Ask three manufacturers for a battery inverter datasheet and you get back three documents that refuse to line up. Not because the products are bad. Because they are not the same product, and two of the three vendors have quietly stopped selling the thing the question was about.
I spent an afternoon with the current PDFs from Fronius, SMA and Enphase, plus SMA's own product pages. What follows compares published records, not hardware I've wired up. Where a number is missing, I say it's missing rather than borrowing one from a retailer.
Did the Battery Inverter Category Just Dissolve?
A battery inverter, strictly, converts between a battery and AC and does nothing else. SMA sold exactly that: the Sunny Boy Storage 3.7, 5.0 and 6.0, with three battery connections and no PV input. Go to that product page today and SMA says plainly that it no longer sells the Sunny Boy Storage, and points you at its hybrid line instead, the Sunny Boy Smart Energy-US.
Fronius never sold a standalone battery inverter in this class. It sells the Primo GEN24, a PV inverter, and the battery function arrives through the UP.storage software upgrade that turns it into a GEN24 Plus. Enphase went the other way entirely: the IQ Battery 5P has six IQ8D-BAT microinverters bolted inside the battery enclosure, so there is no separate box to compare at all. One US-only battery pairing for that GEN24 Plus is covered in Fronius GEN24 Plus and Renon.
So what are you actually choosing between? Three architectures wearing one name.
What Do the Three Datasheets Actually Say?
| Specification | Fronius Primo GEN24 Plus 208-240 | SMA Sunny Boy Smart Energy-US | Enphase IQ Battery 5P |
|---|---|---|---|
| Battery function | Software upgrade on GEN24 Plus | Built in | Microinverters inside the battery |
| Rated AC power | 3.8 to 10.0 kW | 3.84 to 11.52 kVA | 3.84 kVA, 7.68 kVA for 3 s |
| Battery DC range | 150 to 455 V | 90 to 500 V | 76.8 V nominal, 86.4 V max |
| Max charge current | 22 A | 30 A | Not applicable |
| Efficiency published | 97.6% max, 96.5 to 97% CEC | 98.1% max, 96.5 to 97.5% CEC | 96% DC round-trip, 90% AC round-trip |
| Ambient range | -40 to +60 C | -25 to +60 C, with derating | -20 to +50 C charge, -20 to +55 C discharge |
| Derating curve printed | Yes, per class, 30 to 50 C | No | No, an optimum band instead |
| Cooling | Active, with fans | Natural convection | Natural convection |
| Enclosure | Type 4X | IP65 and Type 3R | NEMA 3R |
| Warranty on the datasheet | Absent | 10 years, +5, +10, +15 optional | 15 years or 6,000 cycles |
| UL 9540 listed | No | Yes, with BYD HVL | Yes, and UL 9540A |
Where Do the Datasheets Split on Heat?
Every one of these boxes loses power when it gets hot. Only one of them shows you the shape of that loss.
The Fronius Primo GEN24 208-240 datasheet devotes a page to temperature derating charts, one per power class, plotting output from 30 to 50 degrees C at two DC input voltages. You can read off roughly what a 6.0 will still deliver on a 45 degree wall. SMA gives you a single line: operating range minus 25 to plus 60 degrees C, with derating. Derating starting where, falling how fast, to what floor? The datasheet does not say, and SMA cools by natural convection while Fronius uses active fans, so the two curves cannot possibly have the same shape. How much does a passively cooled box give up on a west wall in August against one with fans in it? That is precisely the question the phrase "with derating" refuses to answer.
Enphase does something different again, and I think it's the most useful of the three for a homeowner. It gives a hard charging ceiling of 50 degrees C, a discharging ceiling of 55, and an optimum operating range of 0 to 30 degrees C. That last row is an admission most vendors will not make: the box works across a wide band and is happy in a narrow one.
A single operating temperature range tells you where the inverter survives. Only a derating curve tells you what it will actually deliver there, and of these three datasheets exactly one prints the curve.
Why Can't You Compare These Efficiency Figures?
SMA's 98.1 percent max beats Fronius at 97.6 percent, and both appear to crush Enphase at 90 percent. Put those in the same column and you'd conclude the Enphase is a disaster. It isn't. The numbers measure different journeys.
The SMA and Fronius figures are peak DC to AC conversion on the PV side, at one load, one voltage, one temperature. Enphase's 90 percent is an AC round-trip: mains into the battery and back out again, at 50 percent of rated power, with both conversions and the cells included. Its one-way equivalent is the 96 percent DC round-trip figure on the same page.
The CEC numbers are closer to comparable and much closer together: 96.5 to 97.5 percent for SMA depending on power class, 96.5 to 97 percent for Fronius. That's a spread of one point between two brands whose quotes can differ by far more than one point of anything. So which of the three figures belongs on a customer quote? I've argued before that the headline efficiency figure misleads, and stacking three architectures side by side makes the case better than two ever did.
One derating threshold is worth pulling out because it has nothing to do with heat. Fronius states that AC power derating occurs at a battery DC input voltage of 419.7 V and higher, inside a 150 to 455 V window. The top 8 percent of the battery voltage range costs you AC output. Neither SMA nor Enphase publishes anything equivalent.
Which Warranty Deadlines Does Nobody Print?
Two of these three datasheets carry a warranty row. SMA prints 10 years standard with optional 5, 10 and 15 year extensions. Enphase prints 15 years or 6,000 cycles, whichever comes first, holding above 60 percent capacity. Fronius prints nothing, and that gap matters more than it looks.
Go to the Fronius USA warranty page and the 10 years is conditional: the unit must be registered on Solar.web within 30 months of delivery. The paid extension to 15 years must be purchased within 6 months of the warranty period starting. Miss either window and the spec you thought you bought is not the spec you have. How many homeowners know a 6 month clock started the day their inverter was commissioned? A warranty with deadlines attached belongs on the datasheet, next to the efficiency row. The warranty term also sets the payback runway, which we turn into a price ceiling in our home battery break-even model.
What Does Backup Mean for Each Inverter?
This is where I'd push back hardest on how these products get sold. Fronius PV Point switches in under 23 seconds; Full Backup in under 35 seconds on the 3.8 to 6.0 units and under 45 on the 7.7 and 10.0, and it needs external grid switching hardware. SMA Backup Secure is a manually activated 1,920 VA outlet at 120 V, available with or without a battery. Enphase forms the grid from the battery itself.
SMA's range goes beyond that outlet. The external Backup Select box transfers automatically and feeds five 15 A / 120 V outlets plus two 20 A / 120 V circuits, up to 3,840 W with the 3.8 to 7.7 classes and up to 5,760 W with the 9.6 and 11.5, per SMA America's product page. It works with or without a battery. SMA's page describes the switch between grid and off-grid mode as automatic and gap-free, yet I couldn't find a transfer time for it in the datasheet or on the page. A description isn't a spec. There's a catch on the other side too. In installations bound by NEC rapid shutdown rules, the datasheet says Backup Secure needs the separate SMA Backup Start module before it will work at all. Neither SMA option is sold as whole-home backup.
Forty-five seconds is not a flicker. Your desk machine reboots, your router renegotiates, your oven clock starts blinking. Would you call that backup, or would you call it a fast restart? If anything in the house cares about continuity, the architecture decides that, not the efficiency table.
Which Battery Inverter Would I Specify?
My opinion, and installers will disagree: on a retrofit where the PV is already up and working, the Enphase IQ Battery 5P is the honest choice, because AC coupling means nothing about the existing array has to be right for the battery to work. On a new build with a complex roof, the Fronius documentation is worth paying for, and the derating charts are a real part of what you're buying. SMA sits between them with the best CEC numbers and the weakest thermal disclosure of the three.
Whichever you pick, size the battery before the inverter. The home solar battery sizing guide does that arithmetic, and if you're weighing the Enphase against the obvious alternative, the Powerwall 3 comparison runs it directly.
Summary
Fronius, SMA and Enphase no longer sell the same kind of product, so a battery inverter comparison is really an architecture comparison. Fronius is the only one of the three that prints temperature derating curves, and the only one that omits warranty terms from the datasheet entirely while attaching a 30 month registration window and a 6 month extension deadline to them elsewhere. SMA posts the best CEC efficiency at 96.5 to 97.5 percent and documents thermal behaviour with a single phrase. Enphase publishes an optimum operating band of 0 to 30 degrees C and a 15 year or 6,000 cycle warranty on the datasheet itself, and its 90 percent AC round-trip is not comparable to the other two figures. Backup separates them most sharply: Enphase forms the grid, SMA offers a manual 1,920 VA outlet or the automatic Backup Select box with no published transfer time, and Fronius takes up to 45 seconds to get there.