Round-trip efficiency on commercial battery datasheets

Round-trip efficiency is the share of the energy you put into a battery that you get back out. It sounds like one number. On commercial battery datasheets it is stated on different bases, and sometimes it is not stated at all. We read the datasheets for six manufacturers’ commercial battery products and found four efficiency figures (more than 89%, 91.7% to 93.7%, 95% and at least 95%), described in four different ways, and two datasheets that give none.

How we did it

On 14 October 2026 we read the manufacturer datasheet for one commercial battery product from each of six makers: Tesla, Fox ESS, Pylontech, BYD, SolarEdge and Sungrow. We took the figures as printed and noted what each sheet says about how the figure was reached. We did not test any equipment, and the list is not a ranking: it is the six sheets we could read in full.

What the six sheets say

Why the figures cannot be lined up

Four things differ from sheet to sheet.

Where the figure is measured. A figure for the battery alone, in direct current, leaves out the inverter. A figure for the whole system, at the alternating-current terminals, includes the inverter, the transformer if there is one, the cooling and the system’s own load. Tesla’s sheet says its figure includes all power conversion and thermal system losses. BYD’s product has no inverter, so its figure cannot. For Fox ESS and Pylontech the sheet does not say.

The conditions. Tesla states 25 °C and a full-depth cycle. Pylontech gives a power level, 0.5C, which means a charge or discharge over two hours. BYD’s footnote gives 0.2C, a five-hour rate, for its usable-energy figure. Efficiency usually falls at higher power and at temperature extremes, and a sheet that gives no conditions leaves you unable to say which case it describes.

What is left out. Two of the six give no round-trip figure, only the efficiency of the inverter or converter. That is one link in the chain and not the chain.

Whether it is measured or calculated. None of the six sheets names a test standard or says who measured the figure.

A worked example shows how much the boundary matters. Take a battery with its own round-trip efficiency of 95%, and an inverter that is 97.5% efficient, the maximum SolarEdge lists for its 50 kW battery inverter. Charging and discharging both pass through the inverter, so the efficiency at the alternating-current terminals is about 0.95 × 0.975 × 0.975, or 90%, before cooling and the system’s own load. That is our arithmetic, to show the effect, and not a measurement of any product. It means a battery-only 95% and a system-level 91.7% cannot be compared as they stand.

Reference points

Two published sources give a sense of the usual range, and both are planning assumptions and not measurements of products. The US National Renewable Energy Laboratory’s Annual Technology Baseline 2025 adopts 85% as a representative round-trip efficiency for commercial battery storage, defined as the ratio of useful energy output to useful energy input. A 2019 worked example from the Electric Power Research Institute, at utility scale, gave about 0.87 for an AC-coupled chain and about 0.91 for a DC-coupled one, from assumed component efficiencies. Our guide to AC-coupled battery storage shows why the arrangement of the inverters changes the chain.

Does it matter?

It matters, but less than the headline suggests. The same EPRI document says round-trip efficiency usually has a lower impact on lifetime project economics than other cost and benefit streams. To get a sense of scale, a 200 kWh battery that is fully discharged on 250 days a year delivers 50 MWh, and each four points of round-trip efficiency on that is about 2 MWh of energy a year (50 MWh × 0.04). Multiply by the gap between the price you pay when you charge and the price you avoid when you discharge to see what it is worth on your tariff. We state no price, because the answer depends on your own data.

What to ask for in a quote

• The round-trip efficiency at the point of connection, in alternating current, including the inverter, any transformer, the cooling and the system’s own load.
• The power level and temperature it applies to, and whether it is for a new battery or an aged one.
• How it was measured, by whom and to which test method, or whether it is calculated.
• Whether the figure is for the battery alone or the whole system, and the same for every bidder you compare.
• Whether the contract commits to any figure, or only the datasheet does.

For the wider buying decision see our guide to commercial battery storage, and for what the warranties promise see what eight makers publish.

Limits of this comparison

It covers six datasheets on one day, one product from each maker, so it says nothing about the rest of their ranges. Datasheets change without notice, and some of the copies we read were hosted by third parties and not by the maker. We tested no equipment. Efficiency in operation depends on how a battery is used, including its power level, its temperature and how often it cycles, so a datasheet figure is a starting point and not a forecast.

Sources

• Tesla, Megapack 2 XL datasheet, Rev. 2.0, 13 November 2025 (copy hosted by Cayuga County, New York)
• Fox ESS, G-MAX datasheet, V1.6, 29 May 2026
• Pylontech, OPTIM US A100-HY datasheet (distributor-hosted copy)
• BYD, Battery-Box Premium LVL datasheet
• SolarEdge, CSS-OU-20 battery cabinet datasheet, DS-000233-ENG (distributor-hosted copy)
• Sungrow, PowerStack ST255CS-2H datasheet
• National Renewable Energy Laboratory, Annual Technology Baseline 2025, commercial battery storage
• Electric Power Research Institute, Solar Plus Storage Cost Assessment and Design Considerations: Executive Summary, 3002016637, December 2019, slide 9

Related reading

• AC-coupled vs DC-coupled battery storage for business
• Commercial battery storage: a buying guide
• What is a BESS?