The Home Battery ReportIndependent · No installer money
spec datadatasheetsround-trip efficiencylosses

Battery round trip efficiency compared: 17 models and their measurement basis

Published round trip efficiency for every home battery we track, from 85.4 to 98 percent. Sort the list by the number and you sort it by measurement basis, because the top of the table is not measuring an AC round trip at all.

The Home Battery Report8 min read✓ Verified

Round trip efficiency should be the easiest spec in home storage to compare. It is one percentage. Every maker publishes it except one. Put the 17 models on our comparison pages in a column, sort descending, and you have a ranking.

What you actually have is a ranking of measurement bases. The three highest numbers in the table, 98 percent, 96 percent and 95 percent, are not measuring an AC round trip. The lowest number in the table, 85.4 percent, is a maker publishing a figure calculated to a state regulator's formula. And the single most-quoted figure in the category, Tesla's 89 percent for the Powerwall 3, is measuring a completely different journey than every other number on the page.

If you want the concept rather than the model-level data, round-trip efficiency explained is the place to start. This page is the table, with each maker's basis attached to each number. Every figure is a published manufacturer specification, and we have not tested any of these products.

The figures as published, highest to lowest

The basis column is the whole story

Group those 17 by what was actually measured and the table stops being a ranking.

  • Full AC round trip, conditions stated. Anker at up to 90.2 percent at 77 degrees F, LG at over 90 percent with charge and discharge powers and an ambient given, Tesla Powerwall 2 at 90 percent at 25 degrees C and 3.3 kW, Panasonic at 89 percent AC coupled at 50 percent of power rating, FranklinWH at 90 percent grid to battery to load, and Enphase at 90 percent AC on both the IQ Battery 5P and the 10C. Seven models, and they span 89 to 90.2 percent. That band is the real answer for what a modern home battery returns.
  • Weighted or formula-based. SunPower's 86 percent CEC weighted and sonnen's 85.4 percent per the SGIP formula. Lower numbers, and not because the hardware is worse. A weighted methodology averages across operating points instead of quoting the best one.
  • Peak, or stated without conditions. SolarEdge's 94.5 percent peak, Savant's 93.8 percent, Generac's 88 percent system figure. Each of these could sit anywhere relative to the AC-to-AC group depending on what was included.
  • Measured at the battery, not across the system. Fortress at 98 percent at 0.5C, BYD at 96 percent with no conditions, Qcells at 95 percent at the module. These are DC batteries. The inverter losses that the AC group is carrying are simply not in these numbers.
  • A different journey entirely. Tesla Powerwall 3 at 89 percent.
  • Nothing published. EG4.

Three makers in this set publish both a system figure and a component figure for the same product, and the gap between them is remarkably consistent. Enphase publishes 90 percent AC and 96 percent DC, a 6-point gap. Panasonic publishes 89 percent AC coupled and 94 percent DC coupled, a 5-point gap. Qcells publishes 95 percent at the module and 88.5 to 92.5 percent as system cycle efficiency, a gap of 2.5 to 6.5 points.

That pattern is the fairest way to read the top of the table. A 96 or 98 percent battery-level figure is not a lie and it is not a triumph; it is the same kind of number the AC-coupled makers publish alongside their headline, and it sits roughly 5 to 6 points above where a full AC round trip for the same product would land. We are not going to translate anybody's figure, because that would be inventing a spec. We are pointing out that the ordering of the table changes if you do.

Why Tesla's 89 percent is a different measurement

Tesla's Powerwall 3 datasheet publishes two efficiency lines and neither one is an AC round trip.

The first is Solar to Battery to Home/Grid Efficiency at 89 percent, footnoted as a typical solar shifting use case. That path takes DC power off the roof, stores it, and delivers AC to the house. It never performs the AC-to-DC conversion at the front end that an AC-coupled battery performs every time it charges. The second is Solar to Home/Grid Efficiency at 97.5 percent, footnoted as tested using CEC weighted efficiency methodology, which is the pass-through path with no storage in it at all.

Neither answers the question an AC round trip answers, which is: if I put a kWh in from the grid, how much comes back out. For a buyer charging from solar and discharging in the evening, Tesla's 89 percent is the relevant number and it is measured on the relevant path. For a buyer charging from the grid at off-peak rates and discharging on peak, which is the whole premise of load shifting on a time-of-use plan, Tesla publishes no applicable figure.

That is worth stating plainly because the 89 percent gets quoted constantly against rivals' 90 percent AC figures as though the Powerwall 3 is a point behind. It is not behind or ahead. It is not the same measurement. And it is a step backwards in disclosure from the Powerwall 2, whose datasheet published a fully conditioned AC-to-battery-to-AC figure.

What the difference is worth in energy

Take a household cycling 10 kWh through the battery every day, which is a reasonable figure for a mid-sized system on a daily arbitrage pattern.

At 90 percent round trip, the loss is 1 kWh a day, about 365 kWh a year. At 88 percent it is 1.2 kWh a day, about 438 kWh a year. At 85.4 percent it is roughly 1.46 kWh a day, about 533 kWh a year. So across the full published AC-and-weighted range in our table, from 90.2 percent down to 85.4 percent, the difference is around 170 kWh a year on that cycling assumption.

That arithmetic is ours, applied to the makers' figures, and it is deliberately in kWh rather than dollars because what a kWh is worth depends entirely on your rate and your rate plan. Put your own numbers into the Worth It calculator to see what it does to the payback.

The honest framing is that round trip efficiency is a real cost and a small differentiator. Across the seven models publishing a fully conditioned AC round trip, the spread is 1.2 percentage points. The specs that actually separate these products are continuous power, usable capacity, and the warranty term and its throughput cap, which are compared in our warranty guide.

Every figure above is a manufacturer datasheet claim recorded with its published basis. Where a maker publishes no figure we say so rather than borrowing one, and none of these numbers came from testing we did.