-
The short version
-
Why you should weigh my opinion at all
-
Mistake 1: the 48V forklift battery charger we forgot to count
-
Mistake 2: fuel pump vs fuel injector, and the $2,100 guess
-
Mistake 3: a Kohler 5E marine generator sized by habit
-
Mistake 4: reading the gasoline number on a dual-fuel unit
-
Where the total cost actually lands
-
Where this checklist stops working
The short version
Three numbers decide whether a generator project goes smoothly: the coincident charging load, the derated output on your actual fuel, and the fuel pressure at the injector inlet. Get those wrong and you'll spend money on the wrong fix — I've done exactly that three times, for about $4,300 in total.
Here's what I'd check before signing anything:
- Add every continuous load that can run at the same moment — including the 48V forklift battery charger bank nobody counted because it's "not really part of the building."
- Take the propane output number, not the gasoline number, if you're running a 4500 watt inverter generator dual fuel setup. The derate is real and it isn't small.
- On diesel units, rule out the low-pressure side before anyone touches an injector. Fuel pump vs fuel injector is a $180 problem vs. a $2,000 problem, and the symptoms overlap.
Everything below is the reasoning, the receipts, and the places where this advice stops applying.
Why you should weigh my opinion at all
I'm a facilities maintenance lead handling generator and power equipment orders for nine years. I've personally made (and documented) eleven significant mistakes on these projects, totaling roughly $4,300 in wasted budget. Now I maintain our team's pre-purchase checklist, mostly so I stop repeating myself.
That checklist has caught 47 potential sizing or fuel-system errors in the past 18 months. Not glamorous. But it means fewer trucks rolling out at 6 a.m.
Mistake 1: the 48V forklift battery charger we forgot to count
September 2022. A 22,000 sq ft warehouse, standby power for the whole building. I walked the floor with a clipboard, added lighting, dock doors, two rooftop units, the office HVAC, and a small server closet. Came up with a little over 60kW of connected load, spec'd a Kohler generator with headroom, felt good about myself.
What I missed: four 48V forklift battery chargers on the far wall, each pulling hard during the first couple hours of a charge cycle. Forklift chargers don't draw a flat load — they pull near maximum current early and taper off as the battery comes up. The facility's charging window overlapped with shift change, which is also when the dock doors run and the lights are all on. Worst-case coincidence, not average load.
The result wasn't dramatic. Nobody's power went out. But when we ran the transfer test, the chargers tripped on input undervoltage one after another, and the unit sat there loaded past where I'd wanted it. We ended up re-doing the load calc and stepping up the unit. That error cost about $1,900 in change-order labor and a three-week delay on the commissioning date.
Two things I do differently now. First, I count chargers as continuous loads, because during a charge cycle they are. Second, I write down the *simultaneity* assumption explicitly — "we assumed HVAC and charging never overlap" is a sentence that will get you in trouble, and writing it down forces you to defend it.
Worth noting: battery charging rooms have their own requirements independent of the generator. OSHA 1910.178(g) covers charging area ventilation and eyewash provisions, and NEC Article 480 covers battery installations generally. That's a separate conversation from generator sizing, but on a facilities project they show up in the same drawing set, so review them together. Verify the current edition adopted in your jurisdiction — local amendments vary more than people expect.
Mistake 2: fuel pump vs fuel injector, and the $2,100 guess
This one hurt. A Kohler SDMO diesel generator at a job site started cranking long, running rough under load, and pushing white smoke. My instinct — and the instinct of the tech I called — was injectors. Diesel guys love injectors, because that's where the drama is.
Wrong. The low-pressure lift pump was weak, and the filter housing was pulling air. The high-pressure side was fine. We replaced a fuel pump for a couple hundred dollars and the unit ran clean. The earlier injector quote had been around $2,100 in parts and labor for four remanufactured injectors plus recalibration.
Here's the distinction that matters, and it's the one I now explain to every new tech on our team: the fuel pump's job is *delivery* — moving diesel from the tank to the high-pressure side at adequate volume and pressure, with the water separator and filters in between. The injectors' job is *metering and atomization* — spraying a precise quantity at a precise moment. A weak lift pump starves the injection system, which produces symptoms that look exactly like bad injectors: rough idle, low power, white smoke, hard starts.
So the order of operations is: tank pickup and fuel level, water separator, primary and secondary filters, lift pump pressure and volume, air leaks on the suction side, then — and only then — the high-pressure side. On a common-rail engine, rail pressure readings tell you a lot before you touch a wrench to an injector. Don't skip the cheap steps because the expensive answer is more interesting.
Rule we now enforce: no injector replacement without a documented lift-pump pressure and volume test on file. No exceptions, no "the tech could tell by ear."
Mistake 3: a Kohler 5E marine generator sized by habit
Boats are where generator assumptions go to die. A customer wanted a Kohler 5E marine generator to run air conditioning and a galley load on a 34-footer. The 5E is a fine little unit — 5kW, diesel, quiet, and common on that size of boat. I sized it by looking at the nameplate on the A/C and adding up the rest, which is the wrong method on water.
Marine loads have inrush. A compressor starting is a momentary current draw several times its running current, and a 5kW set is not enormous. What actually fixed it was a soft-start on the A/C and a documented load shed sequence — not a bigger generator. Cheaper, too, and it kept the customer from dragging around an extra few hundred pounds of machinery.
The other thing I under-weighted: installation conditions. Exhaust routing, cooling water, and ventilation are not accessories on a marine set, they're the difference between a working install and a boat full of carbon monoxide. ABYC P-1 governs exhaust system installation for auxiliary engines, and ABYC E-11 covers AC and DC electrical systems on boats. If your installer shrugs at those references, find a different installer. Cheaper than a surveyor finding the problem later.
Mistake 4: reading the gasoline number on a dual-fuel unit
Short one. A 4500 watt inverter generator dual fuel is often bought for exactly the right reason — propane keeps, gasoline doesn't — and then operated on an assumption that doesn't hold. Propane has lower energy density, so the same engine makes less power on it. The datasheets I've pulled for 4500W-class dual-fuel units show something in the neighborhood of a 10% derate on propane, which puts you closer to 4,000–4,100W. Check the specific datasheet for the unit you're buying; the number moves between models, and altitude makes it worse.
Everything I'd read said "dual fuel means you get the same power either way." In practice, on a job site at 5,000 feet with an RV air conditioner and a microwave, that 10% is the difference between a comfortable afternoon and a breaker that keeps tripping.
Where the total cost actually lands
The temptation is to compare unit price and move on. The $500 quote stops being cheaper the moment it turns into $800 after a change order, a re-calc, and a second truck roll. I now build the estimate in five lines: equipment, installation and site work, fuel over five years at expected run hours, scheduled service and consumables, and a risk line for the things I've historically gotten wrong.
That risk line is not padding. Mine is roughly 8% on new-build installations and near zero on replacements where I can measure the existing load. It came directly out of the $1,900 I burned on the charger mistake.
Where this checklist stops working
Two honest limits. First, if the generator carries life-safety loads — hospital, egress lighting, fire pump — this is not your framework. That's NFPA 110 territory, with a licensed engineer, load testing documentation, and classification requirements that don't care about your spreadsheet. Second, my numbers are commercial and light-industrial; marine installations need a surveyor's sign-off and a sea trial under real load, and dual-fuel portables are consumer equipment where the datasheet really is the authority.
And one caveat on the diesel section: everything I said about ruling out the low-pressure side assumes the engine is mechanically sound and the fault is fuel-related. If you've got a compression or timing issue, no amount of filter changing will save you. That's the point where the checklist hands off to a diagnostic tech — knowing where your checklist ends is part of using it well.
Leave a Reply