Progress of work since start of the project in 2008
At the start of the project it was of imminent importance to specify the requirements, materials and components as well as test procedures used and applied within DECODE. Also a definition of test cells and stacks for all activities was of paramount importance in order to ascertain a coherent activity and comparable results. In the discussions, which are organized within WP2, it was decided to follow a pragmatic approach to enable the comparability of results but also to ascertain the rapid establishment of testing capability within DECODE. Therefore the specifications were performed specifically for each work package taking into accounts the different requirements and degradation mechanisms effective in the various components of PEFC. The work of this WP is now almost finalized and summarized in the deliverable reports “D2.1: Report which materials and components are selected, description of the test cell for standard test” and “D2.2: Report about test conditions for standard tests incl. operation modes”. The specifications started from input given by the automotive industry about operation conditions relevant for automotive applications. Based on these conditions, test conditions were defined for work package 3-5 for standard tests that should at the same time especially enhance the degradation of the component investigated in that WP (membrane, reaction layer, gas diffusion layer and cell) but also ensure a comparability of results between the different work packages. The test procedures for continuous operation and dynamic operation (load cycling) are based on the results of the FCTESQA project. Also the project test cells and stacks were defined. For single cell measurements every partner uses its own single cell, mostly specialized for certain measurements. For stack measurements DANA provides the material in form of short stacks. A membrane by Solvay-Solexis and a GDL by SGL were defined as the single cell standard material. As catalyst coated membranes (CCM) were not available for tests as originally expected, tests with CEA home made CCM and commercial CCM by Solvicore had to be performed before the project test material could be defined. These tests are also terminated now.
According to the project plan the specification phase is be terminated at this point, but it is clear to all partners that modifications to these specifications may be necessary in the future with knowledge gain and shifting focus from understanding to improvement. Therefore updates of the specifications will be necessary and will be reported in future.
For all components state-of-the-art reviews (reviews) are being prepared to have a solid knowledge base from literature to derive the strategies for the improvement phase where strategies to incorporate durability gains are derived. Due to the delayed start of the project but also by the size and extension of the available literature on this topic with many ramifications the reports have been delayed significantly. Most or the reports are available in the final state and no significant impact of the delayed on the project success is expected from the delayed deliveries. The reports on the state-of-the-art will be made public by peer reviewed publications of the project in a suitable journal.
The focus of the membrane and electrode activities in 2008 has been the definition of the CCM material based and the testing procedures to ascertain a good comparability and successful testing in all laboratories. This work which has been mainly preformed in collaboration between CEA and Solexis yielded first ageing results of differently prepared CCM but – more important- defined the reproducible material basis for further investigation at all DECODE laboratories involved in membrane and electrode investigations. Now commercial MEAs are available for comparative in situ tests and laboratory scale CCM formulation is also available for ex-situ characterizations. So, more results will be obtained by several partners in the next weeks/months. Solvay
Solvay Solexis performed the synthesis and extensive characterisation on theAQUIVIONTM membrane used within the project including proton conductivity, water uptake, mechanical stress and hydrogen crossover measurements. In addition, performance data of CCMs and chemical and physical characterisations of the necessary ionomer dispersion are available. The membrane was also preliminary characterised by CEA with SANS (small angle neutron scattering). The necessary detailed analysis to provide indications for further developments to Solvay-Solexis (related to deliverable D3.3) should be possible.
A special focus of the DECODE project is on advanced modelling of porous media and the experimental characterization of gas diffusion layers. On major modelling activity, discussed in this paragraph, is the use of the relatively novel Lattice-Boltzmann technique for the description of porous media and complex fluid systems. In the Lattice-Boltzmann models the fluid consisting of fictive particles, and such particles perform consecutive propagation and collision processes over a discrete lattice mesh. Due to its particulate nature and local dynamics, LBM has several advantages over other conventional CFD methods, especially in dealing with complex boundaries, incorporating of microscopic interactions, and parallelization of the algorithm. In 2008 an adaptation of lattice-Boltzmann tool waLBerla for an efficient parallelisation of the free-surface application was performed. Key assignment for this task is a local management of gas volume changes, and merge of independent gas regions. In addition also new algorithms for describing diffusion in porous media were developed. As a basis for the computational domains of the molecular dynamics of the MPL and the lattice Boltzmann model of the substrate, Opel provided a 3D synchrotron radiography data set of SGL GDM and handed it out to the partners who will convert it into computational domains for their respective models. To make the domain generation more convenient, Opel also provided a small set of image data carrying all relevant features of the original GDM image. Furthermore, Monte Carlo simulations of porous media are an important topic of DECODE based on structural models of the porous media. Three different type structures: fiber-, paper- and spaghetti- structures were generated and investigated. The distribution of PTFE was investigated on model surfaces and the models were verified through plausibility considerations of liquid water interactions.
GDL experimental characterization was performed by x-ray photoelectron spectroscopy, infra red spectroscopy, water porosimetry (in development) of artificially and naturally aged GDLs. Furthermore, neutron and synchrotron imaging and tomography results of fuel cells with relevance to GDL behaviour are generated in other projects, but can be used for DECODE. The methodological basis for investigation of porous media is exceptional in this project. It is expected that the experimental validation of the models will be achieved and that novel scientific results will be generated. This is especially important as the advanced characterisation of the diffusion media is lacking behind compared to the other fuel cell components.
For the investigation of bipolar stacks DANA has provided 5 cell short stacks for Round Robin tests and for durability test. The first ageing results have been generated and are being analysed. It is also planned to use the GDL in the stack for investigating naturally aged GDL (link to WP4) with the methods described above. Within the project, an existing bipolar plate design "FRIEDA" from DANA Victor Reinz was selected for the test stack setup. The tested materials are uncoated SS316L, SS316L with 100nm electroplated gold coating, uncoated SS905L, and milled graphite composite. It is planned to iterate the different materials at different partners for testing under same conditions. The conditions of the durability run were defined within the DECODE Project with input from DLR, Opel and DANA. The tests are still in progress at the partners DLR, CEA, VOLVO, ZSW and DANA.
Up to now four Stacks have been tested and fully analyzed. Two stacks finished the durability run and the post mortem analysis of the components is ongoing. Two stacks to compare the test benches from each partner (round robin test) are still in operation. As well as the ongoing analysis of a single cell stack which was analysed under the lead of ZSW in the CONRAD (Cold Neutron Radiography) setup in Berlin.
The characterization and investigation of bipolar plates in durability measurements has started already and is therefore completely on schedule. It is expected that all expected results will be achieved.
Several publications and presentations concerning DECODE have already been realized. These are summarized below. DECODE was present in the “International Workshop on Accelerated Testing in Fuel Cells” with contributions of different partners (CEA, DLR, DANA, JRC etc. ). The coordinator was in the scientific committee of the work shop and the DECODE logo was on all announcements.
DECODE was presented at:
“International work shop on Accelerated Testing in Fuel cells” in Ulm, Germany in October 2008.
“EUCAR-Workshop” in Brussels in September 2008
“European Fuel Cell and Hydrogen week” in October 2008
“EUCAR Conference & Reception at Autoworld” in Brussels on November 26th, 2008.
In Journal HZwei 01 2009, on page 16-17
In addition, the partners in the project have participated in several conferences in Europe.
In general, the project has been successfully started and all work packages are active in the tasks planned. There are significant delays in some tasks, in particular in the preparation of the state-of-art-reports. Delays are also evident in the investigation of electrodes and membranes in CCMs. However, the overall goals are not endangered and it is expected that the delay can be completely recovered in 2009.