The Editor engages with some of the best of the best in European battery science as they try to forge a better strategy to determine how long batteries will last.
There are some things that people just won’t put up with. It seems to be a national and cultural thing; for the Germans, it’s late trains.
I know this is true because I’ve witnessed it with my own eyes and ears. I was coming back from the seminar organised by the EU Battery Lifetime Prediction Seminar held in Clausthal in January.
It has snowed hard and my train to Hannover was delayed by 20 minutes. The snow was deep. The platforms were cleared, but my fellow travellers were very angry. Old ladies were using their mobile phones furiously and telling their friends and families how late they were going to be.
I can only guess why the train was late, but 20 minutes in those circumstances seemed pretty reasonable. The same week a much smaller snowfall brought the south east of England to a standstill!
It’s no wonder then that Heinz Wenzel, who had organised the meeting at the University, has a contract with the German railway operator to determine how long a particular type of battery will last on the lines that serve Hannover into the Hartz region, where my delayed train was coming from. The operators aren’t concerned about getting a 20-year life or a ten-year life they just want to know how long, given a particular set of operating conditions, the battery will last. Because what they don’t want, as no automobile owner wants, is the surprise of a train not starting at 4am when it’s -20ºC and there’s no maintenance crew to be found.
I wonder if Dr Wenzel will solve the problem? The question of how well batteries will perform, and how long they will last, is a huge stumbling block for the battery business — and certainly an obstacle to batteries’ integration into energy storage systems. It’s an issue that’s bothered the renewable people in Europe and elsewhere for some time —long enough to get a multi-partner study, known as the benchmarking programme off the ground. Its purpose is to evaluate and, if necessary, develop new test procedures which reflect the real operating conditions in renewable energy systems, and to make recommendations to manufacturers and users on how to select the most appropriate components for a particular installation.
All the time, of course, there was this nagging little voice in my head saying: isn’t this the same problem that the hybrid car people are coming across? It’s partial state of charge operation all right!
With less than 40 people present, this wasn’t your average battery meeting. It was meant to stimulate discussion, and it certainly did that. There were good reasons to hold the meeting in Clausthal – it’s where Wenzel lectures in electrical engineering. The place even has its own BESS, just 250kW, and the whole locality is intimately linked with energy storage. The Hartz mountains were a rich source of lead, zinc and silver and from medieval times onwards, and the Germans used water-based energy storage (reservoirs and canals to pump water from the mines). Wenzel teased his audience by inviting in the mechanical engineers from the University, who gave a brief history of fatigue analysis. Well, if only life in the battery business really was as simple as applying a load stress over time and waiting for the battery to fail!
If only battery science had had in its midst the brilliance of the German railway engineer Wohler who, over 150 years ago, worked out how to determine stress and loading curves on mechanical objects and give extremely accurate predictions on how long railway axles would last. The fact is that batteries and their material components are operating at much higher activation energies then the wheels of train, or the suspension components of the latest over-powered automobile. Sure it was fun looking at all the weird and wonderful machines in the engineering workshop, designed to do 100,000 miles-worth of damage to train wheels in ten days, but it’s another world. This fact wasn’t wasted on Eberhard Meissner of JCI. What if you work metals closer to their activation energies, he asked. Did the rules still apply? The answer of course was no.
So how do we get a better idea of how batteries are going to last in service?
It’s a big problem. Or is it ? These battery specialists have taken the view that the differing operating conditions of renewable energy systems gives rise to so-called “battery stress factors”, and it is the sum total of these that produce the ageing effect which eventually kills the battery. But what really kills the patient? To give an analogy, is hypertension in the patient a primary or secondary factor in his coronary heart disease?
Vojtech Svoboda of ZSW listed a number of parameters in the life of a lead acid battery as “stress factors”. They include partial cycling, time at low state of charge, temperature and average time between charges, as well as the number of amp hours put through a battery. But the real interest I think is being able to link the stress with the damage to the system. The damage of course is active material shedding, water loss, gassing, corrosion and sulphation; and Svobda’s work has tried to link one with the other and weigh the stresses: the operating conditions against the damage. So one should be able to calculate how “exercising” a battery (my term) impinges on the lifetime of the system. In the ZSW scale of stress, charge factor gets a one on the scale but high discharge rate gets a five.
But hang on a moment. Don’t we know this already? Aren’t the conditions being imposed on PV system batteries, wind/PV/diesel hybrid batteries and the like just the same as those being imposed on hybrid electric cars? Huge discharges on acceleration, huge current inrushes on regenerative braking, and temperature rise as a consequence of both. What steps has the car industry taken to deal with this? As the automotive battery people know, both Honda and Toyota took decisions to develop a battery management system which protects the battery and puts battery warranty ahead of performance. When the battery is “over-stressed”, so to speak, the hybrid vehicle is increasingly dependent on its internal combustion engine.
Now is this acceptable to renewable energy system users? You can have longer life, but reduced service; or shorter life and a more available battery.
With an abundance of data from RES systems from all over the world and a couple of visualisation tools, Svorb has been able to show six different categories of RES system operation and can interpret the effects on different batteries. But does this give us an accurate indicator of battery life? Most likely not.
The sad fact is that lifetime prediction, for what it’s worth, can only be achieved now by cycling batteries in the manner in which you think the battery will be used, or in a test regime which approximates that. But as we all know, that’s too slow, and too expensive.
Dirk Uwe Sauer from the University of Aachen outlined the work of the EU funded ACTUS project, whose aim is to help battery users get the least lifetime cost out of the battery while getting the full state of function – a sort of no-compromise situation, compared to what our friends in the auto industry presently achieve.
Now since too few manufacturers – and too few customers – want to wait years for the results of years of cycling, accelerated lifetime testing has been the norm, certainly for lead acid, for years. However raised temperatures don’t totally reflect the ageing mechanisms. So he’s not just concentrating on that. I asked Dr Sauer to explain more.
“My opinion is that temperature acceleration is an appropriate method if all typical ageing effects are accelerated in same way by the temperature, and if the ageing of the battery is assumed to be caused mainly by the amp hour throughput. When looking at batteries in autonomous power supply systems, or even in hybrid cars, these conditions are not fulfilled.”
All the effects of temperature on sulphation are very difficult to describe. Furthermore, sulphation is accelerated with increasing temperature if the battery is at rest, or in deep states of charge. But during charging higher temperatures are even helping to dissolve the sulphate and to charge the battery properly. “And charging is the main aspect that needs to be focused in these applications,” says Sauer.
Sauer’s other main point is that a lot of damage happens in the discharged state. Very small currents are those which really hurt the battery. “Batteries in autonomous power supply systems are in a discharge mode for more than 50% of the day, with very small currents where internal variations, such as in acid concentration, cause relaxation currents.”
With existing acid stratification the battery is mainly charged in the upper part and discharged in the lower part. As a result the lower parts die from hard sulphation, which can be seen from all the post-mortem analyses on batteries from such applications.
You can’t show this with simple accelerated testing. However, part of the ACTUS electrochemical tests are at accelerated temperature – precisely to determine the ageing parameters on corrosion and drying out. But by using some of the basic electrochemical tests on a battery to calibrate a model, the Actus programme should be able to predict the lifetime of a battery in just three months: a fraction of the current time.
But is this really fast enough? And does it give us enough information?
Heinz Wenzel’s train problem is not unlike the automobile industry’s “change this battery now” problem that Fiamm has been working on (see BEST Autumn 2003 page 87).
And how interested is the bigger world in having such a solution if it is indeed attainable?
According to Per Lunsager of the Riso laboratory in Denmark, which has a major interest in wind power, batteries are just not going to figure in renewable schemes unless user can have an accurate indication of lifetime.
But, as was apparent at this very useful little seminar, there are really only a handful of scientists working on the problem. The lead acid battery industry has neither the money nor the manpower to solve the problem on its own.
The benchmarking programme has largely been EU funded so it was no suprise to have one of the Commission administrators, one Alex Sorokin doing his very best to encourage those present to put together a proposal covering battery lifetime prediction to meet the criteria of the Sixth framework sustainable energy programme.
Now what he should have said was, “look when it comes to funding science and engineering R&D, no one in Brussels has got a clue, but it seems like they’ve latched on to this sustainable thing. Now if you guys can write a sexy enough proposal explaining why battery lifetime is so important in this renewables stuff, well maybe just maybe you’ll a get a hearing to write some more concrete proposals.”
Of course none of these Brussels types could be so bold or honest. But some of the lead acid people just caught the flavour of what he was driving at. And in their hearts, they know they have to seize this opportunity, because their base markets are flat or being threatened. So maybe, some of them really will log onto www.cordis/lu/eoi/sustdev-energy. By the time this article appears, the closing date will have past. But what the hell?
Wenzell had managed to pull off one of those rare events in the battery calendar – a real exchange of ideas, proper food for thought, questions from the floor, vigorous argument. Let’s hope it sets the standard for all battery events this year.