Nobody Rents a Generator Because It’s Cheap

A contractor called me at 7:15 on a Friday morning last spring. His tower crane was supposed to go up Monday, and the site’s temporary power connection had just been pushed back three weeks. The rental company quoted him $900 a day for a 400kVA set. He was doing the math out loud: twenty-one days, eighteen thousand nine hundred dollars. “That’s insane,” he said. “Should I just buy one?”

I told him to slow down — not because I had the answer ready, but because he was asking the wrong question. Rent vs. buy isn’t about the daily rate. It’s about how many hours of power you actually need, and what each of those hours really costs you. Clients ask me this one constantly, so let me answer it the way I answer it on the phone.

So how do I actually do the math?

The back-of-the-envelope version is easy. Rental quote times days, compared against the purchase price. For my contractor: $18,900 to rent the 400kVA for three weeks, or roughly $40,000 to $50,000 to buy one outright — 粗算， prices move with the engine brand and the canopy, but that band is honest. If you want a fuller picture of what these sets actually cost to buy, I’ve written that up separately. For three weeks, renting wins by a mile. For a two-year job running daily, buying wins by a mile. Most real cases live in the awkward middle.

Here’s where it gets interesting. That comparison is rigged — it prices the machine, not the ownership. A rental quote is an all-in number: the set, the maintenance, the replacement if it breaks down at 2 a.m. A purchase price is just the cover charge.

What owning really costs (the part nobody puts in the spreadsheet)

First, your time or your mechanic’s. A set you own is a set you maintain: oil and filter changes every 250 hours, coolant, fuel filters, air filters. Ignore it and it will remind you, usually during the one week you can’t afford downtime. Across the projects I’ve handled, keeping a mid-size set alive — not the fuel, just the maintenance — runs something like 8 to 12 percent of the purchase price every year. 粗算， but it never shows up in the rent-vs-buy spreadsheet.

Second, idle sets rot. A client of mine bought a 500kVA set for a nine-month job, ran it hard, then parked it behind the warehouse when the job ended. Two years later a new project needed it. The battery was dead, the diesel had gone stale — diesel grows water and microbes if you let it sit, no way around that — and half the seals had dried out. Recommissioning it cost him close to what a year’s maintenance contract would have. Renters never meet this problem, because the set goes back on a truck.

The honest ownership checklist looks like this:

• Scheduled maintenance: oil, filters, coolant — every 250 hours or yearly, whichever comes first
• The battery: plan on replacing it every two or three years, always on the worst possible morning
• Fuel handling: tank drains, and polishing if the set sits for months
• Insurance, transport, and somewhere to park several tons of machine
• Depreciation — and the long, slow process of selling a used set

Add those up and the break-even point moves. The rental quote that looked outrageous for three weeks starts looking reasonable for three months — and for anything under about six months of use, renting is usually the right call, full stop. That’s not the rental company talking. That’s the math talking.

One more honest note: the math above assumes you actually use the set you’re paying for. Renting forces you to re-decide every month — projects change, loads change, and that monthly invoice keeps you honest. Ownership lets you stop thinking, which is exactly how a $50,000 machine ends up parked behind a warehouse for two years. Inertia is expensive too.

When renting wins even if you need power for months

Now the counterintuitive part. Sometimes renting wins even when the calendar says buying should. Three situations, all real.

Your timeline is uncertain. “Six months, maybe nine” is the most expensive sentence in this business. If you buy and the job ends at month four, you’re stuck holding a depreciating asset. Rental contracts extend month to month. You’re buying flexibility, and flexibility has a price — it just happens to be cheaper than a wrong guess.

You don’t have a maintenance crew. A set in a city basement with an automatic transfer switch needs someone to check it monthly — battery voltage, coolant, a test run under load. Rental companies include all of that. Their sets get load-bank tested between jobs; I’ve watched a rental fleet run a bank test on every returned unit, and frankly their fleets are better maintained than most owner-operated sets I see. Their machines are their revenue. Yours is a cost center you forget about.

The requirement is special. A silent canopy for a downtown site, a containerized set for a dusty quarry, parallel operation for a data hall expansion — buying a special-purpose machine for a temporary special-purpose need is how warehouses fill up with equipment nobody wants. Rent the weird one. Buy the boring one.

The call I make for clients

So here’s my rule of thumb, the one I give on the phone: under six months, rent — don’t even do the spreadsheet. Six months to two years, do the honest spreadsheet with the hidden costs above, and be brutal about your real timeline. Permanent load, year after year? Buy, and size the set right the first time.

And one more thing people get backwards: price isn’t the only comparison. Ask what a breakdown costs you. A rental company swaps a failed set, usually same day. Your own set breaking down means your mechanic, your parts lead time, your problem. For a hospital or a data hall, that risk alone can justify the rental premium. For a contractor who owns three sets and a good mechanic, it never will. If you want to run your own numbers past someone who’s seen both sides, talk to our engineers.

My Friday-morning contractor rented, by the way. Nineteen days, the set went back on a truck, and he never thought about it again. Sometimes the best generator is the one you never have to own.