Renewable energy technical article header
Engineering Notes

One Project, Three Rejections: A 2024 Quality Audit Diary on Solar Specs

Posted on 2026-11-10 by Adewale Okoye

On a Tuesday morning in February 2024, I stood on a loading dock with a clipboard and a rejection sticker. The supplier rep next to me was arguing that the 200 modules in front of us were "within industry standard." I had a spec sheet that said otherwise, and a project developer who was about to miss a milestone.

That moment wasn't dramatic. It was just the third rejection in four months on the same project—an off-grid solar-plus-storage install for a rural telecom site, roughly $180k in equipment. The weird part was the product mix. It wasn't just panels. It was JA Solar modules, a LiFePO4 battery bank, a car inverter we'd repurposed for backup, and a Jackery portable power station for the field crew. I'd never had to audit that combination in one go.

Here's what happened, product by product, and what I'd tell you if you're sourcing the same stuff.

First: the JA Solar module spec that didn't add up

The module in question was a JA Solar JAM66D42—a 66-cell bifacial panel from the DeepBlue 4.0 line. The supplier's quote sheet said "bifaciality: 80%." Our approved spec said "bifaciality: 70–80% depending on installation geometry." Those are not the same claim.

If you've ever looked into the ja solar jam66d42 bifaciality percentage, you know the number is a range, not a fixed point. It depends on rear-side irradiance, mounting height, albedo, and temperature. The datasheet gives you a nominal figure with a tolerance. The supplier was quoting the top of the range as if it were guaranteed. On a 200-module order, that's the difference between a performance model that works and one that doesn't.

I pulled the actual spec from the JA Solar official website—not the supplier's PDF, not a distributor's marketing page. The official datasheet listed the bifaciality range and the test conditions. I printed it, highlighted the tolerance, and walked it back to the supplier. He shrugged and said, "Everyone quotes the high number."

Maybe. But I don't sign off on "everyone." I sign off on the spec.

This is the part that still bothers me: the supplier wasn't lying. He was doing what a lot of the industry does—rounding up to the most attractive number. That's a habit that's been drifting for years. Five years ago, bifacial was a premium feature and the numbers were conservative. Now it's standard, and the marketing has gotten loose. The fundamentals haven't changed—a bifacial panel still needs the right mounting and albedo to hit its rear-side gain. But the way people talk about it has changed, and not for the better.

Second: the LiFePO4 BMS low-voltage cutoff that nobody checked

The battery bank was a 48V LiFePO4 pack with a third-party BMS. During the pre-delivery audit, I asked the supplier what the per-cell low-voltage cutoff was set to. He said "2.0V." I asked him to double-check. He came back with "2.5V." Then he came back again with "it depends."

That's a problem. The lifepo4 bms low voltage cutoff per cell is not a cosmetic setting. Set it too low and you'll get a pack that looks fine in testing and loses capacity in the field. Set it too high and you'll get premature shutdowns that the end user will blame on the panels. Most reputable BMS units let you configure it, and most integrators have a house standard—usually somewhere in the 2.0–2.5V per-cell range, with the exact number depending on discharge rate and warranty terms.

We ended up rejecting the first BMS batch because the cutoff was factory-set at 2.0V with no documentation. The vendor's response was, "It's within the cell manufacturer's absolute minimum." That's true. It's also not the same as "it's the right setting for this application."

Here's the mindset shift I'd offer: a BMS setting is a design decision, not a default. If your supplier can't tell you why the cutoff is set where it is, that's not a spec. That's a guess.

And the communication piece was ugly. We were using the same words—"cutoff," "minimum," "recommended"—and meaning different things. I discovered this when the vendor sent a configuration file that showed 2.0V while the quote sheet said 2.5V. Nobody was lying. The sales rep had quoted the cell maker's recommended minimum; the engineering team had quoted the integrator's house standard. Two correct numbers, one wrong assumption. That one cost us two weeks and a re-test.

Third: the car inverter that wasn't rated for what the label said

The backup power side of the project needed a small inverter to run a few critical loads during outages. Someone on the team suggested picking up a power inverter for car autozone use—cheap, available, and "probably fine." I looked at three units. All three listed a peak wattage that was roughly double the continuous rating, and none of them listed a derating curve for ambient temperature.

That's not a knock on retail inverters. It's a knock on reading the box instead of the spec. A 1000W peak / 500W continuous inverter is a 500W inverter. If your load is 700W, it will work—until it doesn't. For a telecom backup application, "until it doesn't" is not an acceptable failure mode.

People think a higher price means better internal components. Sometimes it does. More often, it means the vendor paid for the testing that lets them publish a real number. The price reflects the documentation, not necessarily the hardware. We ended up specifying a unit with a published continuous rating, a temperature derating table, and a THD figure under 3%. The cost difference was about $140 per unit. On a project with four backup nodes, that's $560 for a spec you can actually defend.

Fourth: the Jackery question everyone asks

The field crew wanted a portable power station, and the question that came up—predictably—was where to buy a Jackery solar generator cheapest. I get why people ask. These units aren't cheap, and the price varies a lot by channel.

But here's what I told the team: the cheapest channel is only the cheapest if nothing goes wrong. If you buy from an unauthorized reseller and the unit fails in month three, you're not dealing with Jackery's warranty process—you're dealing with a return window that's already closed. The official site, the major marketplaces, and the warehouse clubs all have different return policies and different levels of warranty support. Price is one variable. It's not the only one.

My rule of thumb: buy from a channel that will still be there when you need service. That's usually the manufacturer's own store or an authorized retailer. It's not always the lowest number on the page.

What I took away from the audit

Three rejections, one project, four months. None of the products were defective. Every single one of them was "within spec" in some loose sense. The problem was that "within spec" had been stretched to mean "close enough to the number on the marketing sheet."

I've only audited maybe 40 projects of this size, mostly in North America and the Middle East. If you're working in utility-scale or in a market with different certification requirements, your experience will probably differ. But the pattern I keep seeing is the same: the gap between the datasheet and the delivery is where quality lives or dies.

What was best practice in 2020—read the spec, trust the vendor, spot-check the delivery—doesn't quite hold in 2025. The products have gotten more complex. The settings have gotten more configurable. The channels have gotten more fragmented. The fundamentals haven't changed: verify the number, document the setting, and don't assume the label tells the whole story. But the execution has transformed.

If you're sourcing any of this stuff, do one thing for me: pull the spec from the manufacturer's own site, not the quote sheet. Then ask your supplier to explain the number. If they can't, you've found your first audit item.

Author avatar

Adewale Okoye

Adewale Okoye is a battery energy storage systems analyst specializing in containerized BESS, power conversion systems, energy management systems, HVAC, fire detection, auxiliaries, and grid services. He uses IEC 62933-5-2:2025 system-safety requirements and UL 9540A thermal-runaway propagation evidence while measuring usable energy, rated power, dispatch duration, auxiliary demand, round-trip efficiency, response time, state-of-charge limits, and separation assumptions. His engineering guides help developers, utilities, and integrators define system boundaries, safety documentation, control interfaces, performance guarantees, and commissioning criteria.

Leave a Reply