KIT-test: Some battery cells last five times longer than others

In comparing battery storage, you will probably look at what kind of lifespan the manufacturer promises and what the system costs. However, there usually is no guarantee for the given lifespan. The results presented by Andreas Gutsch of the Karlsruhe Institute of Technology (KIT) at the PV Symposium in Bad Staffelstein, Germany, last week, are therefore sure to be quite explosive. According to his findings, some lithium-ion cells integrated in storage systems lost up to 30% of their capacity after just 1,000 cycles. That is a possible criterion signifying the expiration of cell’s lifespan. Other cells are well above that even after 5,000 cycles. The range, it turns out, is huge.

However, cells that hold fewer cycles are not necessarily of lesser quality, the KIT storage expert explains. “That depends on the price.” While this is known, the lifespan is not. "The market is completely opaque in terms of performance," Gutsch said.

He and his colleagues therefore put a large number of lithium-ion battery cells under a particularly stressful test, fully charging and discharging the batteries to 0%, as is common for tests in the automotive industry. In battery storage systems, the cells are usually discharged only up to a state of charge of 20%, so that the cycle life in absolute terms is likely to be longer than in the KIT-tests (typically referred to as 80% Depth of Discharge, or 80% DOD). While the KIT tests do not offer an absolute statement about the lifespan of the cells in a battery storage system, they do offer a good comparison between different cells as the test conditions for all of the cells were the same.

Differences by country of origin

Gutsch decline to present the results of individual cells. "It’s not so easy to obtain the cells," he said. It was therefore necessary to agree with manufacturers’ wishes not to publish the results with specific names. He nevertheless revealed a regional distribution of the results that may surprise some.

In a graph, Gutsch offered a comparison of the maximum capacity of cells based on the number of cycles they had behind them. All four or five cells from China that were in the test ranked in the bottom third of the coordinate system, indicating that their performance declined particularly rapidly. All cells from Germany and Japan ranked in the upper third, while those from the United States and South Korea were either in the upper or middle third.

Gutsch did offer detailed information on one particularly successful model, the Tesla battery. It only lasted 400 cycles. “This battery is built specifically for use in a car," he said. If an electric vehicle owner is able to drive 500 kilometers with a fully charged battery, he can drives 200,000 kilometers with a battery that lasts 400 cycles. A Tesla battery can therefore not be used for stationary applications.

For stationary energy storage, Gutsch said batteries with lifespans of at least 3,000 cycles were needed in order for it to operate profitably. Even if the battery storage cost €500 per kilowatt hour capacity, it would still be uneconomical. And he doesn’t expect system costs in the near term to fall below that. Even with the development of costs — German PV expert Winfried Hoffmann, for example, predicts cell costs could fall to as low as €100 per kilowatt hour – it would be difficult for system costs to fall below a few hundred euros.

For a first estimate of current cost of storage, i.e., the cost per kilowatt hour, which is fed into a battery storage system to be used later, divide the storage system cost per kilowatt hour battery capacity by the number of cycles. This results in the minimum possible value. Shown here is how the storage costs for three differently priced systems depend on the cycle lifespan. The cost of the stored power from the system that costs €1,000 per kilowatt hour is always twice as high as that of the system for €500 per kilowatt hour. However, the stored power costs clearly increase for low cycle lifetimes unlinearly, making them still too high for cheap battery storage. Gutsch set limits that even cheap systems will need to reach in the future at 2,500 to 3,000 cycles. Graph: pv magazine / Solarpraxis

Security problems remain an issue

KIT already caused a stir last summer, when it likewise published anonymous tests on battery storage and a checklist to estimate the safety of a system. Problems continue, however. Gutsch exhibited photos of systems that do not comply with the simplest of standards. With one system in particular it was possible to cause a fatal short circuit of the battery by connecting the two cables that are thought to connect the battery to the battery management system. Another system had no redundancy, which would prevent the overcharging of the battery in case a circuit component failed. Yet another system had 16 temperature sensors built into the battery modules, which is very commendable, but it nevertheless had only one that was connected to the battery management system.

Some manufacturers advertise that their systems contain lithium iron phosphate cells and are therefore safe – a claim that visibly upsets Gutsch. This cell type could experience a so-called “thermal runaway,” which could even go off earlier than in some other cell models. "The only thing that’s better is the lower enthalpy of reaction.” This is the energy that is released during the reaction and, in this case, means that "the material starts to go off earlier and you have more time to run away.” In addition, each cell contains a liquid, flammable electrolyte.

According to Gutsch, there’s no way around it but to design systems safely and in compliance with standards, especially the IEC 61508, which describes functional safety. KIT and several other German organizations have jointly published a guideline on how to build storage systems in response to the checklist that shows how best to do this.

Ulimately, Gutsch believes that battery storage will still have great success in the coming years.