Electrolyte Standards, Peter Fischer of Fraunhofer ICT Karlsruhe
Peter started by posing the question – is it possible to have standard electrolytes ? If the answer is yes then what are they – if no then what else in a VRFB structure might be standardised – eg membranes, stack designs etc. Furthermore are there other aspects of VRFB characterisation – eg electrolyte production or testing procedures that might be standardised ?
Peter returned to the general question of electrolytes and outlined the three main types of Vanadium electrolytes currently in use –
First is the classic pure sulphuric acid based electrolyte – this has the lowest Vanadium concentration and uses only sulphuric acid in the mix. This is the composition originally introduced by the UNSW group of Professor Maria Skyllas-Kazacos in 1985. I believe that composition was patented at the time but more than 30 years on seems generally usable without patent restrictions. This is the composition that is being used at ICT Karlsruhe and is, so far, the only type that Peter has been able to undertake comparative testing of numerous manufacturers samples on (13 manufacturers’ samples.)
With a somewhat higher Vanadium concentration, and consequently energy density, is the patented mixed acids electrolyte developed by Pacific Northwest National Lab in 2010. This offers the possibility of greater energy density and also wider operational temperature performance (0-50deg C) as compared with the classic sulphuric-only electrolyte (40deg C is considered its upper operational limit). The downside is that only three organisations have been granted the rights to use the patent – UniEnergy Technologies (founded by the PNNL inventors), WattJoule and, correct me if I am wrong, Pure Vanadium Corp.
The third type is the so-called High-Concentration Sulphuric acid electrolyte – from what I understand this uses other additives to control the precipitation problem (see Vanitec 2ESC) which otherwise limits the classic electrolyte formulation. From what I understand this has not yet been used in any commercial systems and is the subject of patent protection.
Restricting his analysis to only the classic electrolyte formulation that was required for the Fraunhofer ICT 8MWh battery Peter described the variation in properties that they observed amongst samples of electrolyte that they assessed for their project. Firstly the ratio of V(3+) to V(4+) in solution – the Fraunhofer request was to have a slight excess of V3+ as over time whilst standing in contact with air it oxidises to V4+ -as can be seen below there is actually a surprising range of ratios observed in the tested samples – but these are all apparently working electrolytes so perhaps this requirement is not a strong one.
Similarly there are surprising variations of Vanadium concentration across electrolytes that are all nominally quoted at 1.6M concentration. Given the significant cost of Vanadium you might think that these variations would be to the low side but actually there are quite a few that are significantly above 1.6 – even rising to 1.75M. This may not a big problem – it may simply be that Fraunhofer requested 1.6M electrolyte but the manufacturers sent their best solution which had the highest concentration Vanadium in it.
There would seem to be no intrinsic problem in making electrolyte solutions to tighter specifications – the practical question is instead ‘if you are a manufacturer that can use 1.75M solution in your batteries, would you be prepared to lower the specification of your batteries, just so that you could use a standard electrolyte’. 10% greater size for a car battery might be an issue, but might not be so significant for a fixed utility scale installation – this question might only be answered once the operational advantages of having an electrolyte standard – eg better reproducibility, lower production cost and greater electrolyte characterisation are known.
As concentration (of the sulphuric acid primarily) also has a big effect on the electrolyte conductivity some battery structures might actually require higher concentrations than others – in principle high concentration electrolyte suppliers could be probed on why they make their electrolytes higher concentration, but they might consider that this might reveal too much about the rest of their battery design. Similarly you might ask the producers of low concentration electrolytes why they have done so and whether they could accept higher concentrations (assuming they also make batteries). As was suggested in the subsequent discussion perhaps the way to proceed is simply to put out a request for samples of manufacturers working solutions, fully characterise them, and then propose standard bands of concentrations of the required components based upon these working solutions.
Having seen what an standard electrolyte should contain Peter then turned to what it should not contain – impurities such as copper can rapidly plate the inside of the battery structures causing major problems – but others impurities can also affect the operation of the battery over long time scales. Peter listed 4 main sensitivities – electrochemical kinetics, electrolyte stability, thermal stability (coagulation/precipitation as described previously) and membrane contamination.
He illustrated the point with an example of the impurity fingerprint of different electrolyte samples – dirty (red) and clean (blue) shown in the graph below.
Certainly there is significant variation, believed to be coming from the V2O5 feedstock – for example Vanadium feedstock derived from fly-ash processing are seen to contain high levels of Nickel, whilst those from slag processing contain other heavy metals – for example the Chromium shown above. Many of these levels exceed the Fraunhofer requirements, shown by the green bars.
There are also some surprising differences in other elements – eg Boron and Uranium where no specific requirement was set. As was mentioned in the committee discussion it may be a scientifically interesting (and indeed highly challenging) question to establish the failure mode, and thus the tolerable concentration, of every single single impurity species in this table but this would perhaps take a very long time to establish experimentally. Instead further discussion between all the manufacturers, of both batteries and electrolytes, on what they currently make and what currently works looks necessary to move forward on workable maximum allowable impurity proposals.
The question of whether electrolyte standards are required as part of a leasing financial model are returned to in part 3.