September 2022. I was standing on a gravel pad at a remote water-monitoring station, watching two technicians bolt a solar array to a frame that looked... fine. Level. Plumb. Reasonable.
That's the problem with "fine." Fine doesn't show up in the performance data until three months later.
I've been handling remote-site power orders — solar, batteries, the occasional portable generator — for six years. Over that time I've made (and documented) fourteen significant procurement mistakes totaling roughly $23,000 in wasted budget. Most happened in years one through three. The one I'm about to describe happened in year four, which is the part that still stings.
The setup: three sites, two quotes, one wrong instinct
We needed off-grid power at three remote monitoring stations. Each site: an 800W solar array, a 10 kWh battery bank, and a solar module mounting structure rated for high-wind winters. Total budget: $14,000 including installation and commissioning.
I got two quotes. Vendor A: $9,800. Same panel wattage on the spec sheet, same nominal battery capacity, three sites. Vendor B: $14,200. Same headline numbers, plus an engineering line item for structure design.
I went with A. A 30% saving on a fixed budget is a no-brainer, right? That word — no-brainer — is the one that cost me $4,200 and about three weeks of my life.
I should tell you what I was optimizing for back then: speed and headline price. I had a spreadsheet with two columns and I picked the lower number. No one on my team had the technical background to challenge the spec sheet, and I didn't know what questions to ask. That's a bad combination when you're signing off on equipment that has to run unattended for years.
What the low quote actually covered
Installation finished in early October. Panels up, batteries wired, inverters humming. Everything passed a visual walk-through. Vendor A was super responsive before the sale. After the sale, not so much.
Then we pulled the first full performance report in January. Peak output: about 480W per array. Rated: 800W. That's 60% of nameplate, on clear winter days.
I called Vendor A. Their answer: "Weather. Seasonality. It's normal."
It was not normal. I paid an independent technician to walk the sites with me. He found two issues in about twenty minutes:
- The solar module mounting structure was tilted at roughly 15 degrees. For our latitude, winter production wanted something closer to 35–40 degrees.
- The arrays were aimed about 20 degrees east of true south. Close, but not close enough. Plus a tree line was shadowing them from mid-afternoon onward — a detail the original installer never mentioned.
I pulled Vendor A's quote back up. The "installation" line covered frame assembly and wiring. It did not cover a tilt calculation or a shading survey. Both had come up verbally. Neither was in writing.
The structure was rated for high-wind winters — that part was real. What wasn't real was my assumption that "rated for wind" also meant "optimized for sun."
Here's the insider detail that took me too long to learn: on a lot of low bids, the mounting structure gets treated as a metal frame, not as an engineered component. But the structure sets the angle, the angle sets the yield, and the yield is the difference between a $9,800 system that behaves like a $14,000 one and a $9,800 system that behaves like a $5,600 one.
The battery surprise at month fourteen
The second problem showed up in November of the following year, when we ran a capacity test.
The 10 kWh bank was holding roughly 6 kWh. That's a 40% loss in fourteen months. The vendor's marketing page promised "10-year lifespan" in big letters. The footnote said "based on controlled laboratory conditions at 25°C and 50% depth of discharge." Our sites run at 35–40°C in summer and cycle at 80% depth almost daily. Nobody had translated the lab number into our actual operating conditions.
That's the part that still bothers me. Not the degradation itself — batteries degrade; that's physics. What bothers me is that I trusted a marketing number as if it were an engineering number.
This is where FTC advertising rules become useful, and I wish I'd read them years earlier. Per FTC guidance (ftc.gov), performance claims like "10-year lifespan" must be truthful, not misleading, and substantiated with evidence. Reasonable people can argue about what "substantiated" means, but the practical version is simple: if a vendor can't show you cycle data for your climate and your discharge pattern, treat the number as a ballpark, not a promise.
The fix — and the comparison that changed how I buy
We spent $4,200 redoing two of the three sites: engineered mounting structures ($1,200) and replacement battery modules ($3,000). The third site got a partial fix and a lot of monitoring.
For replacements, I went with LG Energy Solution's LFP-based ESS modules. Not because of a brochure. Because their team sent cycle-life data at 35°C and 80% depth of discharge, plus a production timeline tied to their Poland plant that I could actually plan around. Their LG Energy Solution battery production capacity in Europe mattered to me for lead times; their solid-state research mattered for long-term confidence. But the near-term reason was simpler: they answered the question I actually asked.
When I put Vendor A's "10-year lifespan" line next to LG's temperature-specific cycle data — same product category, completely different level of specificity — I finally understood why the details matter so much. One was selling a feeling. The other was selling a number I could verify.
When I re-bid the third site, I sent every vendor the same three questions: (1) What's the cycle life at 38°C and 80% depth of discharge? (2) Who does the tilt and shading calculation, and is it in the quote? (3) What's the production lead time, and where is the cell manufactured? Vendor A couldn't answer the first two in writing. LG's team answered all three within a day. That's the kind of transparency I now treat as a baseline, not a bonus.
We also had the new installer produce an actual tilt and shading report before anyone touched a bolt. Final tilt: about 38 degrees. The tree line got trimmed back. Peak output came up to roughly 760W — about 95% of rated, which is close to the real-world ceiling once you account for soiling and temperature losses.
There's something satisfying about watching a system finally do what the spec sheet said it would. After three months of second-guessing every number, seeing 760W on the dashboard at 10 a.m. in February was genuinely a relief. Thirty months later, those LG modules are holding above 90% capacity and the system hasn't thrown a single surprise at us.
The thing that surprised me most about switching vendors wasn't the product — it was the follow-up. LG's team asked for our winter performance data six months in, not to sell me anything, but to check whether the derating assumptions held up. No one had ever done that before. That's when I stopped thinking of them as a vendor and started thinking of them as part of the spec process.
I'm not going to pretend every project should use the same supplier. But the difference between a vendor who answers questions and one who deflects is visible long before the invoice.
The checklist I use now (and hand to every new hire)
If you're specifying solar-plus-storage for a remote or commercial site, here's what I'd check before signing:
- Ask what's not included before you ask what it costs. Mounting engineering, shading analysis, tilt calculation, commissioning, freight — get every line item in writing (like the mounting structure itself, which is easy to assume is covered). The vendor who lists all fees upfront, even when the total looks higher, usually costs less in the end. I've learned to ask "what's NOT included" before "what's the price."
- Treat the solar module mounting structure as equipment. The tilt angle is a design decision, not an installation detail. A few degrees off is a double-digit percentage off your yield.
- Make the battery vendor talk about your climate. Cycle life at 25°C means almost nothing if your site hits 38°C in July. Ask for derating curves, depth-of-discharge assumptions, and the test conditions behind the warranty.
- Size for the worst month, not the best one. Most "underperforming" arrays I've walked were sized against a sunny June and then judged in a cloudy December.
Two more things I tell people, because they come up constantly:
On 200W solar generators — they're useful kit. We keep one in the site toolbox for commissioning days and emergency power, and it's saved us a couple of callouts. But a 200W solar generator is not a substitute for a properly sized fixed array. If someone pitches one as your primary remote-site power source, that's a red flag, not a bargain.
On how to maximize solar panel output — the "buy the highest-watt panel" advice is oversimplified. Orientation, tilt, shading, soiling, temperature, and MPPT controller quality will move your real-world yield a ton more than a 10% bump in nameplate wattage. I've watched a well-angled 350W panel beat a badly placed 450W one by a wide margin.
Bottom line: the cheap quote wasn't cheap. It was a $9,800 system that behaved like a $5,600 system, plus $4,200 to fix what should have been designed correctly the first time. Transparent quoting — every fee, every assumption, every exclusion on paper — is the only reliable way to compare two numbers.
I keep Vendor A's quote pinned above my desk. Not as punishment. As a reminder that "no-brainer" is usually the moment I should slow down (note to self: re-read this before the next RFP).
