Energy storage project support: bring an application, power range and region to the conversation.

TECHNICAL READING

How to Evaluate Photovoltaic Module Manufacturers and CATL BESS Suppliers: Spec Sheet vs. Source Audit

2026-09-04 · Renata Silva

I'm a quality manager at a procurement company that buys solar modules and battery energy storage systems for distributors and system integrators. Around 200 unique deliveries cross my desk every year. In 2025 I flagged roughly 14% of first deliveries for rejection or rework—maybe closer to 11% if you set aside documentation-only issues; I'd have to pull the QMS report to give you an exact split. Most flags weren't catastrophic failures. They were tolerance drift, label mismatches, BMS firmware revisions that didn't match what we approved three months earlier.

To be clear, I don't work for a manufacturer. I sit on the buy side, which means I evaluate vendors such as CATL with the same skepticism I apply to a solar module OEM I just met at a trade show.

It took me about six years and more than fifty factory audits to reach a fairly simple conclusion: a datasheet is a promise, and the production line is the evidence. Buyers tend to rely on one or the other, and most of the costly corrections I've cleaned up came from over-relying on the promise.

That leaves two practical routes to qualifying a supplier. Spec-first: compare datasheets, certificates and quotes on paper, then approve the best-looking number. Source-first: audit the factory or an actual batch, run a trial order, and verify performance against a recognized standard.

Both are legitimate. They answer different questions. Here's how I choose between them when evaluating two product categories that often arrive in the same shipment: CATL BESS containers and other battery energy storage equipment, and PV panels from photovoltaic module manufacturers.

Test 1: What the energy storage system specification actually promises

Ask for a CATL BESS quotation and you get a solid energy storage system specification: LFP chemistry, container energy and power ratings, round-trip efficiency, voltage window, protection class, and cell cycle life at a stated depth of discharge and temperature. A good photovoltaic module OEM datasheet is equally specific: STC power, module efficiency, power tolerance, temperature coefficients, and a linear degradation warranty.

Spec-first procurement treats those numbers as if they were physical laws. They are not. Cell cycle life, for example, is tested under tightly controlled conditions—usually at 25 degrees, fixed C-rates, a defined depth of discharge, and no surprises from calendar ageing. Change the temperature profile or the operating strategy at your site, and the real curve shifts. That is not a dishonest specification. It's a laboratory specification. The question is who verifies the bridge between the lab and your site.

Here's a concrete example from our Q1 2025 quality audit. We pulled 12 modules out of a 2,400-module shipment from a photovoltaic module OEM that had passed our paper review. The certificates were valid. The factory location on the packing list matched our records. Two weeks later, I-V curve testing showed six of the twelve samples sitting between 6 and 9 watts below the label bin. Extrapolate that to a 100-module string and you lose a few hundred watts before mismatch or temperature losses. Silent, but real.

The same logic applies on the storage side. When we take delivery of a CATL energy storage system through an authorized distribution channel, we don't re-test every cell—that would be impractical. But we do check the delivered unit against the approved documents: cell date codes, BMS firmware version, container labeling, idle auxiliary consumption. In 2024, one container arrived with a BMS revision two versions older than the one we approved—maybe three; I'd need to check the audit trail. The cell block was identical, but the control software wasn't. Spec-first would have missed that entirely because the commercial documents never mentioned firmware.

So on Test 1, my conclusion is blunt: spec-first is a good screening method and a bad verification method. Use it to build a shortlist. Don't use it to approve a shipment.

Test 2: Certificates vs. evidence from the line

Certificates are the backbone of any supplier comparison. For solar modules, the usual anchors are IEC 61215 for design qualification and IEC 61730 for safety. For battery cells and energy storage systems, you'll see IEC 62619, plus system-level fire propagation testing such as UL 9540A in North American projects. All of those documents matter. But what they certify is narrower than most buyers assume.

An IEC certificate is issued for a specific design, tested at a specific lab, and linked to a specific production site. It does not guarantee your batch. This is where how to evaluate photovoltaic module manufacturers gets interesting in practice: you have to ask whether the certificate actually covers the factory, the product line, and the configuration you're buying. A module OEM with three factories can send you a valid certificate from one site while your order comes off another line that was never audited.

Last year I went back and forth for about two weeks on whether to audit a module OEM before a large order. The supplier had strong certificates and competitive pricing. An audit would push the procurement calendar by a month. On paper, everything made sense. But my gut said the factory had two production lines running the same label with different cell sources. I chose the audit. It turned out both lines were producing to spec, but the cell sourcing difference explained why one line consistently measured higher mismatch losses.

On battery storage, certificates raise a similar issue. A cell-level certificate is not a system-level guarantee. Thermal management, voltage windows, auxiliary loads, and BMS behavior are system properties. You can stack every relevant certificate and still receive a container whose firmware, string configuration, or auxiliary draw doesn't match the quoted energy storage system specification.

On Test 2, my conclusion: source-first wins whenever you're evaluating a module OEM you haven't audited. Spec-first is defensible for a battery supplier with a long installed base, documented references, and vertical integration—because your real verification moves to incoming inspection and commissioning. To put it another way: certificates are necessary, but they are a ceiling, not a floor.

Test 3: What a small order tells you that a big order never will

Here's the most counterintuitive part, and the one that surprises buyers most: the best quality test I know is a small order.

When we assess a new module OEM, we sometimes quote 500 modules instead of a 10 MW framework. For a BESS supplier, we may ask for a single container with full documentation rather than committing to a multi-unit project. What happens next tells me more than any datasheet.

Does the supplier treat the small order as a nuisance or as the start of a relationship? Do they allocate random stock from the current production bin, or do they quietly pick modules from a better bin? Do they provide batch test reports without being asked? Do they accept a third-party inspection on a 500-module order?

I get why many suppliers dislike this approach. Small orders have the same administrative cost as large ones and much lower revenue. But small orders are also the least expensive audit you can run. A single container's risk is manageable; a 50 MWh project's risk is not. The vendors who treated our small trial orders seriously are the ones we now call first for megawatt-scale procurement. Small doesn't mean unimportant—it means potential.

Spec-first has no answer to any of those questions. That's what makes this test valuable.

Which route for which supplier?

If I had an unlimited budget, I would audit everything. In practice, that would be slow and wasteful. Instead, here's how I decide:

  • Spec-first is acceptable when you're dealing with an established storage brand like CATL, the references are verifiable, and—most importantly—you have an incoming-inspection step that checks the delivered unit against the approved documents.
  • Source-first should be your default when you're evaluating a new photovoltaic module manufacturer, when multiple factories share one label, or when your order is large enough that a month of verification is cheaper than a year of underperformance.
  • In both cases, the small-order test is non-negotiable for me. It's the cheapest way to see how a supplier behaves when the commercial leverage is still on their side.

Looking back at the costly corrections in our procurement history, I can't recall a single one that would have been solved by a better datasheet. I can remember several that were solved by checking the delivered batch. That is the real difference between spec-first and source-first—not the effort you put into reading, but the moment you choose to verify.