Project note

Vivint Solar Credits, Battery Fires, and the Checks Nobody Runs Before Signing

2026-11-06 · Renata Silva

The short version

Three things decide whether a commercial solar-plus-storage project is a good deal, and none of them show up on page one of any proposal I've been handed:

  1. Who legally receives the solar energy credits. If you sign a lease or a PPA, the third party that owns the system usually claims the federal credit, the state SRECs, and the depreciation. You get the kilowatt-hours. Those are not the same thing, and on a mid-size array the gap is often five figures a year.
  2. What listing the battery cabinet carries. UL 9540 for the system, UL 9540A fire testing behind it, and an install that actually matches NFPA 855. A spec sheet that says "UL-listed cells" is telling you about the cells. Not the cabinet. Not the spacing.
  3. Who read the as-built drawings. Every monitoring portal and MPPT solar charge controller app I've touched reports what the hardware chooses to report. It is a dashboard, not a safety instrument, and definitely not a revenue meter.

The system you're buying is not the equipment. It's the contract, the listing, and the as-built drawings. Get those three right and the panels become almost interchangeable. Get them wrong and you can have a perfectly functioning array that loses money for a decade.

Why you should weigh my opinion, and also not

I've been handling commercial solar and battery storage procurement for 8 years. I've personally made and documented 14 significant mistakes totaling roughly $61,000 in wasted budget — some of it ours, some of it a client's, all of it avoidable. I now maintain the pre-signature checklist our team uses on every project.

In my first year (2017), I approved a set of deep-cycle lithium packs for a warehouse fleet without checking the charger's profile. It looked fine on paper. Six months in, two packs had swollen and one had gone into a hard shutdown that took a dock door offline for a day. That mistake cost about $4,100 in replacements plus a very unpleasant Monday.

The expensive one was September 2022. We signed a third-party-owned PPA on a 340 kW roof and our internal pro forma assumed the federal credit and the state SRECs flowed to us as offsets. They didn't. They flowed to the owner — which is how third-party ownership is supposed to work. Our model had double-counted roughly $22,400 of first-year benefit that never arrived. Nobody lied to us. We just never asked the question out loud.

After the third permit rejection in Q1 2024 — an AHJ that wanted separation distances documented against the 2023 edition of NFPA 855 rather than the 2020 edition — I stopped trusting my memory and built the checklist. We've used it on 31 projects since. It has caught something on 19 of them.

Point 1: The credits follow ownership, not the roof

This is the part that trips up almost every commercial buyer, and it's worth being precise about it.

In a direct purchase, you own the system. You claim the federal commercial credit under IRC §48 (or §48E for projects placed in service after 2024), you claim any state SREC or production incentive, and you depreciate the asset on a 5-year MACRS schedule with bonus depreciation. As of January 2025, the base commercial credit sits at 30%, with adders for domestic content, energy community siting, and prevailing wage plus apprenticeship compliance. Verify the current numbers with your tax counsel — the rules were rewritten in August 2022 and have been moving since.

In a lease or PPA, the third party owns the system. They claim all of the above. You get a fixed or escalating rate per kWh, and whatever the contract says about escalators, buyouts, and end-of-term transfers. That's not a trick. It's the entire business model, and it's why leasing works for buyers with no tax appetite.

To be fair to Vivint Solar here: this structure is standard across the whole industry, and it's not something Vivint invented or hid. But it matters that Vivint Solar as a brand is largely gone — Sunrun closed its acquisition in October 2020 and folded the residential origination business in. If you're searching "solar energy credits available for Vivint Solar," you're most likely looking at a legacy lease, an inherited contract from a property acquisition, or a resale disclosure. In those cases the credits are almost certainly not yours, and the useful question isn't "how do I claim them" — it's "what does the transfer clause say, and who's on the hook for the roof warranty."

One more thing people conflate: net metering credits are not tax credits. They're a utility billing mechanism, and they've been shrinking. California's NEM 3.0 took effect in April 2023 and the CPUC's own modeling put the average export compensation cut at roughly 75%. If your 2019 pro forma assumed retail-rate export, it's wrong now. That's not a Vivint problem or a Sunrun problem. It's the whole industry's problem, and it's why storage suddenly pencils where it didn't before.

And on the hardware itself: the vivint solar panel on your roof was, in most vintages, a commodity Tier-1 module from a rotating set of manufacturers. The model number on the as-built drawing is what governs your warranty. Not the brand on the truck.

Point 2: What actually causes lithium battery fires

Here's the honest framing: thermal runaway almost always needs two things at once — a trigger, and a failed barrier that was supposed to stop the trigger from becoming a fire.

The common triggers:

  • Overcharge — charge current continues past full, electrolyte breaks down, pressure and heat build.
  • Internal short — manufacturing defects, contamination, or lithium dendrites formed by charging below 0°C.
  • External heat or physical damage — a cabinet next to a hot surface, a forklift impact, a compression failure.
  • Wrong charge profile — lead-acid absorption and float stages applied to a lithium pack that expects constant current then constant voltage.

The failed barriers are usually the BMS not disconnecting, the pack not being tested as a system, or the installation not matching the listing.

Which brings me to the marine question, because it's a genuinely useful illustration. A counterpart at one of our sister sites manages a small golf-cart and utility-vehicle fleet. They bought West Marine lithium batteries — solid product, sold for deep-cycle marine use — and dropped them into carts still running the original lead-acid taper chargers, because the marketing said "drop-in replacement."

That's the historical myth worth killing: the "drop-in lithium" thinking comes from an era when LiFePO4 packs were rare and vendors competed on ease of adoption. That's changed. A lithium pack is only a drop-in if the charger's profile, the low-temperature cutoff, and the alternator or DC-DC regulation all get addressed. On boats that's what ABYC E-13 exists for. On land it's why NFPA 855 and UL 9540A exist.

I'm not going to claim I know exactly which cell failed in their case, because I don't. What I know is that the BMS was reporting normal right up until it wasn't, and that the charger had never been reconfigured.

From the outside, a battery cabinet looks like a box with a fan. The reality is that its fire behavior is a function of cell chemistry, spacing, suppression, and whether anyone ran the propagation test.

Point 3: Your MPPT controller app is a dashboard, not a meter

I like these apps. I use them. They're genuinely better than they were in 2018. But I've been burned twice by trusting them.

The one that still bothers me: a 2023 site where the monitoring portal showed production running about 11% above our model. Great news, right? It wasn't. The production CT had been clamped on the load side of the service, so the portal was counting the same electrons twice. Everything looked excellent until the utility true-up, when the actual export numbers came in and the gap was obvious. Roughly $6,800 of forecast savings that didn't exist, and a very awkward conversation with a CFO.

The MPPT sweep logic itself is usually fine. What isn't fine is assuming that a green checkmark means the array is healthy. A charge controller app will happily report voltage and current from a string that's dirty, partially shaded, or mis-oriented — because it's measuring what reaches the terminals, not what reaches the terminals correctly. Faults that produce current look like production.

Two habits fix most of this. First, commission with an independent meter, not the vendor's portal. Second, put eyes on the as-built drawing set before you accept the system — the CT placement, the string map, and the firmware revision should all match what's physically installed. That last one is where we caught a controller running firmware two revisions behind its own release notes.

Where all of this stops being true

I'd rather be honest about the edges than pretend this applies everywhere.

If you're a small commercial site under roughly 50 kW buying cash, most of the credit complexity evaporates — you claim §48 directly, and the real risk is documentation, not structure. If your utility still offers full retail net metering, storage arbitrage often doesn't pencil at all and you're better off waiting. If you're in a jurisdiction that hasn't adopted a recent NEC cycle or NFPA 855, permitting is a different animal — sometimes easier, and "easier" is not automatically better.

And if you're a residential customer rather than commercial, don't map any of this onto your situation. The residential credit under §25D works differently, the lease structures are different, and the decision is a household budget decision, not a depreciation schedule.

Demanding an as-built review costs money and adds a week. On a 12 kW system, that's not worth it. On anything with a battery over 20 kWh, I'd argue it's the cheapest hour of the project.

Renata Silva

Renata Silva

Renata Silva is a photovoltaic module analyst covering monocrystalline solar panels, bifacial modules, TOPCon and heterojunction designs, glass-glass construction, junction boxes, and module warranties. She interprets IEC 61215 and IEC 61730 evidence while comparing rated power, conversion efficiency, temperature coefficient, bifaciality, insulation, mechanical-load results, degradation assumptions, and tolerance. Her technical guides help EPC engineers, distributors, and project buyers separate qualification evidence from site-specific energy yield, climate exposure, installation constraints, and long-term performance risk.