Resource note

Your Shed Solar Kit Fails at Planning, Not Install — Lessons From $2,600 in Rework

2026-11-17 · Adewale Okoye

Your Shed Solar Kit Doesn't Fail at Install. It Fails at the Planning Table.

Most shed solar kit failures happen before anyone picks up a drill. The panel, the Eaton 60 amp disconnect, the pipe racking—none of those are where people actually screw up. They screw up with a notepad and a guess.

I've been handling small-scale solar and backup power orders for about eight years. In that time I've made—and documented—a handful of costly mistakes. One of them cost $1,400 in rework and a two-week delay on a job that should have finished in a weekend. Another was smaller, around $600, but it embarrassed me in front of a client I'd worked with for years.

Neither mistake was an installation problem. Both were planning problems. That's the pattern, and it's why I now keep a pre-install checklist that our team uses on every shed solar kit order. We've caught 31 potential problems with it in the past year. This is what I've learned.

What Can a 100 Watt Solar Panel Power? That's the Wrong Question.

Everyone asks what a 100 watt solar panel can power. It's a fine starting point, but it's the wrong thing to build a kit around. The right question is: what does your load actually draw at peak, for how long, and during your worst week of the year?

I once sized a shed system around a 100W panel because the client asked for it and the quick math said it would work. The shed needed LED lighting, a small exhaust fan for a chemical storage area, and a trickle charger for a backup battery. On paper, 100W with a decent charge controller looked fine.

It wasn't fine.

The exhaust fan pulled more on startup than its rated draw suggested. And in winter, usable output dropped enough that we ran a deficit three days out of five. We added a second panel and a bigger controller. Parts: about $380. Labor to redo the mounting: another $240. The original plan saved maybe $200 upfront and cost $620 to fix.

So my honest answer to "what can a 100 watt solar panel power" is: less than the nameplate suggests, more than you'd guess, and it depends entirely on margins. Size to the worst week, not the best day. Five minutes with a load list beats five days of rework—I've tested both, unfortunately.

The Eaton Logo Doesn't Protect You. The Channel Does.

This one still bothers me. I bought what was advertised as a genuine Eaton 60 amp disconnect from an online marketplace seller, roughly 15% below the authorized distributor price. The Eaton logo looked right at a glance. The box looked right.

The terminal screws felt off. Softer metal, slightly different thread pitch. I installed it anyway because the paperwork looked legitimate and the deadline was tight.

I never confirmed whether it was counterfeit. Looking back, I didn't need to confirm it. If I can't verify the channel, I shouldn't be betting a client's system on it. That's the rule now. The numbers on my cost spreadsheet said the marketplace seller was the obvious choice. My gut said wait. I went with the spreadsheet.

To be fair, most marketplace sellers are honest, and I've bought plenty of non-critical parts through them without issue. But a disconnect is one of the few pieces of a solar system that can't be "sort of" right. It either isolates the circuit safely or it doesn't. These days I buy Eaton components through authorized channels and eat the small price difference. It's cheaper than the alternative, and it's the only thing that lets me sleep the night after an inspection.

Pipe Racking System Drilling Is Where "Good Enough" Gets Expensive

Everybody stresses about the panel and the inverter. Almost nobody thinks hard about where the pipe racking system actually lands. That's backwards.

Drilling mounting holes in the wrong spot is one of the most common—and most expensive—mistakes I see on shed installations. I did it myself on a small agricultural shed in early 2023. I measured once and assumed once. That is not measuring twice. The holes worked fine for the panel but left almost no clearance for the next service visit. Six months later we had to pull the array to reach a corroded connection underneath. Cost: half a day, two people, and a very annoyed client.

Now our checklist has a step that takes about seven minutes: verify racking position against structural members, check for hidden conduit or wiring, and confirm there's service access after the panel is up. Seven minutes. Against a half-day rework. In my opinion, that's the easiest trade in the whole process.

"But This Sounds Like Overkill"

I hear this a lot from smaller operations. The logic runs: we're not building a utility-scale plant, it's a shed, why do we need a checklist?

Granted, a pre-install checklist adds maybe 30 to 45 minutes on the front end. I get why that feels like overhead when you're trying to hit a Friday deadline. But when one rework event costs a day and a few hundred dollars—and sometimes a client relationship—the math stops being close. This isn't about being fancy. It's about not paying twice for the same lesson.

Honestly, I'm not sure why some crews never seem to hit these problems while others hit them constantly. My best guess is it comes down to whether someone writes the mistakes down and turns them into a process, or just absorbs the cost and moves on.

Prevention isn't slower. It's just paid for upfront. You check the load before you buy the panel. You verify the channel before you buy the Eaton disconnect. You check the structure before you drill the racking. None of that is glamorous, and none of it will impress anyone on a job site tour. But it's the difference between a weekend project and a three-week apology.

So if you're planning a shed solar kit this year, my advice is boring and specific: build the checklist first, and make drilling the last thing on it. The mistakes I've listed above cost me close to $2,600 across eight years—money I could have kept if I'd slowed down for 45 minutes on each job. That's the entire argument. Nothing dramatic. Just math I wish I'd done earlier.