradation rates.
3.3 Results of final SOFC round robin test
At the end of the SOCTESQA project the final round robin test was dedicated to validate the developed test modules with focus on applicability, robustness, reproducibility and reliability. The test results of all optimized test modules which were applied in this final round robin test are very consistent with the former testing campaigns. Moreover, the observed differences of the stack behaviour are in good agreement with previous sensitivity analyses of the test input parameters. The following sections show some test results of all testing campaigns for some selected test modules. The detailed results are reported in deliverable D 4.5 (Report of the final round robin test).
Validation of TM 03: Current-voltage characteristics
The sensitivity analysis of the previous testing campaigns identified several test input parameters, especially the gas inlet temperatures, as important influences on the stack performance. Therefore, Figure 15 shows the stack power densities gained from the jV-characteristics of all stacks tested in the SOCTESQA project as a function of the average temperature of the inlet gases (Tav,gas in). Both values were taken at a current density of 500 mA/cm². The stacks have been operated at 750°C with 0.5 H2 + 0.5 N2 NLPM/RU at the negative electrode and 4 NLPM/RU air at the positive electrode.
Figure 15: Stack power density calculated at 500 mA/cm² as a function of the average temperature of the inlet gases (Tav,gas in) measured at 500 mA/cm²
Due to the optimization of the gas preheating subsystems the gas inlet temperatures for most of the stacks are in a range of 737°C and 753°C. Therefore, only a small dependency of the gas inlet temperatures on the stack performance is visible. The power densities of all tested stacks, except of two, have a variation of about 5.1 % between all stacks (highest and lowest value). This is very close to the quality variation threshold of 5 % given by the stack supplier in deliverable D 2.1 (List of SOC specifications). Hence,, a high quality and a high reproducibility of the results are proven.
Validation of TM 03 (jV characteristics) and TM 04 (EIS)
In order to validate the different test methods, investigations across the test modules and the several testing campaigns have also been performed. The reproducibility of TM 03 and TM 04 is shown in Figure 16 which summarizes the correlation between the low frequency resistances RLF measured with the electrochemical impedance spectra (TM 04) and the calculated area specific resistances of the jV-curves (TM 03) of the RUs of all stacks with increasing current. Both values were determined at a direct current density of 238 mA/cm² with gas flow rates of 0.5 H2 + 0.5 N2 NLPM/RU on the negative electrode side and 4 NLPM/RU air on the positive electrode side at 750°C.
Figure 16: Correlation between RLF of EIS and ASR of jV-curves of RUs at DC current of 238 mA/cm², 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU) and 750°C
Most of the values are located quite near or underneath the dotted line which indicates the ideal theoretical relation of both test methods. However, the area specific resistances calculated from the jV-curves are slightly higher than the total resistances (RLF) determined by electrochemical impedance spectra. The deviations are mainly related to different temperatures of the two test methods. Usually the stack temperature when measuring EIS spectra is slightly higher compared to the temperature when measuring a jV-curve with increasing current density. Hence, the ASR values are higher compared to RLF. In the case of the jV-curve with decreasing current density, this is vice-versa. This effect is explained in detail in the deliverable report D 4.4 of the second testing campaign. Considering the different thermal conditions both test methods show a good reproducibility and prove the reproducibility of both test methods.
4 SOEC (WP 5)
Work package 5 (WP 5) “Solid oxide electrolysis cell (SOEC)” is dedicated to the evaluation and validation of the test procedures defined in WP 3 in the electrolysis mode through the testing of the cell/stack assembly unit specified and procured in WP 2. Similar to the above described SOFC section, the results of the validation processes enabled the optimization of the test modules with respect to SOEC operation.
4.1 Adaptation of test stations for SOEC testing
Depending on the initial status of each partner’s test station at the beginning of the project, this adaptation and commissioning step was composed of different types of actions and conducted differently in each laboratory. To list the specifications of each test station available in each laboratory, all the thoughts and discussions around deliverable D 5.1 helped the consortium:
• to choose the ElringKlinger (EK) stack as the most easily testable common commercial cell/stack assembly unit,
• to identify some important specifications to be fulfilled by the test stations for SOEC operation,
• to point out the last mandatory adaptations to be finalized for each test station.
Following test station adaptations, modifications and upgrades were realized successfully:
• mass flow controller and thermocouple changes,
• gas diffusion plates and voltage probes integration,
• gas preheating subsystems implementation and optimisation,
• upgrade of acquisition subsystem in particular to follow dynamic profiles,
• steam production stability’s improvement, which plays an important role to avoid noise on voltage and EIS signals in SOEC mode.
The details of these adaptations are described in the deliverable D 5.2 (Report of the initial SOEC test bench validation). Some results of the test adaptations and optimizations are shown in the following figures. Figure 17 and Figure 18 display the results of the optimisation of the gas preheating subsystem and the furnace thermal homogeneity of the different partners. These two diagrams show the differences between relevant temperatures at OCV for each partner’s test station in SOEC mode before optimization (Fig. 17) and after optimization (Fig. 18). The corresponding symbols are: TTP = top plate temperature (= Tstack in this project), TBP = bottom plate temperature, Tneg,in and Tpos,in = inlet gas temperatures on both negative electrode and positive electrode sides, respectively, Tneg,out and Tpos,out = outlet gas temperatures on both negative electrode and positive electrode sides, respectively. In particular the temperature homogeneity of the test stations at ENEA and NTU was clearly improved.
Figure 17: Differences between relevant temperatures at OCV for each partner’s test station in SOEC mode before optimization
Figure 18: Differences between relevant temperatures at OCV for each partner’s test station in SOEC mode after optimization
Figure 19 shows the improvements which have been achieved in terms of voltage stability in SOEC mode. Due to optimization of steam supply stability the voltage fluctuations were significantly reduced. Hence, the quality of the electrochemical impedance spectra signals could be increased especially in the low frequency (LF) region. Moreover, the electromagnetic interferences, which disturb the high frequency (HF) region in particular at CEA and DTU, were significantly minimized.
Figure 19: EIS diagrams of RU 3 in SOEC obtained at CEA before and after optimization (750°C, -12 mA/cm² and 0.5 H2O+ 0.125 H2 // 1 air (NLPM/RU))
The test station adaptations have led to further increase the reproducibility of results of the first and of the second testing campaigns. Moreover, during the project duration reliable interfaces between the stack and the test station were provided, operating conditions as similar as possible between all partners were achieved and dynamic profiles could be followed.
4.2 Results in SOEC mode
The following sections highlight selected results for the validation of the test modules which were mainly obtained in the first and the second testing campaigns in SOEC. The TMs were validated under SOEC operating conditions as realistic as possible, in particular at higher electrical current density in the second testing campaign. The overall, detailed results are reported in deliverable D 5.3 (Report of the first SOEC testing campaign) and D 5.4 (Report of the second SOEC testing campaign).
Validation of TM 03 (j-V characteristics) and TM 04 (EIS)
Above all, a very high stability and homogeneity of the test station components are required for a good reproducibility of results in particular for the very sensitive EIS technique (TM 04). Figure 20 shows the current-voltage characteristics of the stacks of the partners in SOEC mode at 750°C, 80 % H2O + 20 % H2 // air. Similar to the fuel cell operation the OCVs of the stacks are exactly the same. The current voltage curves are almost linear with some higher differences among the partners compared to the SOFC mode. However, the performance trend of the stacks is similar to the SOFC operation. It is shown that all test stations are able to supply the targeted H2/air compositions to the tested stacks within ± 5 % of deviation and in a stable way all along the test. However, the steam production is less stable which induces voltage fluctuations bothering the jV-curves and EIS measurements analysis.
Figure 20: Current voltage behavior of tested stacks of the different partners in SOEC mode at 750°C, 0.5 H2O+ 0.125 H2 // 1 air (NLPM/RU)
Due to the induced voltage fluctuations, a satisfying steam supply stability is required and it is recommended to maintain voltage fluctuations below ± 10 mV for lowest results dispersion especially during EIS measurements which are very sensitive to this phenomenon (see Figure 19). Depending on the steam supply subsystem, a kind of nitrogen “dilution” can be used to enhance steam supply stability and cleanness of the EIS low frequency region. Additionally, the use of numerically fitting of the spectra by an equivalent electrical circuit for ASR determination can be efficient to avoid this drawback when low frequency disturbances remain and to succeed in determining the total resistance. This point is also very important to allow the determination of the ASR at one particular current in a small zone of the jV curve. In case of remaining voltage fluctuations when jV-curves are recorded (TM 03), it is preferable to determine ASR values on a large zone of the jV-curve (0.1 A/cm² at least) for a lowest results dispersion among partners. A linear fit or a polynomial one can also be used depending on the global shape of the jV-curve (see Figure 21 below). In case of voltage fluctuations and thermal inhomogeneity (first testing campaign), it is lower when a larger zone of the jV-curves at both current densities or a linear fit of the full jV-curves is taken for ASR determination. If voltage stability and thermal homogeneity are improved, the dispersion difference between the different calculation methods is reduced.
Figure 21: ASR dispersion among partners obtained from jV-curves in SOEC mode with different ASR calculation methods.
As already mentioned in section 4.1 (Fig.17 and Fig.18) thermal homogeneity strongly influences dispersion of results and hence the reproducibility of the stack results. The correlation with performance is presented in Figure 22, which shows a clear dependency of the stack ASR (linear regression) from the inlet gas temperatures. So it is highly recommended to measure temperature at different locations in the stack (top and bottom plates, gas inlet and outlet on both sides for instance) and try to limit the difference between those temperatures below ± 6 °C for a lowest results dispersion.
Figure 22: Stack ASR in SOEC mode at 500 mA/cm2, 750°C and 0.5 H2O + 0.125 H2 // 1 air (NLPM/RU) as a function of average temperature of the fuel and oxidant inlet gases
Validation of TM 12 (operation at constant current) and TM 13 (operation at varying current)
In steady-state operation at constant current density, it appears that internal pollution brought by the inlet gases and the test station itself can increase voltage and ASR degradation rates during long-term operation (TM 12 and TM 13). Figure 23 shows that initially stack voltage evolution with time at a constant operating point was very different between CEA and EIFER leading to very different voltage degradation rates. Some issues as temperatures homogeneity, gases and water quality and steam flow stability have been invoked.
After changing the steam generator, the preheating loops and the purity of hydrogen gas (from grade 3.5 to 6.0) CEA measured a much lower degradation (from 9%/1000h to 3%/1000h) with nominally identical stacks. This improved voltage degradation rate is probably due to the reduction of contamination. The results of EIFER show that the degradation rate depends on test duration. So the necessity to extend the test to a reasonable duration is clearly recommended in both TM 12 and TM 13. To complete that point, Figure 24 below shows that after reduction of the internal pollution on CEA test station; voltage degradation rate value obtained in this laboratory was decreased for a higher reproducibility between both testing partners.
Figure 23: Long term behavior of 4 stacks of two partners (EIFER and CEA ) in SOEC mode at 750°C, constant current of -300 mA/cm², 0.5 H2O+ 0.125 H2 // 1 air (NLPM/RU)
Figure 24: Voltage degradation rates calculated during the initial test bench validation (in blue) and the first testing campaign (in orange) under constant current operation at two testing partners EIFER and CEA with two different inlet gases qualities at CEA
For a comprehensive understanding of the SOC assembly unit degradation behavior the determination of different degradation rates (voltage, ASR from jV-curve and from EIS) is recommended. For instance the continuous recording of voltage allows to follow the degradation dependency with time (continuous or by sequence ...) whereas the EIS diagrams recorded before and after a long-term operation allow to separate ohmic, polarization and gas concentration resistance degradation rates and to give ideas about possible degradation phenomena.
4.3 Results of final SOEC round robin test
The final round robin test allowed to fully validating five test modules in single SOEC mode. The results are described in detail in deliverable D 5.5 (Report of the final SOEC round robin test). It is worth noticing here that globally results dispersion was a little bit higher for this final testing campaign. In contrast to previous testing campaigns which showed satisfying improved dispersion due to efficient test station adaptation/improvement, the results of the final round robin test scatter slightly higher. This can be attributed to the fact that the corresponding tests were performed on stacks from different fabrication batches and the lower robustness of the stacks coming from the last production batch. However, the results are reproducible to the results of stacks from the same batch of previous testing campaigns. Therefore, despite the higher result dispersion within the final round robin test, the robustness of the final test modules was confirmed.
5 Combined SOFC/SOEC (WP 6)
The objective of WP 6 (combined SOFC/SOEC) is the evaluation and validation of the test procedures for cells and stacks in combined SOFC/SOEC mode defined in WP 3. This includes similar to WP 4 and WP 5 the validation of the applicability, the reproducibility and the repeatability of the test results but also the sensitivity analysis of the corresponding test input parameters in reversing SOFC/SOEC operation.
5.1 Adaptation of test stations for combined SOFC/SOEC testing
A specification of the requirements for test stations with respect to combined SOFC/SOEC testing was made in the deliverable D 6.1 (Specification list of test benches for combined SOFC/SOEC Operation). The most important requirements are steam generator, fuel gas quality, electronic load, power supply, equipment for electrochemical impedance spectroscopy, and stability and noise levels. Basically, all these items are also considered in the single mode operation in SOFC mode and SOEC mode, and hence dealt with in WP 4 and WP 5, respectively. A few further items for special attention for the combined SOFC/SOEC operation have been identified. They deal with the time for switching and the automation of switching and with limitation in reversible and continuous measurement of the jV-characteristics across OCV.
The time required for switching between SOFC and SOEC mode (and vice versa) depends on the maximum ramp rate in the shift (any change in temperature, gas, and humidity change) and the time required for any manual switch (shift of gases and shift between SOFC electrical load and SOEC power supply). When changing from high humidification (SOEC) to a lower value (SOFC), in some cases it will take some time (up to 1 hour) before all the steam is removed and the stack voltage value is stable again. A maximum time for switching of 2 hours was specified and a required holding time at OCV has been accepted in the requirement for the test stations.
Five partners have completed the initial test station validation program (described in D 3.2) including combined SOFC/SOEC operation. The test results from the combined operation have been reported in D6.2 (Initial test station validation in combined SOFC-SOEC mode). The period with "combined operation" included followed the sequence: SOFC-SOEC-SOFC-SOEC, with 24 hours for each step, including a 2 hour period for switching. And overview of two SOFC/SOEC switching procedures is given in Figure 25 below.
Figure 25: Overview of the SOFC/SOEC switching procedure with two periods in SOFC and two periods SOEC
All the test stations of the partners are able to follow the planned set points. All the test laboratories have shown that they are able to perform the switching between the operation modes within the specified 2 hours period. When returning to the second SOFC step and the second SOEC step, all stacks were found to return to the same performance level, which means that the results are at least reproducible for same stack, same test station within a short period of days. This is shown for the SOFC mode in Figure 26, and for the SOEC mode in Figure 27. These diagrams show for each partner the cell voltages at +/- 300 mA/cm2 and stack temperatures for the 1st and 2nd SOFC operation period in SOFC and SOEC operation, respectively. All partners have a high reproducibility of own results, whereas somewhat larger differences are observed between partners
Figure 26: Average voltage of RUs and temperature of stacks of the partners for the 1st and 2nd SOFC operation period
Figure 27: Average voltage of RUs and temperature of stacks of the partners for the 1st and 2nd SOEC operation period
For a simple combined testing and short-term testing (< 5 days) automation of the switching is not required. However, in the further work for validation of the application specific test program (power-to-gas-to-power) automation of switching is foreseen to be required in order to be able to carry out the work within normal work hours, and to be able to carry out the actual test without an unreasonable high work load involved (for manual shifting). During the second period of the project, NTU and DTU were mainly responsible for the testing of SOC stacks in combined SOFC/SOEC mode. Both partners have succeeded in modifying their test stations with fully automatic SOFC/SOEC mode switching. This allowed for long term testing with numerous switches and switches at any time of the day or the week. Moreover, laboratory manual working resources have been saved significantly.
The initial test station validation results also indicated that NTU had low gas inlet temperatures. During the second period of the project, NTU has improved their gas pre-heating system and increased the negative electrode gas inlet temperature (Tneg,in) for about 20 °C. The corresponding details were reported in the deliverable report D 5.3 (Report of the first SOEC testing campaign). DTU also identified and solved the issues associated with the stack voltage fluctuation in SOEC gas composition and low stack voltage when shifting from SOEC mode to SOFC mode. The results are reported in deliverable D 6.3 (Report of the first testing campaign in combined SOFC-SOEC).
5.2 Results in combined SOFC/SOEC mode
After the initial test station validation campaign of the first project period, the first and second test campaigns were performed in combined SOFC/SOEC operation in order to optimize the test modules of WP 3. In the following section the most important results of these validation processes are reported. The results were mainly generated by the two responsible partners NTU and DTU.
Figure 28 shows the stack voltage and current density evaluation as function of operating time of the stacks tested in the first test campaign. The detailed results of this test campaign are reported in D 6.3 (“Report of the first SOFC/SOEC testing campaign”). The stacks were tested at 750°C with 50% N2 + 50% H2 in SOFC and 20% H2 + 80% H2O in SOEC supplied to the negative electrode compartment. In both modes air was supplied to the positive electrode compartment. The stacks show very similar performance at the beginning, but different degradation behavior during the combined SOFC/SOEC mode operation. Nevertheless, after the initial period (first 200 hours) the two stacks show quite low and similar voltage degradation rates.
Figure 28: Stack voltage and current density evaluation as function of testing time. The stacks were tested in the first test campaign at 750°C under 0.3 A/cm2 and 50% N2 + 50% H2 in SOFC mode and -0.3 A/cm2 and 20% H2 + 80% H2O inSOEC mode.
In the second test campaign, validation of the test modules by testing the stacks under “harsher condition” was suggested. The current densities in the test program TP04 (“Test program for the application of combined SOFC-SOEC in a power-to-gas-to-power system”) were changed from -0.3 A/cm2 to -0.5 A/cm2. The results are presented in Figure 29. The entire results are reported in D 6.4 (Report of the second SOFC-SOEC combined mode testing campaign).
Figure 29: Stack voltage and current density evaluation as function of testing time. The stacks were tested in the second test campaign at 750°C under 0.5 A/cm2 and 50% N2 + 50% H2 in SOFC mode and -0.5 A/cm2 and 20% H2 + 80% H2O in SOEC mode.
Higher overall voltage degradation rates were observed on both tested stacks when operated at higher current densities. However, stacks tested at the same institute show overall similar degradation trend, for example, the stacks tested at DTU show high initial voltage degradation rates in the first 200 hours and stabilize afterwards. Stacks tested at NTU show almost linear degradation in the entire testing period. Nevertheless at the end of the tests (the last 150 hours), the stacks tested at DTU an NTU showed similar degradation rates when operated under the same test conditions. The different initial degradation behavior between stacks tested at DTU and stacks tested at NTU may be caused by different gas inlet gases qualities with different impurity levels.
5.3 Results of final SOFC/SOEC round robin test
In the final round robin test campaign, besides validation of the same test modules, DTU and NTU have carried out the final test on combined SOFC/SOEC operation in order to check the robustness of the testing procedures. The results of the combined SOFC/SOEC mode operation have been reported in deliverable D 6.5 (Report of the final round robin testing test).
Figure 30 shows the stack voltage and current density evaluation of DTU and NTU stacks as function of testing time of the final round robin test. Similar degradation behavior/results were observed compared with the second testing campaign.
Figure 30: Stack voltage and current density evaluation as function of testing time. The stacks were tested in the final round robin campaign at 750°C under 0.5 A/cm2 and 50% N2 + 50% H2 in SOFC mode and -0.5 A/cm2 and 20% H2 + 80% H2O in SOEC mode.
Moreover, as can be seen from Figure 31 and Figure 32, stacks tested at DTU and NTU in different testing campaigns show very good repeatability. Only minor difference can be seen for the stacks tested at the same partner, which can be explained by differences in the stack hardware itself. The results prove the very good robustness of the testing procedures, even though different initial degradation behavior is observed between partners, which may be related to different impurities from the test stations or to different inlet gases qualities supplied to the stacks. Longer time operation is therefore recommended for better evaluation and comparison of stacks degradation.
Figure 31: Stack voltage and current density evaluation as function of testing time. The stacks were tested at DTU in the second test campaign and the final round robin test campaign at 750°C under 0.5 A/cm2 and 50% N2 + 50% H2 in SOFC mode and -0.5 A/cm2 and 20% H2 + 80% H2O in SOEC mode.
Figure 32: Stack voltage and current density evaluation as function of testing time. The stacks were tested at NTU in the second test campaign and the final round robin test campaign at 750°C under 0.5 A/cm2 and 50% N2 + 50% H2 in SOFC mode and -0.5 A/cm2 and 20% H2 + 80% H2O in SOEC mode.
Potential Impact:
The ultimate objective – to come up with procedures at the end of the project that are not only validated inside the laboratory, but are already shaped towards the requirements of regulations codes and standards as well as industrial productivity and reliability – has been successfully achieved. The interaction with relevant industrial stakeholders, the broad dissemination of the project results and the liaison with standards developing organizations (SDO) have been important issues in the project. Figure 33 shows the overall consortium of the project. The test procedures incorporate the inputs from both entities and foster worldwide discussion and awareness of the topic. In this way the maximum exploitation of the project outcome was achieved. The details of dissemination and exploitation of the project results are reported in deliverables D7.2 (“Interim report on liaison and dissemination activities”) and D7.3 (“Final report on liaison and dissemination activities”).
Figure 33: Overall consortium of the SOCTESQA project
1.1 Liaison to SDO
SOCTESQA has carried out a comprehensive survey of ongoing standardization activities in the field and has entered into contact and liaison with the main bodies currently working on regulations for hydrogen and fuel cell technologies. Essentially, these are grouped under the international bodies of the ISO Technical Committee 197 on hydrogen technologies – which looks mainly at safety issues of electrolyzers and hydrogen handling – and the International Electrotechnical commission (IEC) TC105 on fuel cell technologies. The latter is much more focused on the technology and the definition of practical guidelines in terms of system performance, installation and characterization. Also CEN/CENELEC, the main European SDO in the field, has been liaised with, through an official agreement.
The work in SOCTESQA is considered most in line with IEC, and in fact the Technical Specification on solid oxide fuel cell and stack test procedures (issued in 2014 as IEC TS 62282-7-2) has been a fundamental seed document for further elaboration of the SOCTESQA procedures, which however break new ground by providing guidelines also for the operation of SOC stacks in electrolysis mode (SOEC) and reversible mode (combined SOFC/SOEC). After these initial approaches a formal Liaison was initiated with the IEC TC105. This underpinned the initiatives on behalf of SOCTESQA project taken up by the IEC, i.e. submission of a New Work Item Proposal (NWIP) and subsequent initiation of the new Working Group (WG) 13 on Standards for fuel cells operated in regenerative mode. The leader of this work package 7 of SOCTESQA also acts as project leader for the first standard and as convener for the entire WG 13 of IEC. Through this WG and the formal liaison, the entire SOCTESQA consortium was able to participate in and guide the activities that will lead to the publication of Standard 62282-8-101: “Solid oxide single cell and stack performance including regenerative operation”, which incorporates all the work carried out in the SOCTESQA project.
Through the IEC TC105 committee, liaison is also ensured with ISO’s TC197 on hydrogen technologies, which will be monitored through the former activity. Simultaneously, CENELEC, the European technical standardization body, has also initiated a new working group on Hydrogen. The scope of the working group covers the production of hydrogen through electrolysis and the transportation, distribution and usage of that hydrogen in pure form or as a natural gas dominant mixture (H2NG). In addition, actions in cross-cutting fields such as safety and training of personnel are identified. The final objective of this working group is to set a long term collaborative framework (liaison) with major bodies for strengthening cooperation between regulatory work, standardization work and RDI programs (e.g. European Commission, JRC, FCH2 JU, IEA, ISO, IEC). DLR has joined this working group, namely task force 2: electrolyzers, and participated at several meetings with the background of transferring the results and experience so far achieved in SOCTESQA. Moreover, DLR has entered into a formal liaison with CENELEC to monitor the progress in this working group on behalf of the SOCTESQA consortium. The corresponding liaison agreement between DLR and CENELEC was signed in Sept. 2015.
1.2 Liaison to industry
Interaction with industry has been part of SOCTESQA since the time of writing the project proposal: a number of letters were gained at that time where key industrial players manifested their interest in following the project’s activities and achievements. These industries were chosen to constitute the project’s industrial advisory board (IAB).
At the beginning of operations, fact sheets were compiled by SOCTESQA partners and sent out to the IAB for gaining technical input as to the operating conditions of SOC systems in the applications targeted: systems for combined heat and power generation (µ-CHP), auxiliary power units (APU) and electrolysis systems mainly. All industries of the IAB have been invited to participate in the first of two workshops organized by SOCTESQA, which was held on December 15th, 2015, in Naples, in conjunction with the 6th European Fuel Cell Technology & Applications “Piero Lunghi” conference. The scope of this workshop was to present SOCTESQA activities and trigger input from international industries and SDO for a hands-on discussion on how to formulate robust, univocal procedures relevant for industry and that can facilitate quality assurance in the production process and confident market entry.
During the project progress the IAB as well as all industrial stakeholders were kept up to date with SOCTESQA progress through personal exchanges and the SOCTESQA newsletters. Finally, the SOCTESQA project outcome workshop reached out to industrial stakeholders directly, through a booth and two presentations at the Hannover Fair, 24-28 April 2017, the biggest exhibition on fuel cells and hydrogen technologies in Europe. This liaison/dissemination event is reported more detailed in deliverable D 7.4 (“Project outcome workshop”).
1.3 Dissemination of project results
The basic tool for dissemination, namely the project website with a “corporate” identity, was developed early in the project and went on line as planned (see also deliverable D 7.1:”Project website”). A list of dissemination activities carried out by SOCTESQA partners with details regarding dates, presentation types and titles, and outcome is given in deliverable D 7.3 (“Final report on Liaison and dissemination activities”).
The SOCTESQA project was intensively promoted, with a number of items disseminated to the general and scientific public. Two peer-reviewed articles and three articles were published in high-profile scientific conferences, like the Int. Conference on SOFC, the European SOFC&SOC Forum and the European Fuel Cell Conference. Moreover, the results of the project were broadly disseminated to the scientific audience with many presentations, e.g. 7 presentations in scientific conferences and 8 presentations in workshops or key international events including assemblies of the International Electrotechnical Commission (IEC). Especially the latter events have generated the highest dissemination impact. Thanks to this engagement and formalized in the liaison mentioned above, the project has provided the procedures for solid oxide cell and stack testing that are to become internationally applicable standards, for the benefit of industrial production all round.
As mentioned above, a crucial dissemination event that catalysed the interaction with industry and standardization, an important milestone within the SOCTESQA project, was the workshop organized on 15.12.2015 in the frame of the 2015 European Fuel Cell “Piero Lunghi” conference in Naples/Italy. Keynote speakers from standardization bodies (JRC, IEC and Japanese electrical manufacturers’ association – JEMA) and collaborating projects such as STACKTEST (FCH-JU) interacted with industry (CERES Power, Elcogen) and academia to demonstrate achievements, identify gaps and bottlenecks and ways forward in a lively debate. Moreover, two presentations were given to general audience in the frame of the final workshop including a project booth at Hannover Fair 2017.
Altogether 14 posters were presented in scientific conferences, workshops, fairs and the FCH-JU review days. Additionally, two newsletters were disseminated to relevant stakeholders. As already described above two project flyers and all important test module documents were handed out to scientific audience at Hannover Fair 2016 and 2017. Moreover, interaction with other European and National funded projects was established by participating and exchanging of the results at project meetings. And finally, all public deliverable reports and all test modules were uploaded on the project website.
The results and knowledge generated by the SOCTESQA project have led to a straightforward and open approach for the universal adoption of harmonized, quality-assuring procedures, which is the key for the successful technological progress and market implementation of the ceramic solid oxide technology and the corresponding entire industry.
List of Websites:
http://www.soctesqa.eu/
Contact details: Dr. Michael Lang, German Aerospace Center (DLR), Pfaffenwaldring 38-40, 70569 Stuttgart, e-Mail: michael.lang@dlr.de Tel: +49 711 6862605
The aim of the project is to develop uniform and industry wide test modules and programs for solid oxide cell and stack (SOC) assembly units. New application fields which are based on the operation of the SOC cell/stack assembly in fuel cell (SOFC), in electrolysis (SOEC) and in combined SOFC/SOEC mode are addressed. This covers the wide field of energy conversion systems, e.g. stationary SOFC μ-CHP, mobile SOFC APU, stationary SOEC systems (power-to-gas) and combined SOFC/SOEC (power-to-gas-to-power) systems. The project partners are DLR (coordinator, Germany), CEA (France), DTU (Denmark), ENEA (Italy), JRC (Belgium), EIFER (Germany) and NTU (Singapore). During the project a close interaction with an industrial advisory board (IAB) ensured to achieve industrial relevant outcome of the project. Additionally, a continuous liaison with standards developing organizations (SDOs) was established in order to implement the outcome of the project successfully into international standards.
Altogether 11 test modules were developed in the project, which are TM 00: General SOC testing guidelines, TM 02: Start-up, TM 03: Current-voltage characteristics, TM 04: Electrochemical impedance spectroscopy, TM 07: Reactant utilization, TM 08: Reactant gas composition, TM 09: Temperature sensitivity, TM 12: Operation under constant current, TM 13: Operation under varying current, TM 14: Thermal cycling and TM 16: Shut-down. These test modules were experimentally validated by the project partners in several testing campaigns with 5 test programs (TPs) for the different applications.
For the development of the test modules several steps were performed. At the beginning of the project all important specifications, especially of the interfaces between the test station and test object and the nomenclatures, were defined. For the validation of the test modules short stacks with 5 repeat units from ElringKlinger were used. In the first testing campaign the test stations were modified and harmonized in order to provide reliable and reproducible interfaces between the stacks and the test stations, to achieve operating conditions as similar as possible among all partners and to follow dynamic profiles. In the following two testing campaigns the test modules were validated and optimized with application-specific test programmes for SOFC, SOEC and combined SOFC/SOEC applications. In this context, technical input for the operating conditions of SOC systems from relevant industrial stakeholders were integrated in the test programmes. The validation process mainly addressed the quality, reproducibility and repeatability of the results between the partners and between the different test modules. The optimization of the test modules included thermal aspects like stack temperature, gas temperatures, temperature homogeneity inside the furnace, instabilities of the fuel gas humidification unit and electrical wiring setup for electrochemical impedance spectroscopy. Additionally, aspects for the reproducible determination of degradation rates were implemented in the long term operation test modules. Moreover, recommendations for the reliable calculation of derived quantities, e.g. area-specific resistance (ASR), reactant utilizations and low and high frequency resistances, were included in the test modules. At the end of the project the final round robin test was performed, which showed a high quality, reproducibility and repeatability of the results of the different test laboratories. Hence, the robustness of the optimized test modules was confirmed.
The exploitation of the project mainly focused on the transfer of the results to standards developing organizations, e.g. the International Electrotechnical Commission (IEC) and CEN/CENELEC, the main European SDO in the field. A formal liaison of the SOCTESQA consortium with both entities was established. In case of IEC a new working group (WG 13) was initiated within the technical committee TC105. The entire SOCTESQA consortium participated in and guided the activities, leading to the publication of Standard 62282-8-101: “Energy storage systems using fuel cell modules in reverse mode - Test procedures for solid oxide single cell and stack performance including operation in reverse mode”. Moreover, the project results have been broadly disseminated to general and scientific public, e.g. by scientific papers, conference contributions, workshops, fairs and the SOCTESQA website (http://www.soctesqa.eu). The results and knowledge generated in the SOCTESQA project have led to a straightforward and open approach for the universal adoption of harmonized, quality-assuring procedures, which is the key for the successful technological progress and market implementation of the ceramic solid oxide technology.
Project Context and Objectives:
The aim of the project is to develop uniform and industry wide test modules and programs for solid oxide cell and stack (SOC) assembly units. New application fields which are based on the operation of the SOC cell/stack assembly in fuel cell (SOFC), in electrolysis (SOEC) and in combined SOFC/SOEC mode are addressed. This covers the wide field of power conversion systems, e.g. stationary SOFC μ-CHP, mobile SOFC APU, stationary SOEC systems (power-to-gas) and combined SOFC/SOEC (power-to-gas-to-power) systems.
The project builds on the experience and the methodology gained in previous European projects, e.g. “FCTESTNET” and “FCTESQA”. However, these projects focused mainly on single cell tests and system level tests under steady state conditions. On the other hand other relevant stack projects, e.g. the “STACKTEST” project, do not address the high temperature fuel cell technology. None of these projects focuses on the development of test procedures of solid oxide cell/stack systems under dynamic operating conditions or in the electrolysis mode. Moreover, established advanced characterization techniques, e.g. the electrochemical impedance spectroscopy (EIS), have not yet been integrated in the test protocols of these previous projects. Only few results were validated by round robin tests.
Figure 1 shows a complete solid oxide cell/stack testing system with the corresponding control subsystems, which is very complex. Numerous test input parameters have a high impact on the test object output results. Moreover, a clear definition and specification of the interfaces between test object and test station is obligatory. Especially for high temperature solid oxide systems the temperature control subsystem and the gas control subsystems play an important role for the test output results. Moreover, the complexity of the test system is significantly increased when the operation mode is transferred from steady state operation to dynamic operating conditions. It is clear that detailed test schemes, procedures and protocols are essentially necessary for the development of the SOC cell/stack assembly unit.
Figure 1: Schematic graph of a test system for high temperature solid oxide assembly unit
The objective of the present project therefore concentrates on the development of uniform test procedures for high temperature ceramic solid oxide cell/stack assembly units both in the fuel cell and in the electrolysis mode. It is intended to develop a full set of application specific test modules and programs addressing function, performance, durability and degradation. These test procedures and programs usually consist of a combination of different test modules. The test procedures address three different operation modes, which are solid oxide fuel cell (SOFC), solid oxide electrolysis cell (SOEC) and combined SOFC/SOEC operation. Each of these cases includes both steady state and dynamic operations.
Within the project, the test modules and test programs are established and experimentally validated. The test procedures are validated on stack relevant test specimens, e.g. on short stacks with 5 cells. The test procedures include and specify the complete testing system, the different operating modes, the test conditions and the electrochemical characterisation methods, e.g. current-voltage curves, long-term operation and electrochemical impedance spectroscopy. The test modules and programs consider all relevant operating conditions of the complete SOC test system, e.g. input gases, pressure levels, temperatures, current loads, voltage level, mechanical loads, dynamic transients etc. In order to systematically study these effects, a sound understanding of the interaction of the test object with the test station is developed.
For the SOCTESQA project an organization and management structure has been built that suits the multitude of activities. In this project six European partners and one non-European partner are involved. These are DLR (coordinator, Germany), CEA (France), DTU (Denmark), ENEA (Italy), JRC (Belgium), EIFER (Germany) and NTU (Singapore). During the project a close interaction with an industrial advisory board (IAB) ensures to achieve industrial relevant outcome of the project. Additionally, a continuous liaison with standards developing organizations (SDOs) is aspired with the aim to implement the outcome of the project successfully into international standards. The activities are organized in work packages (WP), which are coordinated by work package leaders. These are WP 1: Coordination (DLR), WP 2: Specifications and Procurement (JRC), WP 3: Testing Procedures (EIFER), WP 4: Solid Oxide Fuel Cell (DLR), WP 5: Solid Oxide Electrolysis Cell (CEA), WP 6: Combined SOFC/SOEC (DTU), WP 7: Dissemination and Liaison (ENEA). Each work package contains several tasks, which are coordinated by task leaders. In this project the work package leaders are also the leaders of the corresponding tasks in their WP.
For the achievement of the project objectives, two workflow paths have to be taken into account. The first path is the development of the test modules and programs and the second path refers to the liaison activities to industry and to standard development organizations (SDO). Figure 2 show the workflow diagram for the project.
Figure 2: Workflow diagram for the project
The first step of the project concerns the specification of all relevant terms, definitions and terminology. After that a test matrix with all test modules for the SOFC, the SOEC and the combined SOFC/SOEC operation modes is developed. It is important that the test matrix reveals and confirms the industry needs from the very beginning of the project. This test matrix and the available test facility characteristics are the basis for the drafting and definition of the generic test modules and programs. In parallel the test specimens are defined and procured from solid oxide cell/stack manufacturing stakeholders. The specific test programs are defined by integration of the application specific conditions into the generic test modules. After this initial definition stage altogether 4 testing campaigns were performed. The first test bench validation campaign among the testing project partners figures out possible test station differences. For this purpose, the most critical test modules are selected. The review of this test bench validation focuses on the harmonization of the test benches by minimization of the possible differences, and the subsequent integration of any minor differences in equipment within the text of the test-specific procedures.
In the second testing campaign the specific test programs are applied to the SOC cell/stack assembly. The results of the specific test programs are the basis for the validation of the generic test modules. The test modules are reviewed and modified based on experimental results. In this context the repeatability of several measurements, the reproducibility of the different partners and methods are mainly addressed. Other issues are a sensitivity analysis of the test input parameters and the calculation of derived quantities. Establishment and validation of the generic test modules shall observe compliance with the applied specific test programs and consider any relevant feedback from these activities. With the modified test modules a third validation and review campaign takes place. After that a final round robin test is performed and reviewed to ensure comparability of the results obtained in different test laboratories. It is planned that each partner tests about 6 cell/stack assembly units during the 4 testing campaigns.
The second path refers to the liaison activities to industry and to standard development organizations (SDO). It is intended to integrate several independent SOFC and SOEC relevant industrial companies forming an industrial advisory board (IAB) in the overall project consortium instead to involve individual companies as project partners. Several important companies of the solid oxide cell technology have agreed and committed to support this project. This ensures a broad industrial advisory impact from different SOFC and SOEC stakeholders, e.g. material suppliers, cell and stack manufacturers or system developers on the project. Moreover, the test procedures can be developed independently from any possible individual industrial stakeholder`s interests thus enabling their broad international acceptance after the project. Since fuel cell testing is an integral part of international standardization activities, liaison with international standardization activities is integrated in the project and established in a separate work package. Contacts with the IEC TC105 responsible for the IEC 62282 family of standards as well as relations with the ISO TC 197 and other relevant working groups are considered appropriate in this context and the existing reciprocal influence will be enhanced. The output of the project are standardized and industry wide test procedures and protocols according to the different system applications.
Project Results:
The present section summarizes the results, which have been achieved during the three years of SOCTESQA project. For each work package (WP) the work progress and the main achievements are summarized according to the different tasks of the description of work (DoW). The relevant deliverables are referred.
1 Specifications and Procurement (WP 2)
This work package focuses on the identification and definition of all important specifications at the beginning of the project, the review of already existing test procedures and the procurement of testing specimen.
1.1 Definition of SOC specifications
At the beginning of the project all important specifications were identified and defined in work package 2. This includes the specification of the SOC test station, of the interfaces between the test station and the cell/stack assembly unit, of the testing system environment and of the nomenclatures. Figure 3 shows the corresponding interfaces between test station and short stack with the relevant test input and output parameters. Moreover, all the specifications for the solid oxide cell/stack assembly unit (SOC short stack) have been defined in cooperation with the potential stack suppliers. These specifications are compiled in the deliverable D 2.1 (List of SOC Specifications).
Figure 3: Interfaces between test station and short stack
1.2 Compilation of existent SOC test procedures
In the second task all existent SOC test procedures for single cells and stacks including those from previous projects and from the open literature were surveyed and compiled. The results were compiled in the deliverable D 2.2 (List of SOC test procedures). The definitions & terminology for use in the test procedures are taken from the IEC document (IEC 62282-1 TS: “Fuel Cell technologies Part 1: Terminology”) where needed supplemented by the FCTESTNET Glossary to avoid repetition and to preserve consistency with international standards. Definitions missing in these documents were defined by the consortium during the drafting of the test procedures. All definitions & terminology are listed in the SOCTESQA master document (Test module TM 00: General testing guidelines). The investigated literature already offers a number of tests modules for performance, reactant gas compositions, lambda sensitivity and cell/stack endurance whereas other modules especially for dynamic operation can hardly be found. Figure 4 shows the number of literature references for the different test modules. Most of the modules refer only to SOFC testing. The lack of SOEC test modules showed that the development of test programs for electrolysis mode needed the most effort within the present project.
Figure 4: Number of literature documents for the different test modules
1.3 Procurement of SOC materials and test samples
In this task the SOC short stacks intended for the validation of the test procedures and the round robin test campaign were procured. Several technical discussions with potential stack suppliers (ElringKlinger, Sunfire and SolidPower) were necessary to select the most suitable one. By weighing all pros and cons the partners decided to use short stacks with 5 repeat units from ElringKlinger for the validation of the test modules. The corresponding stack design as shown in Figure 5 required the lowest effort to be integrated in the test stations of the different partners.
Figure 5: Short stack with 5 repeat units from ElringKlinger used in the SOCTESQA project
After having organized all administrative issues the stacks for the first three testing campaigns were procured from ElringKlinger. The details of the procurements and deliveries are summarized in deliverables D 2.3 D 2.4 D 2.5 and D 2.6. For the round robin test no further stack procurement was necessary because most partners still had stacks from previous campaigns in stock, which were not yet tested. Stack shipment between different partners was organized. In this way, all partners managed to complete the full test schedule without stack redundancy.
2 Test Procedures (WP 3)
In work package 3 generic harmonized test modules and programs have been developed and elaborated for the cell/stack assembly unit. The test procedures are defined for both steady-state and dynamic operation conditions for SOFC, SOEC and combined SOFC/SOEC systems.
2.1 Test matrix definition
A test matrix (see Table 1) containing altogether 18 test modules which are relevant for four major applications has been defined. The four applications identified are:
• SOFC for micro combined heat and power (μ-CHP) and distributed power generation
• SOFC for auxiliary power unit (APU)
• SOEC for hydrogen production (power-to-gas)
• Combined SOFC/SOEC for electricity storage (power-to-gas-to-power)
In the SOCTESQA project altogether 10 test modules (bold in Table 1) and the general test module TM 00 (General SOC testing Guidelines) were developed. In order to make sure that the test matrix reflects the needs for the final applications, industrial stakeholders who are developing SOFC/SOEC products have been contacted to collect information regarding the required operation modes during the lifecycle of the product for each application. Feedbacks from 7 industrial stake holders have been received and taken into account.
Table 1: Test matrix for application-specific testing
2.2 Test procedures for initial test bench validation
In order to validate the applicability of test procedures defined in the present project, tests of solid oxide cells/stacks have to be performed in different laboratories and results should then be compared. It is therefore necessary to harmonize the specifications of test stations among different institutions. The purpose was to make sure that the facilities and the test object interfaces in all testing partners’ laboratories are suitable for the test procedure evaluation. All project partners have adapted their test stations accordingly in order to realize nearly identical operating conditions of stacks. For this purpose a test program has been developed.
In this test program, all three operation modes of the SOC stack (SOFC, SOEC and combined SOFC/SOEC) have been considered. Operating parameters have been defined by taking into account the stack specifications supplied by the stack manufacturer as well as constraints coming from test stations. Five generic test modules have been created and embedded in the test program. These are:
• TM 02: Start-up
• TM 03: Current-voltage characteristics
• TM 04: Electrochemical impedance spectroscopy
• TM 12: Operation under constant current
• TM 16: Shut-down
Figure 6 shows exemplarily the structure and the content of the test module TM 03 (Current voltage characteristics). A test module contains all relevant information and guidelines for the reliable testing of the SOC test object, e.g. the objective and scope of the test, all test input and output parameters, the formula for the derived quantities. Moreover, the test procedure and the data processing and presentation are described in detail.
Figure 6: Simplified example of test module: TM03 (Current-voltage characteristics)
In addition, a lot of effort has been dedicated to the creation of the master document TM 00 (General SOC Testing Guidelines), which contains common issues applicable for all test modules defined in the test matrix. Among others the following issues are addressed:
• General safety aspects
• Description of test object (cell and stack)
• Description of test system
• Interface between test object and test system
• List of quantities which are important to be controlled / measured
• Reporting of testing results
2.3 Test procedures for first validation
In this task the 5 test modules of the test station validation (TM 00: General SOC testing guidelines, TM 02: Start-up, TM 03: Current-voltage characteristics, TM 04: Electrochemical impedance spectroscopy, TM 12: Operation under constant current and TM 16: Shut-down) were optimized. Moreover, another 4 test modules have been drafted. They are TM 07: Reactant utilization, TM 08: Reactant gas composition, TM 09: Temperature sensitivity and TM 13: Operation under varying current.
In addition, three application-specific test programs (TP01, TP03 and TP04) have been defined, which have been used by the project consortium in the first test campaign. These are:
• TP01: Test program for the stationary application of SOFC in a µ-CHP system
• TP03: Test program for the application of SOEC in a power-to-gas system
• TP04: Test program for the application of combined SOFC/SOEC in a power-to-gas-to-power system
In these three test programs, all 9 test modules have been integrated. Figure 7 shows as an example the scheme of test program TP01 (Test program for the stationary application of SOFC in a micro-CHP system). This test program is a sequence of all important test modules for this system application. Moreover, some of the test modules were also run in SOEC mode in order to investigate their reproducibility in electrolysis mode (see below section 3: SOEC). All test procedures are documented in deliverable D 3.3 (Test procedure document for the first validation).
Figure 7: Test program for the stationary application of SOFC in a µ-CHP system
2.4 Test procedures for second validation
According to the results obtained during the first validation test campaign, test modules have been revised and optimized to a new version. In addition, a new test program and a new test module have been developed. They are:
• TP02: Test program for the mobile application of SOFC in an APU
• TM 14: Thermal cycling
TM 14 has been integrated into TP02. During the second validation test campaign, TP02, TP03 and TP04 have been applied by the project consortium. Deliverable D 3.4 (Test procedure document for the second validation) compiles all 11 test modules and 4 test programs.
2.5 Test procedures for final round robin test
Based on test results achieved during the second validation test campaign, all test modules and test programs have been reviewed and further optimised. In order to check the robustness of these test procedures, another test program (TP05) has been defined for the final round robin test, see deliverable D 3.5 (Test procedure document for the final Round Robin Test).
2.6 Final test protocols
The final round robin test results supplied the last feedbacks for optimising all 11 test modules developed in the project. After harmonising the format of all test modules, the final versions of test protocols have been worked out. In addition, a one-page summary has been created for each test module in order to provide a quick overview. Final test protocols have been compiled in the deliverable D 3.6 (Final document of test protocols) and are also available for the public on the SOCTESQA website.
3 SOFC (WP 4)
This work package aims at testing and validation of the cell/stack assembly unit in SOFC operation with the test procedures specified in WP 3. Recommendations based on the experimental validations are directly integrated into the improvement of the developed test modules.
3.1 Adaptation of test stations for SOFC testing
As the test facilities differ among the project partners, one of the first steps in WP 4 was a compilation of the test station specifications and a comparison with the specifications of different stacks and the test program requirements. As most of the test station furnaces of the partners are limited in geometrical dimensions it has been shown that the size of the stack will be a critical factor. In cooperation with WP 2 a suitable stack supplier (ElringKlinger) was chosen which fits the test stations of all partners. The results of this survey and the recommended modifications for each test station are documented in the deliverable D 4.1 (Specification list of test benches for SOFC operation).
According to the stack specifications and the aims of the project the test stations were adapted, particularly with regard to uniform test input parameters at the stack/test station interfaces. For the homogeneous gas supply to the stack an additional gas distribution plate had to be integrated in the test stations. This component was produced and customized specifically for each test station by an external supplier. In order to harmonize the adaptation and the integration of the stacks among the partners a manual called “Mounting of SOC-stack (EK) on gas distribution plate (GDP) in test station” was developed.
The adaptation and harmonization of the test stations among the partners have been continued during the testing campaigns. For instance two partners have been able to improve their gas preheating systems significantly after identifying demands during the initial test station validation. Moreover, the set-up for measuring electrochemical impedance spectra was modified by all project partners in order to minimize high frequency artefacts. The details of these test station optimizations are described in deliverable D 4.2 (Report of the initial SOFC test bench validation). Finally all test stations have been adapted successfully and have been able to perform the required test programs.
3.2 Results in SOFC mode
The following sections highlight the most important results for the validation of the test modules which were mainly obtained during the initial test station validation and in the first and the second testing campaign. The overall, detailed results are reported in deliverable D 4.3 (Report of the first SOFC testing campaign) and D 4.4 (Report of the second SOFC testing campaign).
Validation of TM 03: Current-voltage characteristics
The SOFC current-voltage characteristics of the stacks of the partners at 750°C, 50 % H2+50 % N2 // air are shown in Figure 8. The test results show a very high reproducibility between the different test laboratories. The OCV of all partners is almost the same and the voltage curves especially at lower current densities are very similar. However, at higher current densities small differences in stack performance among the partners exist.
Figure 8: Current voltage behavior of tested stacks of the different partners in SOFC mode at 750°C, 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU)
In the next step of development a complete analysis of the SOFC current-voltage results on the test input parameters was performed. This so-called sensitivity analysis qualifies the influence of the test input parameters on the performance and helps to better understand the differences in performance. Figure 9 shows as an example the stack power density at 500 mA/cm2 as a function of the average temperature of the fuel and oxidant inlet gases. A clear increase in performance with increase in gas inlet temperature can be observed. In contrast to that, nearly no influence of the performance from the measured temperatures at the top and bottom plate of the stack was observed. Hence the gas inlet temperature could be identified as one of the most important influences. The results obtained by this sensitivity analysis were used to optimize the test module TM 03 (Current-voltage characteristics).
Figure 9: Stack power density at 500 mA/cm2 in SOFC mode as a function of average temperature of the fuel and oxidant inlet gases
Validation of TM 04: Electrochemical impedance spectroscopy
Figure 10 shows the electrochemical impedance spectra of repeat unit 3 of the stacks of the different partners at 750°C, 12 mA/cm2 and 50 % H2+50 % N2 // air. The spectra show uniform results among the partners. Especially in the high-frequency range the spectra are almost similar which indicates a high reproducibility for the determination of the ohmic resistance of a repeat unit. However, often high-frequency disturbances make it difficult to determine the ohmic resistance of a repeat unit. This can be observed in the spectra of CEA. This issue was integrated in the optimized test module and solved by twisting the voltage probes to each other in order to minimize magnetic interference to each other.
Altogether three frequency dependent processes can be identified in the spectra, which are the polarization impedances of the anode and the cathode and the gas concentration impedance at the fuel gas side. In the low frequency range variations of the overall impedances among the partners can be observed, which is due to the high sensitivity of the gas concentration arc on the fuel gas composition. However, the total resistance of the electrochemical impedance spectrum correlates very well with the ASR calculated with the jV-curve, which proves a good reproducibility between both methods.
Figure 10: Electrochemical impedance spectra of repeat unit 3 of the stacks of the different partners in SOFC mode at 750°C, 12 mA/cm2 and 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU)
Validation of TM 07: Reactant utilization
The influence of the reactant utilization on the stack power density can be investigated in different ways which are described in TM 07 (Reactant utilisation). Figure 11 shows exemplarily the stack power densities as a function of the fuel gas utilization where the gas flow rates were kept constant at 0.5 H2 + 0.5 N2 NLPM/ RU on the negative electrode side and 4 NLPM/RU of air on the positive electrode side at 750°C. The current was increased stepwise from 0.3 A/cm² to 0.5 A/cm² to 0.6 A/cm² and to 0.7 A/cm².
Figure 11: Stack power density at different current densities as a function of fuel gas utilization at constant gas flows of 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU), 750°C
The test results of the different test methods show a good homogeneity among the tested stacks of both testing campaigns and also for both fuel and oxygen utilization. According to the theory the stack power density increases almost linearly with increasing fuel gas utilization and oxygen utilization. Minor deviations can be explained by temperature differences, stack differences and rounding errors. Compared to the negative electrode side the influence of decreasing O2 supply on the stack power density is less pronounced due to the applied excess amount of air. It has been found that reproducible and accurate settings for the operating value and the step size of the electrical current are essential in order to identify inhomogeneity among the tested stacks.
Validation of TM 08: Temperature Sensitivity
The temperature sensitivity has been investigated in both testing campaigns. Figure 12 shows representatively the power densities of the tested stacks of the second validation campaign at different electrical current densities as a function of the average stack temperature. The stacks were operated at furnace set points of 720°C, 750°C and 780°C with 0.5 H2 + 0.5 N2 NLPM/RU on the negative electrode side and 4 air NLPM/RU on the positive electrode side. After temperature stabilization the current density was increased from 0 A/cm² to 0.3 A/cm² and 0.5 A/cm² up to 0.7 A/cm².
Figure 12: Stack power densities at varying current densities and different operating temperatures with operating gases of 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU)
The test results among the partners are very homogenous and according to the theory. Hence, only minor improvements in the test module have been necessary. In the first testing campaign the stack power densities have been represented as a function of the specified operating temperature. But due to the heat generation of the proceeding exothermal reactions the average stack temperatures at given current densities differ strongly from the specified operating temperature. This temperature increase becomes more obvious with increasing current densities (up to 29°C at 700 mA/cm²). However, it can be seen that the temperature dependency among the partners is nearly the same which is indicated in Figure 12 as dotted lines.
Validation of TM 12: Operation under constant current
The operation under constant current in SOFC mode was performed at 300 mA/cm², 750°C and 0.5 H2 + 0.5 N2 / 4 air (NLPM/RU). These test conditions were set stable during the testing period of about 1000 h. The voltage behavior of the stacks of two partners (DLR, DTU) is shown in Figure 13. The OCVs of both stacks are very stable and slightly increase with time. This can be explained by an increase of the gas tightness of the sealing between stack and gas distribution plate. Moreover, nearly similar degradation rates of the stack voltages were obtained by both partners. This is due to the reliable control and the accurate monitoring of all interface conditions between stack and test station. All this knowledge for obtaining reproducible degradation values was used to optimize test module TM 12.
Figure 13: Long term behaviour of two stacks (DLR, DTU) at 750°C, constant current density of 300 mA/cm² and 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU)
Validation of TM 14: Thermal Cycling
TM 14 was investigated during the second testing campaign. Detailed information concerning the operating procedure is given in test module TM 14 and deliverable D 4.4 (Report of the second SOFC testing campaign). During thermal cycling two types of degradation can be distinguished: the degradation of OCV and the voltage degradation under constant current. Figure 14 shows the OCVs and stack voltages of three different stacks during 30 thermal cycles. For each cycle the stacks have been operated at 750°C for approximately 90 minutes at 500 mA/cm² with 0.5 H2 + 0.5 N2 NLPM/RU on the negative electrode side and 4 NLPM/RU air on the positive electrode side
Figure 14: Long term behaviour of different stacks (DTU, JRC, ENEA) during 30 thermal cycles (200°C - 750°C) at 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU) and 750°C
The OCVs of the stacks remain almost constant during thermal cycles which prove a very good gas tightness of the stacks with low thermo-mechanical stresses. The voltages of the stacks at 500 mA/cm² differ in a range from 3.9 V to 4.2 V which is in good accordance with the initial jV-curves of these stacks. Regarding the total degradation rates after 30 thermal cycles stacks with similar performance seem to degrade in a same range. There is also a tendency that stacks with lower performance show more stable behaviour thanks to the lower degradation rates. Similar to TM 12 the very accurate control of all stack input parameters is necessary in order to achieve reproducible degradation rates.
3.3 Results of final SOFC round robin test
At the end of the SOCTESQA project the final round robin test was dedicated to validate the developed test modules with focus on applicability, robustness, reproducibility and reliability. The test results of all optimized test modules which were applied in this final round robin test are very consistent with the former testing campaigns. Moreover, the observed differences of the stack behaviour are in good agreement with previous sensitivity analyses of the test input parameters. The following sections show some test results of all testing campaigns for some selected test modules. The detailed results are reported in deliverable D 4.5 (Report of the final round robin test).
Validation of TM 03: Current-voltage characteristics
The sensitivity analysis of the previous testing campaigns identified several test input parameters, especially the gas inlet temperatures, as important influences on the stack performance. Therefore, Figure 15 shows the stack power densities gained from the jV-characteristics of all stacks tested in the SOCTESQA project as a function of the average temperature of the inlet gases (Tav,gas in). Both values were taken at a current density of 500 mA/cm². The stacks have been operated at 750°C with 0.5 H2 + 0.5 N2 NLPM/RU at the negative electrode and 4 NLPM/RU air at the positive electrode.
Figure 15: Stack power density calculated at 500 mA/cm² as a function of the average temperature of the inlet gases (Tav,gas in) measured at 500 mA/cm²
Due to the optimization of the gas preheating subsystems the gas inlet temperatures for most of the stacks are in a range of 737°C and 753°C. Therefore, only a small dependency of the gas inlet temperatures on the stack performance is visible. The power densities of all tested stacks, except of two, have a variation of about 5.1 % between all stacks (highest and lowest value). This is very close to the quality variation threshold of 5 % given by the stack supplier in deliverable D 2.1 (List of SOC specifications). Hence,, a high quality and a high reproducibility of the results are proven.
Validation of TM 03 (jV characteristics) and TM 04 (EIS)
In order to validate the different test methods, investigations across the test modules and the several testing campaigns have also been performed. The reproducibility of TM 03 and TM 04 is shown in Figure 16 which summarizes the correlation between the low frequency resistances RLF measured with the electrochemical impedance spectra (TM 04) and the calculated area specific resistances of the jV-curves (TM 03) of the RUs of all stacks with increasing current. Both values were determined at a direct current density of 238 mA/cm² with gas flow rates of 0.5 H2 + 0.5 N2 NLPM/RU on the negative electrode side and 4 NLPM/RU air on the positive electrode side at 750°C.
Figure 16: Correlation between RLF of EIS and ASR of jV-curves of RUs at DC current of 238 mA/cm², 0.5 H2 + 0.5 N2 // 4 air (NLPM/RU) and 750°C
Most of the values are located quite near or underneath the dotted line which indicates the ideal theoretical relation of both test methods. However, the area specific resistances calculated from the jV-curves are slightly higher than the total resistances (RLF) determined by electrochemical impedance spectra. The deviations are mainly related to different temperatures of the two test methods. Usually the stack temperature when measuring EIS spectra is slightly higher compared to the temperature when measuring a jV-curve with increasing current density. Hence, the ASR values are higher compared to RLF. In the case of the jV-curve with decreasing current density, this is vice-versa. This effect is explained in detail in the deliverable report D 4.4 of the second testing campaign. Considering the different thermal conditions both test methods show a good reproducibility and prove the reproducibility of both test methods.
4 SOEC (WP 5)
Work package 5 (WP 5) “Solid oxide electrolysis cell (SOEC)” is dedicated to the evaluation and validation of the test procedures defined in WP 3 in the electrolysis mode through the testing of the cell/stack assembly unit specified and procured in WP 2. Similar to the above described SOFC section, the results of the validation processes enabled the optimization of the test modules with respect to SOEC operation.
4.1 Adaptation of test stations for SOEC testing
Depending on the initial status of each partner’s test station at the beginning of the project, this adaptation and commissioning step was composed of different types of actions and conducted differently in each laboratory. To list the specifications of each test station available in each laboratory, all the thoughts and discussions around deliverable D 5.1 helped the consortium:
• to choose the ElringKlinger (EK) stack as the most easily testable common commercial cell/stack assembly unit,
• to identify some important specifications to be fulfilled by the test stations for SOEC operation,
• to point out the last mandatory adaptations to be finalized for each test station.
Following test station adaptations, modifications and upgrades were realized successfully:
• mass flow controller and thermocouple changes,
• gas diffusion plates and voltage probes integration,
• gas preheating subsystems implementation and optimisation,
• upgrade of acquisition subsystem in particular to follow dynamic profiles,
• steam production stability’s improvement, which plays an important role to avoid noise on voltage and EIS signals in SOEC mode.
The details of these adaptations are described in the deliverable D 5.2 (Report of the initial SOEC test bench validation). Some results of the test adaptations and optimizations are shown in the following figures. Figure 17 and Figure 18 display the results of the optimisation of the gas preheating subsystem and the furnace thermal homogeneity of the different partners. These two diagrams show the differences between relevant temperatures at OCV for each partner’s test station in SOEC mode before optimization (Fig. 17) and after optimization (Fig. 18). The corresponding symbols are: TTP = top plate temperature (= Tstack in this project), TBP = bottom plate temperature, Tneg,in and Tpos,in = inlet gas temperatures on both negative electrode and positive electrode sides, respectively, Tneg,out and Tpos,out = outlet gas temperatures on both negative electrode and positive electrode sides, respectively. In particular the temperature homogeneity of the test stations at ENEA and NTU was clearly improved.
Figure 17: Differences between relevant temperatures at OCV for each partner’s test station in SOEC mode before optimization
Figure 18: Differences between relevant temperatures at OCV for each partner’s test station in SOEC mode after optimization
Figure 19 shows the improvements which have been achieved in terms of voltage stability in SOEC mode. Due to optimization of steam supply stability the voltage fluctuations were significantly reduced. Hence, the quality of the electrochemical impedance spectra signals could be increased especially in the low frequency (LF) region. Moreover, the electromagnetic interferences, which disturb the high frequency (HF) region in particular at CEA and DTU, were significantly minimized.
Figure 19: EIS diagrams of RU 3 in SOEC obtained at CEA before and after optimization (750°C, -12 mA/cm² and 0.5 H2O+ 0.125 H2 // 1 air (NLPM/RU))
The test station adaptations have led to further increase the reproducibility of results of the first and of the second testing campaigns. Moreover, during the project duration reliable interfaces between the stack and the test station were provided, operating conditions as similar as possible between all partners were achieved and dynamic profiles could be followed.
4.2 Results in SOEC mode
The following sections highlight selected results for the validation of the test modules which were mainly obtained in the first and the second testing campaigns in SOEC. The TMs were validated under SOEC operating conditions as realistic as possible, in particular at higher electrical current density in the second testing campaign. The overall, detailed results are reported in deliverable D 5.3 (Report of the first SOEC testing campaign) and D 5.4 (Report of the second SOEC testing campaign).
Validation of TM 03 (j-V characteristics) and TM 04 (EIS)
Above all, a very high stability and homogeneity of the test station components are required for a good reproducibility of results in particular for the very sensitive EIS technique (TM 04). Figure 20 shows the current-voltage characteristics of the stacks of the partners in SOEC mode at 750°C, 80 % H2O + 20 % H2 // air. Similar to the fuel cell operation the OCVs of the stacks are exactly the same. The current voltage curves are almost linear with some higher differences among the partners compared to the SOFC mode. However, the performance trend of the stacks is similar to the SOFC operation. It is shown that all test stations are able to supply the targeted H2/air compositions to the tested stacks within ± 5 % of deviation and in a stable way all along the test. However, the steam production is less stable which induces voltage fluctuations bothering the jV-curves and EIS measurements analysis.
Figure 20: Current voltage behavior of tested stacks of the different partners in SOEC mode at 750°C, 0.5 H2O+ 0.125 H2 // 1 air (NLPM/RU)
Due to the induced voltage fluctuations, a satisfying steam supply stability is required and it is recommended to maintain voltage fluctuations below ± 10 mV for lowest results dispersion especially during EIS measurements which are very sensitive to this phenomenon (see Figure 19). Depending on the steam supply subsystem, a kind of nitrogen “dilution” can be used to enhance steam supply stability and cleanness of the EIS low frequency region. Additionally, the use of numerically fitting of the spectra by an equivalent electrical circuit for ASR determination can be efficient to avoid this drawback when low frequency disturbances remain and to succeed in determining the total resistance. This point is also very important to allow the determination of the ASR at one particular current in a small zone of the jV curve. In case of remaining voltage fluctuations when jV-curves are recorded (TM 03), it is preferable to determine ASR values on a large zone of the jV-curve (0.1 A/cm² at least) for a lowest results dispersion among partners. A linear fit or a polynomial one can also be used depending on the global shape of the jV-curve (see Figure 21 below). In case of voltage fluctuations and thermal inhomogeneity (first testing campaign), it is lower when a larger zone of the jV-curves at both current densities or a linear fit of the full jV-curves is taken for ASR determination. If voltage stability and thermal homogeneity are improved, the dispersion difference between the different calculation methods is reduced.
Figure 21: ASR dispersion among partners obtained from jV-curves in SOEC mode with different ASR calculation methods.
As already mentioned in section 4.1 (Fig.17 and Fig.18) thermal homogeneity strongly influences dispersion of results and hence the reproducibility of the stack results. The correlation with performance is presented in Figure 22, which shows a clear dependency of the stack ASR (linear regression) from the inlet gas temperatures. So it is highly recommended to measure temperature at different locations in the stack (top and bottom plates, gas inlet and outlet on both sides for instance) and try to limit the difference between those temperatures below ± 6 °C for a lowest results dispersion.
Figure 22: Stack ASR in SOEC mode at 500 mA/cm2, 750°C and 0.5 H2O + 0.125 H2 // 1 air (NLPM/RU) as a function of average temperature of the fuel and oxidant inlet gases
Validation of TM 12 (operation at constant current) and TM 13 (operation at varying current)
In steady-state operation at constant current density, it appears that internal pollution brought by the inlet gases and the test station itself can increase voltage and ASR degradation rates during long-term operation (TM 12 and TM 13). Figure 23 shows that initially stack voltage evolution with time at a constant operating point was very different between CEA and EIFER leading to very different voltage degradation rates. Some issues as temperatures homogeneity, gases and water quality and steam flow stability have been invoked.
After changing the steam generator, the preheating loops and the purity of hydrogen gas (from grade 3.5 to 6.0) CEA measured a much lower degradation (from 9%/1000h to 3%/1000h) with nominally identical stacks. This improved voltage degradation rate is probably due to the reduction of contamination. The results of EIFER show that the degradation rate depends on test duration. So the necessity to extend the test to a reasonable duration is clearly recommended in both TM 12 and TM 13. To complete that point, Figure 24 below shows that after reduction of the internal pollution on CEA test station; voltage degradation rate value obtained in this laboratory was decreased for a higher reproducibility between both testing partners.
Figure 23: Long term behavior of 4 stacks of two partners (EIFER and CEA ) in SOEC mode at 750°C, constant current of -300 mA/cm², 0.5 H2O+ 0.125 H2 // 1 air (NLPM/RU)
Figure 24: Voltage degradation rates calculated during the initial test bench validation (in blue) and the first testing campaign (in orange) under constant current operation at two testing partners EIFER and CEA with two different inlet gases qualities at CEA
For a comprehensive understanding of the SOC assembly unit degradation behavior the determination of different degradation rates (voltage, ASR from jV-curve and from EIS) is recommended. For instance the continuous recording of voltage allows to follow the degradation dependency with time (continuous or by sequence ...) whereas the EIS diagrams recorded before and after a long-term operation allow to separate ohmic, polarization and gas concentration resistance degradation rates and to give ideas about possible degradation phenomena.
4.3 Results of final SOEC round robin test
The final round robin test allowed to fully validating five test modules in single SOEC mode. The results are described in detail in deliverable D 5.5 (Report of the final SOEC round robin test). It is worth noticing here that globally results dispersion was a little bit higher for this final testing campaign. In contrast to previous testing campaigns which showed satisfying improved dispersion due to efficient test station adaptation/improvement, the results of the final round robin test scatter slightly higher. This can be attributed to the fact that the corresponding tests were performed on stacks from different fabrication batches and the lower robustness of the stacks coming from the last production batch. However, the results are reproducible to the results of stacks from the same batch of previous testing campaigns. Therefore, despite the higher result dispersion within the final round robin test, the robustness of the final test modules was confirmed.
5 Combined SOFC/SOEC (WP 6)
The objective of WP 6 (combined SOFC/SOEC) is the evaluation and validation of the test procedures for cells and stacks in combined SOFC/SOEC mode defined in WP 3. This includes similar to WP 4 and WP 5 the validation of the applicability, the reproducibility and the repeatability of the test results but also the sensitivity analysis of the corresponding test input parameters in reversing SOFC/SOEC operation.
5.1 Adaptation of test stations for combined SOFC/SOEC testing
A specification of the requirements for test stations with respect to combined SOFC/SOEC testing was made in the deliverable D 6.1 (Specification list of test benches for combined SOFC/SOEC Operation). The most important requirements are steam generator, fuel gas quality, electronic load, power supply, equipment for electrochemical impedance spectroscopy, and stability and noise levels. Basically, all these items are also considered in the single mode operation in SOFC mode and SOEC mode, and hence dealt with in WP 4 and WP 5, respectively. A few further items for special attention for the combined SOFC/SOEC operation have been identified. They deal with the time for switching and the automation of switching and with limitation in reversible and continuous measurement of the jV-characteristics across OCV.
The time required for switching between SOFC and SOEC mode (and vice versa) depends on the maximum ramp rate in the shift (any change in temperature, gas, and humidity change) and the time required for any manual switch (shift of gases and shift between SOFC electrical load and SOEC power supply). When changing from high humidification (SOEC) to a lower value (SOFC), in some cases it will take some time (up to 1 hour) before all the steam is removed and the stack voltage value is stable again. A maximum time for switching of 2 hours was specified and a required holding time at OCV has been accepted in the requirement for the test stations.
Five partners have completed the initial test station validation program (described in D 3.2) including combined SOFC/SOEC operation. The test results from the combined operation have been reported in D6.2 (Initial test station validation in combined SOFC-SOEC mode). The period with "combined operation" included followed the sequence: SOFC-SOEC-SOFC-SOEC, with 24 hours for each step, including a 2 hour period for switching. And overview of two SOFC/SOEC switching procedures is given in Figure 25 below.
Figure 25: Overview of the SOFC/SOEC switching procedure with two periods in SOFC and two periods SOEC
All the test stations of the partners are able to follow the planned set points. All the test laboratories have shown that they are able to perform the switching between the operation modes within the specified 2 hours period. When returning to the second SOFC step and the second SOEC step, all stacks were found to return to the same performance level, which means that the results are at least reproducible for same stack, same test station within a short period of days. This is shown for the SOFC mode in Figure 26, and for the SOEC mode in Figure 27. These diagrams show for each partner the cell voltages at +/- 300 mA/cm2 and stack temperatures for the 1st and 2nd SOFC operation period in SOFC and SOEC operation, respectively. All partners have a high reproducibility of own results, whereas somewhat larger differences are observed between partners
Figure 26: Average voltage of RUs and temperature of stacks of the partners for the 1st and 2nd SOFC operation period
Figure 27: Average voltage of RUs and temperature of stacks of the partners for the 1st and 2nd SOEC operation period
For a simple combined testing and short-term testing (< 5 days) automation of the switching is not required. However, in the further work for validation of the application specific test program (power-to-gas-to-power) automation of switching is foreseen to be required in order to be able to carry out the work within normal work hours, and to be able to carry out the actual test without an unreasonable high work load involved (for manual shifting). During the second period of the project, NTU and DTU were mainly responsible for the testing of SOC stacks in combined SOFC/SOEC mode. Both partners have succeeded in modifying their test stations with fully automatic SOFC/SOEC mode switching. This allowed for long term testing with numerous switches and switches at any time of the day or the week. Moreover, laboratory manual working resources have been saved significantly.
The initial test station validation results also indicated that NTU had low gas inlet temperatures. During the second period of the project, NTU has improved their gas pre-heating system and increased the negative electrode gas inlet temperature (Tneg,in) for about 20 °C. The corresponding details were reported in the deliverable report D 5.3 (Report of the first SOEC testing campaign). DTU also identified and solved the issues associated with the stack voltage fluctuation in SOEC gas composition and low stack voltage when shifting from SOEC mode to SOFC mode. The results are reported in deliverable D 6.3 (Report of the first testing campaign in combined SOFC-SOEC).
5.2 Results in combined SOFC/SOEC mode
After the initial test station validation campaign of the first project period, the first and second test campaigns were performed in combined SOFC/SOEC operation in order to optimize the test modules of WP 3. In the following section the most important results of these validation processes are reported. The results were mainly generated by the two responsible partners NTU and DTU.
Figure 28 shows the stack voltage and current density evaluation as function of operating time of the stacks tested in the first test campaign. The detailed results of this test campaign are reported in D 6.3 (“Report of the first SOFC/SOEC testing campaign”). The stacks were tested at 750°C with 50% N2 + 50% H2 in SOFC and 20% H2 + 80% H2O in SOEC supplied to the negative electrode compartment. In both modes air was supplied to the positive electrode compartment. The stacks show very similar performance at the beginning, but different degradation behavior during the combined SOFC/SOEC mode operation. Nevertheless, after the initial period (first 200 hours) the two stacks show quite low and similar voltage degradation rates.
Figure 28: Stack voltage and current density evaluation as function of testing time. The stacks were tested in the first test campaign at 750°C under 0.3 A/cm2 and 50% N2 + 50% H2 in SOFC mode and -0.3 A/cm2 and 20% H2 + 80% H2O inSOEC mode.
In the second test campaign, validation of the test modules by testing the stacks under “harsher condition” was suggested. The current densities in the test program TP04 (“Test program for the application of combined SOFC-SOEC in a power-to-gas-to-power system”) were changed from -0.3 A/cm2 to -0.5 A/cm2. The results are presented in Figure 29. The entire results are reported in D 6.4 (Report of the second SOFC-SOEC combined mode testing campaign).
Figure 29: Stack voltage and current density evaluation as function of testing time. The stacks were tested in the second test campaign at 750°C under 0.5 A/cm2 and 50% N2 + 50% H2 in SOFC mode and -0.5 A/cm2 and 20% H2 + 80% H2O in SOEC mode.
Higher overall voltage degradation rates were observed on both tested stacks when operated at higher current densities. However, stacks tested at the same institute show overall similar degradation trend, for example, the stacks tested at DTU show high initial voltage degradation rates in the first 200 hours and stabilize afterwards. Stacks tested at NTU show almost linear degradation in the entire testing period. Nevertheless at the end of the tests (the last 150 hours), the stacks tested at DTU an NTU showed similar degradation rates when operated under the same test conditions. The different initial degradation behavior between stacks tested at DTU and stacks tested at NTU may be caused by different gas inlet gases qualities with different impurity levels.
5.3 Results of final SOFC/SOEC round robin test
In the final round robin test campaign, besides validation of the same test modules, DTU and NTU have carried out the final test on combined SOFC/SOEC operation in order to check the robustness of the testing procedures. The results of the combined SOFC/SOEC mode operation have been reported in deliverable D 6.5 (Report of the final round robin testing test).
Figure 30 shows the stack voltage and current density evaluation of DTU and NTU stacks as function of testing time of the final round robin test. Similar degradation behavior/results were observed compared with the second testing campaign.
Figure 30: Stack voltage and current density evaluation as function of testing time. The stacks were tested in the final round robin campaign at 750°C under 0.5 A/cm2 and 50% N2 + 50% H2 in SOFC mode and -0.5 A/cm2 and 20% H2 + 80% H2O in SOEC mode.
Moreover, as can be seen from Figure 31 and Figure 32, stacks tested at DTU and NTU in different testing campaigns show very good repeatability. Only minor difference can be seen for the stacks tested at the same partner, which can be explained by differences in the stack hardware itself. The results prove the very good robustness of the testing procedures, even though different initial degradation behavior is observed between partners, which may be related to different impurities from the test stations or to different inlet gases qualities supplied to the stacks. Longer time operation is therefore recommended for better evaluation and comparison of stacks degradation.
Figure 31: Stack voltage and current density evaluation as function of testing time. The stacks were tested at DTU in the second test campaign and the final round robin test campaign at 750°C under 0.5 A/cm2 and 50% N2 + 50% H2 in SOFC mode and -0.5 A/cm2 and 20% H2 + 80% H2O in SOEC mode.
Figure 32: Stack voltage and current density evaluation as function of testing time. The stacks were tested at NTU in the second test campaign and the final round robin test campaign at 750°C under 0.5 A/cm2 and 50% N2 + 50% H2 in SOFC mode and -0.5 A/cm2 and 20% H2 + 80% H2O in SOEC mode.
Potential Impact:
The ultimate objective – to come up with procedures at the end of the project that are not only validated inside the laboratory, but are already shaped towards the requirements of regulations codes and standards as well as industrial productivity and reliability – has been successfully achieved. The interaction with relevant industrial stakeholders, the broad dissemination of the project results and the liaison with standards developing organizations (SDO) have been important issues in the project. Figure 33 shows the overall consortium of the project. The test procedures incorporate the inputs from both entities and foster worldwide discussion and awareness of the topic. In this way the maximum exploitation of the project outcome was achieved. The details of dissemination and exploitation of the project results are reported in deliverables D7.2 (“Interim report on liaison and dissemination activities”) and D7.3 (“Final report on liaison and dissemination activities”).
Figure 33: Overall consortium of the SOCTESQA project
1.1 Liaison to SDO
SOCTESQA has carried out a comprehensive survey of ongoing standardization activities in the field and has entered into contact and liaison with the main bodies currently working on regulations for hydrogen and fuel cell technologies. Essentially, these are grouped under the international bodies of the ISO Technical Committee 197 on hydrogen technologies – which looks mainly at safety issues of electrolyzers and hydrogen handling – and the International Electrotechnical commission (IEC) TC105 on fuel cell technologies. The latter is much more focused on the technology and the definition of practical guidelines in terms of system performance, installation and characterization. Also CEN/CENELEC, the main European SDO in the field, has been liaised with, through an official agreement.
The work in SOCTESQA is considered most in line with IEC, and in fact the Technical Specification on solid oxide fuel cell and stack test procedures (issued in 2014 as IEC TS 62282-7-2) has been a fundamental seed document for further elaboration of the SOCTESQA procedures, which however break new ground by providing guidelines also for the operation of SOC stacks in electrolysis mode (SOEC) and reversible mode (combined SOFC/SOEC). After these initial approaches a formal Liaison was initiated with the IEC TC105. This underpinned the initiatives on behalf of SOCTESQA project taken up by the IEC, i.e. submission of a New Work Item Proposal (NWIP) and subsequent initiation of the new Working Group (WG) 13 on Standards for fuel cells operated in regenerative mode. The leader of this work package 7 of SOCTESQA also acts as project leader for the first standard and as convener for the entire WG 13 of IEC. Through this WG and the formal liaison, the entire SOCTESQA consortium was able to participate in and guide the activities that will lead to the publication of Standard 62282-8-101: “Solid oxide single cell and stack performance including regenerative operation”, which incorporates all the work carried out in the SOCTESQA project.
Through the IEC TC105 committee, liaison is also ensured with ISO’s TC197 on hydrogen technologies, which will be monitored through the former activity. Simultaneously, CENELEC, the European technical standardization body, has also initiated a new working group on Hydrogen. The scope of the working group covers the production of hydrogen through electrolysis and the transportation, distribution and usage of that hydrogen in pure form or as a natural gas dominant mixture (H2NG). In addition, actions in cross-cutting fields such as safety and training of personnel are identified. The final objective of this working group is to set a long term collaborative framework (liaison) with major bodies for strengthening cooperation between regulatory work, standardization work and RDI programs (e.g. European Commission, JRC, FCH2 JU, IEA, ISO, IEC). DLR has joined this working group, namely task force 2: electrolyzers, and participated at several meetings with the background of transferring the results and experience so far achieved in SOCTESQA. Moreover, DLR has entered into a formal liaison with CENELEC to monitor the progress in this working group on behalf of the SOCTESQA consortium. The corresponding liaison agreement between DLR and CENELEC was signed in Sept. 2015.
1.2 Liaison to industry
Interaction with industry has been part of SOCTESQA since the time of writing the project proposal: a number of letters were gained at that time where key industrial players manifested their interest in following the project’s activities and achievements. These industries were chosen to constitute the project’s industrial advisory board (IAB).
At the beginning of operations, fact sheets were compiled by SOCTESQA partners and sent out to the IAB for gaining technical input as to the operating conditions of SOC systems in the applications targeted: systems for combined heat and power generation (µ-CHP), auxiliary power units (APU) and electrolysis systems mainly. All industries of the IAB have been invited to participate in the first of two workshops organized by SOCTESQA, which was held on December 15th, 2015, in Naples, in conjunction with the 6th European Fuel Cell Technology & Applications “Piero Lunghi” conference. The scope of this workshop was to present SOCTESQA activities and trigger input from international industries and SDO for a hands-on discussion on how to formulate robust, univocal procedures relevant for industry and that can facilitate quality assurance in the production process and confident market entry.
During the project progress the IAB as well as all industrial stakeholders were kept up to date with SOCTESQA progress through personal exchanges and the SOCTESQA newsletters. Finally, the SOCTESQA project outcome workshop reached out to industrial stakeholders directly, through a booth and two presentations at the Hannover Fair, 24-28 April 2017, the biggest exhibition on fuel cells and hydrogen technologies in Europe. This liaison/dissemination event is reported more detailed in deliverable D 7.4 (“Project outcome workshop”).
1.3 Dissemination of project results
The basic tool for dissemination, namely the project website with a “corporate” identity, was developed early in the project and went on line as planned (see also deliverable D 7.1:”Project website”). A list of dissemination activities carried out by SOCTESQA partners with details regarding dates, presentation types and titles, and outcome is given in deliverable D 7.3 (“Final report on Liaison and dissemination activities”).
The SOCTESQA project was intensively promoted, with a number of items disseminated to the general and scientific public. Two peer-reviewed articles and three articles were published in high-profile scientific conferences, like the Int. Conference on SOFC, the European SOFC&SOC Forum and the European Fuel Cell Conference. Moreover, the results of the project were broadly disseminated to the scientific audience with many presentations, e.g. 7 presentations in scientific conferences and 8 presentations in workshops or key international events including assemblies of the International Electrotechnical Commission (IEC). Especially the latter events have generated the highest dissemination impact. Thanks to this engagement and formalized in the liaison mentioned above, the project has provided the procedures for solid oxide cell and stack testing that are to become internationally applicable standards, for the benefit of industrial production all round.
As mentioned above, a crucial dissemination event that catalysed the interaction with industry and standardization, an important milestone within the SOCTESQA project, was the workshop organized on 15.12.2015 in the frame of the 2015 European Fuel Cell “Piero Lunghi” conference in Naples/Italy. Keynote speakers from standardization bodies (JRC, IEC and Japanese electrical manufacturers’ association – JEMA) and collaborating projects such as STACKTEST (FCH-JU) interacted with industry (CERES Power, Elcogen) and academia to demonstrate achievements, identify gaps and bottlenecks and ways forward in a lively debate. Moreover, two presentations were given to general audience in the frame of the final workshop including a project booth at Hannover Fair 2017.
Altogether 14 posters were presented in scientific conferences, workshops, fairs and the FCH-JU review days. Additionally, two newsletters were disseminated to relevant stakeholders. As already described above two project flyers and all important test module documents were handed out to scientific audience at Hannover Fair 2016 and 2017. Moreover, interaction with other European and National funded projects was established by participating and exchanging of the results at project meetings. And finally, all public deliverable reports and all test modules were uploaded on the project website.
The results and knowledge generated by the SOCTESQA project have led to a straightforward and open approach for the universal adoption of harmonized, quality-assuring procedures, which is the key for the successful technological progress and market implementation of the ceramic solid oxide technology and the corresponding entire industry.
List of Websites:
http://www.soctesqa.eu/
Contact details: Dr. Michael Lang, German Aerospace Center (DLR), Pfaffenwaldring 38-40, 70569 Stuttgart, e-Mail: michael.lang@dlr.de Tel: +49 711 6862605
