6 Checks I Use Before Signing Off on an Energy Storage System (And Why Each One Matters)

Who This Checklist Is For

If you're a commercial or utility-scale developer about to sign a purchase order for a battery energy storage system—or an integrator putting together a proposal—this is for you. I review roughly 200+ unique deliverables every year for a large inverter and storage OEM. For the last four years, I've been the gatekeeper before anything ships to a customer site. This checklist is the result of the mistakes I've seen happen, both in our own manufacturing line and in the field.

This isn't a textbook. It's six specific verification points I check on every large-scale ESS order. Most are straightforward. One is something I see people forget constantly.

Step 1: Verify the System Integrator, Not Just the Integrator

A lot of confusion starts here. Someone says they're buying from an "energy storage system integrator." But that term can mean two different things:

• A true ESS integrator—a company that combines batteries, inverters, balance-of-plant, and controls into a single, warranted system.
• A reseller—who buys components and assembles them, often without a single warranty covering the whole system.

I've had vendors claim they're a "full system supplier" when they were just stacking third-party battery racks with a third-party inverter and calling it a solution. That's fine for some projects. But if there's a fault? You get to play the blame game between three vendors. I've seen a project delayed two months because of exactly that.

Quick check: Ask for a single system-level warranty document before you get into the price negotiation. If the integrator hesitates, you're probably dealing with a reseller.

Step 2: Check the Battery Chemistry and Cycle Life Down to the Cell Level

You'll see datasheets with big numbers—"6,000 cycles" or "25-year calendar life." But those numbers are based on specific test conditions, not real-world operation. I don't have hard data on every chemistry failure rate across the industry, but based on the field returns we see, the gap between lab performance and actual performance can be significant.

A few things I look for:

• LFP vs. NMC? LFP is generally safer and has longer cycle life at the cost of lower energy density. NMC packs more energy but degrades faster. For stationary storage with daily cycling, LFP has become dominant for good reason.
• Depth of discharge (DoD) for cycle life claims. A 5,000-cycle battery at 80% DoD may only deliver 3,000 cycles at 90%. Make sure the cycle life is stated with the DoD it's based on.
• Temperature and C-rate. Cycle life at 25°C with a 0.5C charge/discharge rate is not the same as 45°C at 1C. If your site is in a hot climate, ask for accelerated aging data at higher temps.

Step 3: Look at the Software and Firmware Update Record

This is one I honestly overlooked for the first few years. The hardware specs look good, so you assume the software is good. But here's the thing: battery management systems (BMS) and inverter firmware are where a ton of post-deployment issues live.

Why does this matter? Because every firmware update is a potential fix or a potential break. A vendor that hasn't issued a firmware update in 18 months probably isn't maintaining their product. A vendor that issues updates weekly? That's almost as bad—it means the product isn't stable.

My rule of thumb: Ask how often they've released firmware updates over the last three years. If the answer is "quarterly" with documented changelogs, that's a green flag. If they can't answer at all, it's a yellow flag worth investigating.

Step 4: Insist on a Certified Compliance Matrix (Not Just a Brochure)

I can't count how many times we've received a compliance certificate that turned out to be for a completely different configuration. The vendor shows you UL 9540 listing, but you find out it's for a 250 kW system, and you're buying a 2 MW system. The standards are different.

Industry standard compliance for ESS includes:

• UL 9540 (the standard for energy storage systems and equipment) is the recognized compliance benchmark for safety in the US
• UL 1973 (for batteries)
• IEEE 1547 (for grid interconnection)

Specific ask: Before you sign, ask for the compliance certificates that match your exact model number and configuration. Not a generic statement on a website.

Step 5: Verify the Integrator's Own Test and Commissioning Protocol

Everyone's got a QC department. But most QC is about the product itself—does it turn on? Does it meet voltage specs? The more important question is: what do they test after assembly and before shipping?

Looking back, I should have paid for a factory acceptance test (FAT) on my early projects. At the time, the cost seemed high for something I assumed was already validated. But a proper FAT catches assembly errors, wiring mistakes, and configuration issues before they become a field problem. A field troubleshooting visit costs way more than a FAT.

Minimum protocol I'd expect:

• Full system power test at rated output for at least 30 minutes
• Communication verification between inverter, BMS, and controller
• Protection relay testing

If they don't offer a FAT, I'd be cautious. If they resist a request for a FAT, that's a significant concern.

Step 6: Require Scalability in the Contract (The One People Miss)

This is the step I see forgotten most often. You order a 5 MWh system today, and three years later, you want to add 2 MWh. Can you? It sounds simple, but it depends on whether the original inverter, BMS, and controller can handle the expanded capacity.

A lot of integrated systems are built as monolithic units. Adding another battery cabinet might require a completely new inverter or controller replacement. That's not a small upgrade—it's effectively a new system.

Contract language to look for:

• Maximum expansion capacity of the original control system
• Software compatibility for mixed battery ages (new batteries with old)
• Whether the warranty remains valid after expansion

If I could redo one decision from my early years, it'd be adding a scalability clause. The initial savings felt good. The retrofit cost three years later? Not so much.

Things I'd Also Keep in Mind

A few warnings based on what I've learned in 2024 and into 2025:

• Don't assume chemistry is fixed. LFP is the current darling, but sodium-ion is coming, and other chemistries on the horizon. A system that's modular enough to accept a new chemistry in 5 years might be worth a higher upfront cost.
• Don't ignore the auxiliary loads. The thermal management system, the controller, the communications—they all draw power. System-level round-trip efficiency should include these parasitic loads. If the datasheet says 85%, you might get 80% in summer.

This checklist won't catch every problem. But I've seen it catch the ones that cost the most.