Every week, someone sends us a purchase request that reads like this: "25 units, 22 kW home standby, best price, ship in 30 days, send the Caterpillar generator data."
That's a reasonable request. If you're buying home generators for wholesale, price per kW and lead time feel like the two numbers that decide the deal. I thought so too, for about the first six months on this job.
I'm the quality and brand compliance manager at a power systems distributor. I review every genset package before it reaches a customer — roughly 300 submittals a year. In 2025, 21% of the first-article packages we received got rejected. Not because the equipment was bad. Because the data didn't match the machine.
The problem you think you have
Most advice on how to choose home generator for wholesale starts with load calculations and price per kW. That's not wrong. You do need to know whether the load is 14 kW or 22 kW, and you do need to know what the unit costs landed.
So buyers build a comparison sheet. Rated kW, price, lead time, warranty length. Three columns, maybe five. And then they pick.
Here's the thing: almost none of the failures I've seen in four years came from picking the wrong column. They came from the columns meaning different things in different rows.
The problem you actually have: "22 kW" is three different numbers
Ask five suppliers for the rated output of a "22 kW" standby unit and you can get five defensible answers, because the rating depends on which standard you're reading from.
ISO 8528-1 defines several distinct power ratings for the same machine — emergency standby power (ESP), prime power (PRP), and continuous power (COP), among others. A unit labeled 22 kW under a standby rating might be 20 kW under prime and lower still under continuous. Same engine. Same alternator. Three nameplates, depending on who's printing it.
That matters more than it sounds. When a homeowner's load calculation was done against a standby number and the AHJ reads a prime number off the submittal, you've got a permit problem that no one can talk their way out of.
Then there's derate. A genset rated at 22 kW at sea level and 25°C does not deliver 22 kW at 1,800 meters and 40°C. Published derate guidance for diesel units generally runs around 3% per 1,000 ft of altitude and roughly 1% per 10°F above 77°F for naturally aspirated machines, with turbocharged units derating less. I'm generalizing — the exact curve is on the manufacturer's data sheet, and that's the curve you should be quoting. If the data sheet you have doesn't include one, you don't have the data.
This is the buried part. When someone asks us for caterpillar generator data, they usually mean the marketing spec sheet. The spec sheet is often the least useful document in the package, because it describes the platform, not the configuration you're buying.
Why it keeps happening: platform data isn't your data
Most home generator OEM programs work the same way. You take a common platform, put your brand on the enclosure, choose a transfer switch, choose a control panel, and ship it. That's a legitimate business. I'm not knocking it.
The failure mode is that the data travels with the platform instead of the configuration. Add an acoustic enclosure and the weight goes up, the airflow requirement changes, and the sound rating moves. Bundle a different ATS and the wiring schematic on file is now wrong. Specify a different controller and the annunciation list is wrong too.
So a single Caterpillar generator spec sheet can end up covering six SKUs, and nobody updates the derate table when the customer's site moves from a coastal install to a mountain one.
The most frustrating part of vendor management here: the same defect shows up from three different suppliers in the same quarter. You'd think written specs would prevent that, but a written spec with no rating definition just moves the argument downstream — to the dock, or to the inspector, or to the homeowner's driveway at 11 p.m.
What it costs when the data is wrong
Three places it shows up, in order of how much it hurts.
Freight and rework. In Q1 2025, we held a 60-unit private-label order at the port because the enclosure on the drawing was four inches wider than the crates we'd already built. $14,000 in scrapped crates, three weeks of schedule, and a very unhappy logistics team.
Inspection and permitting. A rejected submittal at the AHJ stage can cost you the install window. Crews get rescheduled, the homeowner waits, and the distributor eats the difference. This is the cost that never shows up in the quote.
Warranty and reputation. I still kick myself for not requiring a signed derate calculation on our first 60-unit private-label run back in 2023. We sized for 22 kW at 25°C. The site ran at 38°C in July. Three units went into overload on a Tuesday afternoon in August, and we paid for the service calls out of margin.
And then there's the one that's harder to price: time. As a standby generator distributor, you live or die by storm season. In August 2025, we paid a roughly 12% premium — about $9,600 — to lock a confirmed six-week delivery window on a 40-unit order instead of accepting an eight-week "should be fine" from a lower quote. The customer's install window was fixed. Missing it meant rescheduling crews on a $250,000 project. I'd make that trade again without blinking. What that premium bought wasn't speed. It was certainty.
To be fair, I get why buyers chase the cheapest confirmed-looking date. Budgets are real, and a 12% premium is a 12% premium. But "probably on time" isn't a date. It's a hope with a PO number attached.
What actually fixes it
This part is short, because the problem is already the hard part.
Every contract we write now includes five things. First, the rating definition in writing, with the standard it's measured against — ESP, PRP, or continuous, named explicitly. Second, a site-specific derate calculation at the actual altitude and ambient temperature, signed by the supplier's engineering contact. Third, the full submittal package before the PO: dimensioned drawing, wet and dry weight, sound level at a stated distance, fuel consumption at 25/50/75/100% load, wiring schematic, emissions certification, and the derate curve. Fourth, a locked delivery date, not an estimate. Fifth, a parts and service answer — who stocks the AVR, the control board, and the fuel filter for this specific model, and who's authorized to service it within 48 hours.
If a supplier can't produce all five, that's information too. It might be a fine supplier with a thin documentation process. But it's not a supplier you want on a deadline.
One caveat. This approach works for us because we're a mid-volume distributor with fairly predictable ordering — 80 to 120 units a month and a service arm we control. If you're a one-off private-label buyer sourcing through a trading company, you probably don't have the leverage to demand a signed derate calculation. In that case, get as much as you can in writing before money moves. At minimum: the rating definition and the derate curve. Those two documents prevent most of what I described above.
Everything else — the enclosure finish, the badge on the panel — is packaging. The data is the product.