Battery Storage Inspection 2026: Large differences in efficiency and warranty
Twelve solar power storage systems in lab test: New frontrunner achieves an overall efficiency of 97 percent. There are significant gaps in warranty conditions.
(Image: Crovik Media / Shutterstock.com)
HTW Berlin and the spin-off company aquu tested and evaluated twelve solar power storage systems in the 5 kW and 10 kW performance classes. In addition to a new frontrunner in the 10 kW class, the study provides the first systematic analysis of the warranty conditions of 20 manufacturers – with sometimes significant differences. The test field included products from the usual suspects: BYD, Fox ESS, Fronius, Kostal, Sax Power, and SMA.
The heart of the battery storage inspection is the determination of the System Performance Index (SPI). Simply put, it is an efficiency metric that considers four categories of losses: sizing losses, conversion losses, control losses, and standby losses.
10 kW class with new efficiency winner
According to the test results, the Fox ESS PQ-H3-Ultra-10.0 hybrid inverter in combination with the EQ3300-5 battery took the lead in the 10 kW class with an SPI of 97.0 percent. This is a record value, as the authors write. Behind it is the RCT Power Power Storage DC 10.0 with the Power Battery 11.5, which achieves an SPI of 96.4 percent. In third place is the Energy Depot Centurio 10 together with the Domus 2.5 battery with an SPI of 95.4 percent.
The system comprising the Fox ESS PQ-H3-Ultra-10.0 hybrid inverter and the EQ3300-5 battery achieves a 97 percent SPI in the 2026 Battery Storage Inspection.
(Image:Â HTW Berlin)
The last place in the 10 kW class is an anonymously listed system. According to the study, it only achieved an SPI of 89.3 percent and landed in efficiency class G. The SPI losses of this system are therefore 3.5 times higher than those of the top performer from Fox ESS. The authors attribute this primarily to a battery efficiency of only 87.9 percent and a standby consumption of 64 watts.
In the smaller performance class up to 5 kW, the Sax Power Home Plus AC-coupled battery system emerged as the leader according to the study, with an SPI of 93.2 percent. The system is based on the so-called multi-level technology, which combines power electronics and battery cells in a single device.
Behind it follows SMA with the Sunny Boy Smart Energy 5.0 and the Home Storage 6.5 battery with an SPI of 92.8 percent. Kostal is close behind with the Plenticore MP G3 M 4.6 in the DC-coupled version with the BYD Battery-Box HVS+ 7.7, achieving an SPI of 92.7 percent. The detailed loss analysis of the study reveals interesting differences: the SMA inverter had lower conversion losses than the Kostal system, but Kostal compensated for this with more precise control and more economical standby operation.
Financial implications of efficiency differences
According to the scientists' calculations, the overall system efficiency directly affects the annual cost savings. The basis is a reference household with a 10 kW PV system, heat pump, and electric car, which consumes 9363 kWh per year and pays 40 cents per kilowatt-hour for grid electricity.
In this scenario, according to the study, each SPI loss percentage point in the 10 kW class costs an average of 26 euros per year. With a difference of 7.7 percentage points between the top performer from Fox ESS and the last place, the annual cost advantage of the more efficient system amounts to around 200 euros.
Videos by heise
The study also shows: a highly efficient system can achieve the same electricity cost savings with less battery capacity than a less efficient system with a larger storage. A specific example from the authors: the RCT Power storage system with only 10.6 kWh of usable capacity and an SPI of 96.4 percent reduces the annual electricity bill almost as much as a system with 14.2 kWh capacity that only achieves an SPI of 92.3 percent. Choosing a more efficient system thus allows you to save on additional capacity – an advantage for resource efficiency as well, according to the authors.
Wide range in partial load and response times
The efficiencies in the so-called partial load range are particularly relevant in practice. After all, at night, when only the refrigerator is humming, household electricity consumption is usually below 300 watts.
At a discharge power of only 100 watts, the efficiencies of the 10 kW inverters range from 54 percent for the least efficient to 86 percent for the RCT Power system. Sax Power leads in the smaller 5 kW devices with 84 percent at 100 watts output power. Specifically, this means that in the worst case, half of the battery power going into the inverter is lost as waste heat.
Consumers should also pay close attention to the settling times. This is the time it takes for a system to react to load changes. The results vary between 0.2 seconds for Energy Depot and RCT Power and 13.7 seconds for the anonymous system.
To avoid unnecessarily drawing power from the grid, a system must react quickly to changes in load. This works better for some than for others.
(Image:Â HTW Berlin)
The ability of PV storage systems to react quickly to load fluctuations is relevant when you turn on the electric kettle, which immediately draws 1500 watts, or switch on cycling loads like a hob. Speed is crucial to avoid unnecessarily drawing power from the grid – or feeding it in.
Significant differences: Warranty conditions of 20 manufacturers analyzed
In addition to efficiency, the Berlin scientists systematically examined the warranty conditions of 20 well-known manufacturers. The results reveal significant differences. For example, the guaranteed remaining capacity at the end of the warranty period varies between 60 and 85 percent of the initial value, depending on the manufacturer.
The product warranty for the named participants ranges from 5 to 10 years, and the performance warranty from 10 to 12 years. SMA offers the longest performance warranty among the test participants, with 12 years, but guarantees only 60 percent remaining capacity – the lowest value in the comparison.
Furthermore, the study identifies numerous differences in warranty requirements and exclusions. Some manufacturers require registration within 30 days, while others allow up to 12 months. A permanent internet connection is a basic requirement for warranty claims for many manufacturers.
The scope of cost coverage also varies considerably in the event of a claim: while some manufacturers cover all ancillary costs, including installation and removal as well as travel, others only provide a replacement device and charge for labor costs – sometimes at 150 euros per hour.
“Looking at the warranty conditions before purchasing a storage system can prevent unpleasant surprises in the event of a claim,” says Nico Orth from HTW Berlin. Advantageous conditions can be recognized, among other things, by the fact that the manufacturer does not impose extensive proof obligations and covers the costs of replacement in the event of a claim.
(ssi)