Short answer: a home standby generator sold as a 26 kW unit delivers 22.5 kW on natural gas, and one sold as 22 kW delivers 19.5 kW. The model name is the propane rating. Since roughly two thirds of American standby installations run on a natural gas line, most buyers are quoting a number their generator will never produce. We pulled the published ratings for eight current standby units and calculated the gap for each. It ranges from zero to 14.3 percent, and it does not scale the way you would expect.
The Natural Gas Gap
| Model | Name implies | Propane (kW) | Natural gas (kW) | Gap |
|---|---|---|---|---|
| Generac Guardian 14kW (7224) | 14 | 14.0 | 14.0 | 0.0% |
| Generac Guardian 18kW (7228) | 18 | 18.0 | 17.0 | 5.6% |
| Kohler 26RCA | 26 | 26.0 | 24.0 | 7.7% |
| Kohler 20RCA | 20 | 20.0 | 18.0 | 10.0% |
| Cummins QuietConnect RS20A | 20 | 20.0 | 18.0 | 10.0% |
| Generac Guardian 22kW (7042) | 22 | 22.0 | 19.5 | 11.4% |
| Generac Guardian 26kW (7291) | 26 | 26.0 | 22.5 | 13.5% |
| Kohler 14RCA | 14 | 14.0 | 12.0 | 14.3% |
Two things stand out. The Generac 14 kW is the only unit in the table with no gap at all, delivering its full 14 kW on both fuels, while the Kohler of the same nominal size gives up 14.3 percent and delivers 12 kW. Two generators both marketed as 14 kW differ by 2 kW on the fuel most people will actually use. That is an entire refrigerator, freezer and furnace blower of difference, decided by a number that appears nowhere in the product name.
Second, the gap grows with size within the Generac line: zero at 14 kW, 5.6 percent at 18, 11.4 percent at 22 and 13.5 percent at 26. The larger units are pushing the same 999cc engine harder, and natural gas cannot keep up at the top of the range. Kohler runs the opposite way, with its largest unit showing the smallest gap.
What This Does to Sizing
The practical error is straightforward. A homeowner with a 20 kW calculated load looks at a 22 kW Generac, assumes 2 kW of margin, and installs a unit that delivers 19.5 kW on the gas line feeding it. The generator is undersized from day one, and the symptom appears only under full winter load, when the furnace, well pump and range are running together.
Three rules follow:
- Size from the natural gas column if you are on a gas line. The propane figure is irrelevant to you and exists mostly for the model name.
- Leave 20 percent headroom on top of that. Standby generators are rated for continuous duty, but running at 100 percent for days shortens engine life and leaves nothing for a motor start.
- Confirm the gas meter and line can feed it. This is the failure mode installers see most. A 22 kW generator needs a substantial gas supply, and an undersized meter or a long run of narrow pipe starves it regardless of what the spec sheet says. The utility can upgrade a meter, usually free, but it takes weeks.
Our guide to what size whole-house generator you need works from actual loads. Apply the derate before you compare.
Noise: The Weekly Exercise Trick
Every standby generator runs itself weekly to keep the engine healthy, and that self-test is the noise your neighbors hear most often, because it happens on a schedule whether or not there is an outage.
| Model | Normal operation | Weekly exercise | Reduction |
|---|---|---|---|
| Generac Guardian 14kW | 65 dBA | 55 dBA | 10 dBA |
| Generac Guardian 18kW | 65 dBA | 55 dBA | 10 dBA |
| Generac Guardian 22kW | 67 dBA | 57 dBA | 10 dBA |
| Generac Guardian 26kW | 67 dBA | 57 dBA | 10 dBA |
| Kohler 26RCA | 67 dBA | 56 dBA | 11 dBA |
| Kohler 14RCA | 69 dBA | 63 dBA | 6 dBA |
| Kohler 20RCA | 69 dBA | 64 dBA | 5 dBA |
| Cummins QuietConnect RS20A | 65 dBA | not published | n/a |
Generac’s quiet test drops the engine to a lower speed and takes a consistent 10 dBA off across the range, which is roughly a halving of perceived loudness. Kohler’s larger unit does even better at 11 dBA, but its 14 and 20 kW models only manage 5 to 6 dBA, so their weekly test is noticeably louder than a Generac’s.
One caution when comparing these figures: Generac publishes its numbers at 23 feet under normal load, while Kohler publishes at 7 meters with no load. Seven meters is 23 feet, so the distance matches, but a no-load measurement flatters any generator. Kohler’s 69 dBA at no load would be higher under real load. Treat the brands as roughly comparable and the within-brand comparisons as reliable. General decibel context is in generator noise levels explained.
Home Batteries: Power Versus Energy
Batteries have the same naming problem in a different form. A battery’s headline number is energy in kilowatt-hours, but whether it can run your house depends on power in kilowatts. Dividing one by the other shows what each system was designed for.
| System | Usable kWh | Continuous kW | kW per kWh | Minutes at full power | Chemistry |
|---|---|---|---|---|---|
| Tesla Powerwall 3 | 13.5 | 11.5 | 0.85 | 70 | LFP |
| Enphase IQ Battery 5P | 5.0 | 3.84 | 0.77 | 78 | LFP |
| Enphase IQ Battery 10C | 10.0 | 7.08 | 0.71 | 85 | LFP |
| FranklinWH aPower 2 | 15.0 | 10.0 | 0.67 | 90 | LFP |
| SolarEdge Home Battery | 9.7 | 5.0 | 0.52 | 116 | NMC or NCA |
The Powerwall 3 leads on power density at 0.85 kW per kWh, which is why a single unit can start large motor loads that would trip a comparably sized competitor. The SolarEdge sits at the other end at 0.52, meaning it discharges more slowly and lasts longer at its own maximum, which suits a house with modest continuous loads and a long overnight rather than one with a well pump.
The last column is the one worth internalizing. Every one of these systems empties in 70 to 116 minutes if you actually draw its full rated power. Batteries are sized for a night of essentials, not for running a house at capacity. That is the arithmetic behind how many kWh of home battery you actually need.
One chemistry outlier worth knowing
Every battery in this group uses lithium iron phosphate except the SolarEdge Home Battery, whose own emergency response documentation describes nickel and cobalt oxide cells, consistent with NMC or NCA chemistry. That is the same older chemistry family that produced the sixteen-fold lifetime difference we found among portable power stations. For a wall-mounted unit that will sit in a garage for fifteen years, chemistry deserves a place in the decision, and it is worth asking any installer to confirm it in writing.
Warranties are not comparable
Three different structures appear across these products. Enphase warrants 15 years or 6,000 cycles to 60 percent capacity. FranklinWH warrants 15 years or 60 MWh of total throughput. Generac warrants its PWRcell modules for 10 years or 7.56 MWh per module. A cycle warranty, a throughput warranty and a calendar warranty cannot be compared directly, and the throughput figures are the most honest of the three because they measure what you actually took out of the battery.
What We Could Not Verify
Every figure above came from a manufacturer specification sheet or datasheet read in September 2026, with two exceptions worth stating plainly. Tesla’s site blocked automated access, so the Powerwall 3 capacity and power figures come from a secondary industry source, and several of its other fields could not be confirmed and were left out rather than estimated. Cummins does not publish a weekly exercise decibel figure.
We also found no current product called the Anker SOLIX X1 on Anker’s own site, so it is absent from the battery table rather than filled in with a similar model’s specs.
Installed prices are not published by most of these manufacturers and vary widely by region and installer, so they are omitted. Generator unit prices from dealers are included in no table here because they move constantly and because the bundling of a transfer switch, which can be a 1,500 dollar difference, makes headline prices misleading. Whether the switch is included is noted in manual vs automatic transfer switches.
Frequently Asked Questions
Why is my standby generator rated lower on natural gas?
Natural gas arrives at low pressure and carries less energy per unit volume than propane, so the engine produces less power from the same displacement. The drop ranges from zero to about 14 percent depending on the model, and manufacturers name the product after the propane figure.
Which is better for a standby generator, natural gas or propane?
Natural gas, for almost anyone who has a gas line. It never runs out, costs the least per kWh, and requires no tank. Propane’s advantages are that it works where no gas line exists and that it stores indefinitely. The fuel comparison is in natural gas vs propane generators.
Does the model name tell me the real output?
Only if you run propane. On natural gas, look up the specific spec sheet figure for your model. Two generators with the same name from different brands can differ by 2 kW on gas.
How loud is a standby generator during its weekly test?
Between 55 and 64 dBA at 23 feet depending on brand and model, against 65 to 69 dBA under load. Generac’s quiet test reduces noise by a consistent 10 dBA; Kohler’s smaller units reduce it by only 5 to 6.
Can one home battery replace a standby generator?
For essentials overnight, yes. For whole-house loads over several days, no. Every system here empties in under two hours at its full rated power, so a battery covers a night and a generator covers a week. Many households with frequent long outages end up with both.