Installer SupportResource Center

PROJECT READING

ESS Procurement by Scenario: Why My $52,000 Sourcing Mistake Won't Fit on a Single Advice Sheet

2026-11-17 · Camille Lefevre

In Q2 2023, our team placed an order for a commercial ESS deployment — a few megawatt-hours behind the meter, mid-size project. The spec sheets had LG Energy Solution branding on them, the thermal data looked right, and the distributor's lead time was three weeks shorter than the second bid. I signed off. (Note to self: this is exactly the kind of "clean" purchase that should make alarm bells ring.)

Four months later, we were running performance tests and the degradation numbers were nowhere near the "typical" curve the distributor had been showing at trade shows. Not catastrophic. But off enough that we had to renegotiate the customer's SLA and eat roughly $52,000 in restructured system design. The product wasn't defective. The problem was that I'd pulled our residential screening checklist and applied it to a commercial deployment. My mistake, not the battery's.

Every time I scroll through energy storage system ESS news, the advice splits into two camps: "buy the cheapest LFP" or "buy the most premium NMC," as if everyone should follow the same rule. I don't think there's a universal ESS procurement answer. What you ought to do depends on which of three deployment scenarios you're actually in — and they need completely different sourcing logic.

The three scenarios (and how to tell which one you're in)

Here's the classification my team uses now. Before we even open a specification sheet, we slot the project:

  1. Utility-scale / front-of-meter. 20 MWh and up. Long interconnection timelines. Performance guarantees are the thing you fight about.
  2. Commercial and industrial (C&I). 100 kWh to a few MWh. Mixed buyers — facilities teams, EPCs, sustainability managers. Integration partners matter more than cell brand.
  3. Residential and small commercial. Sub-100 kWh. Often installed by electrical contractors and handled downstream by solar battery maintainers, not the original purchaser.

Scenario A: Utility-scale — buy the performance guarantee, not the cell

On utility-scale projects, the cell itself is almost a commodity line item. What gets you in trouble is the round-trip efficiency guarantee and the capacity degradation curve written into the PPA. On a $5M+ project, a 2% slippage in round-trip efficiency over 15 years can be worth more than the battery procurement spend itself.

The pitfall I've seen (and once helped cause): teams load up on cell-level spec comparison — pulling datasheets from the LG Energy Solution official website and comparing Wh ratings — but they don't push hard enough on the system-level efficiency guarantee. The cell manufacturer's spec is a ceiling, not a floor.

Two things I'd insist on now:

  • A performance guarantee backed by measured availability data from at least three commissioned sites of similar size and climate — not a "typical" curve from a brochure.
  • Liquidated damages tied to actually measured round-trip efficiency, not modeled efficiency.

On price: the lowest per-kWh bid will almost always win the tender sheet. And that's the wrong number. If you can get 4% better round-trip efficiency for $18/kWh more in capex, the NPV advantage is usually north of $40/kWh over 15 years, depending on your dispatch profile. That's before downtime costs.

So glad I double-checked the round-trip efficiency guarantee language on the last tender. Almost signed a version that used "modeled" instead of "measured." That one word would have been a six-figure problem by year three (unfortunately, that project had already absorbed one of those before I joined the team).

Scenario B: C&I — look at the integrator, not the cell badge

This is where I made the $52,000 mistake. In C&I, the deployment is usually small enough that no single manufacturer wants to own the whole system, and bulky enough that a residential-grade install crew can get it wrong.

The counter-intuitive bit: on most C&I ESS deployments, the deciding factor in lifecycle cost is not the cell technology. It's the integrator. LFP vs. NMC matters — but a good LFP integration with proper thermal management will beat a poorly integrated NMC every time on 10-year TCO. I've watched this play out on both sides now.

One specific communication pitfall that cost us three weeks: our purchase order said "ESS unit, 500 kWh capacity." The integrator's contract said "ESS unit, 500 kWh nominal capacity at 25°C." Those are not the same thing at our site temperatures. We were using the same words but meaning different numbers. We discovered it when the acceptance test results came in 8% below what our dispatch model had assumed.

What I check on the integrator side:

  • How many systems of comparable size they have actually commissioned this year — not how many they plan to.
  • Whether they have a real thermal management plan documented, not just "we followed the OEM recommendation."
  • Response time on service calls, verified with reference calls — their sales team's number is not a data point.

To be fair, if you're a large enough C&I buyer, you can sometimes go direct to the OEM. But I've watched that go badly for teams without in-house ESS engineering — the OEM will happily sell you modules, and then you discover that installing and integrating them is a completely different skill set. (Note to self: stop assuming procurement ends at the P.O.)

One more thing I'd flag in this scenario: a lot of spec comparison happens by staring at trademarked logos. The LG Energy Solution logo on the datasheet is a quality signal — the cells are fine — but it tells you nothing about the integrator's thermal design. That's the variable that actually moves the number.

Scenario C: Residential — the maintenance contract matters more than the battery

This is where I see the worst advice floating around, and honestly, I'm not entirely sure why. The economics of a home battery are so different from a utility or C&I one that comparing them feels almost like comparing a car to a bike.

The residential mistake I want everybody to avoid: buying the cheapest $/kWh battery and assuming the installer will handle everything. Solar battery maintainers — the people who actually get called when a system underperforms — will tell you that a solid majority of their service calls are not cell failures. They're inverter communications issues, firmware drift, and BMS calibration problems that fall on whoever's still around in year four.

So the real decision in the residential segment is: who's fixing this in 2033?

  • A long-term maintenance contract with the installer matters more than a 5% price advantage on the battery itself.
  • Some brands (including LG's RESU line) have a service network that survives even if your installer doesn't. That's not always worth the premium — but it's worth knowing whether your installer has that fallback.
  • Check whether the firmware update path is handled by the manufacturer or requires a site visit. If it's the latter, you're paying for a technician every two years for no good reason.

Granted, this segment changes fast, and a lot of my benchmarks come from 2023–2024 deployments. The market is evolving quickly. Your mileage may vary if you're installing for a jurisdiction with different interconnection rules or net-metering tariffs.

How to figure out which scenario you're actually in

You'd think classification would be obvious from project size. It's not always. A 400 kWh system for a grocery store can pull you into utility-style interconnection requirements. A 5 MWh system behind the meter can be a C&I problem funded like a utility one. Size doesn't classify it — the interconnection point does.

Three questions to run through before you start sourcing:

  1. Who owns the interconnection risk? If it's you, you're in utility territory. If it's a third party, you're closer to C&I.
  2. Who's on the hook for year-7 maintenance? If the answer isn't clear before you sign, you've found your problem. I've never fully understood why this question isn't asked first on every project. Perhaps because maintenance discussions are boring until they cost you a week.
  3. What's the failure mode that actually scares you? If it's performance guarantee shortfall, Scenario A. If it's integration problems, Scenario B. If it's "who calls me when this thing glitches," Scenario C.

I can only speak to the deployments I've touched — mostly mid-size C&I in the EU, plus a handful of utility projects. If you're doing grid-scale work in Asia-Pacific, there are buyer dynamics I'm not familiar with. Someone else has that playbook.

One last thing. The $52,000 mistake didn't come from buying a bad product. It came from assuming a good product would behave the same way in every context. If I could wave a wand and undo one habit, it wouldn't be the purchase order. It'd be the sourcing shortcut that preceded it. (Note to self: don't shop for a category, shop for a scenario.)