Technical note

Four Versions of the Solar Question (and Why the Answer Keeps Changing)

2026-11-20 · Isabel Moreno

I answer service calls for a living. Not sales calls — the other kind. The ones that come in when an inverter is offline, when a roof is leaking around a penetration, or when something with feathers has moved into the array and nobody noticed for eight months.

That job changes how you read a question like "how much is a solar panel cost" or "do I need a battery." I mostly see the systems that went sideways. Which is why I'll say this up front: there isn't one answer to any of these. There are at least four, and they don't resemble each other.

What actually decides your answer:

  • What your utility pays you for export versus what it charges you for a kilowatt-hour. That ratio moves sizing decisions more than panel efficiency ever will.
  • Whether you need power when the grid is down. That's a resilience question. It is not a savings question, and mixing the two up is how people end up bitter about a purchase that was actually fine.
  • Whether the array already exists. New build and retrofit are different projects with different failure modes and very different cost curves.

Here are the four versions I run into. Find yours before you call anybody.

Scenario 1: Hawaii — the grid rules drive the design, not the sunshine

Hawaii is the one market where I'd tell almost anyone to start with the tariff sheet before the equipment list. Publicly reported average commercial rates have sat somewhere in the mid-thirty-cent to low-forty-cent range per kWh on Oahu, higher on the neighbor islands, depending on rate class and riders. Verify against the current HECO, HELCO or MECO filings, because rate classes and surcharges move and I'm not your interconnection engineer.

The thing that trips people up: Hawaii closed its original net energy metering program to new applicants back in 2015. What replaced it compensates exported power at a lower rate than what you pay to buy it. That single change flipped the design logic. Oversizing an array to dump power onto the grid is, more often than not, the wrong move now. Sizing to your daytime load and pairing storage to shift production into the evening is usually the better fit — which is also why "vivint solar hawaii" searches tend to end up in battery conversations rather than panel-count conversations.

On the brand point, since it comes up constantly: Vivint Solar's residential business was folded into Sunrun in 2020. If you're holding an older Vivint lease or PPA, your service and warranty requests route through Sunrun now. Nothing dramatic about it — just saves you twenty minutes on hold if you know it going in.

One warning on utility battery incentive programs in Hawaii: they've been capacity-capped and the terms have changed more than once. Anyone quoting you a per-kWh storage incentive from a two-year-old blog post is quoting fiction. Pull the current program page yourself.

Scenario 2: Mainland commercial, grid is reliable, you just want the bill down

This is the scenario where I disagree with the most people, so here it is plainly: if your grid is reliable and you have a working generator, a battery is not an ROI purchase. It's an insurance purchase. Price it that way or don't buy it.

I've watched facility managers run battery payback math that quietly assumed demand-charge savings they never actually captured, because the dispatch logic wasn't tuned to their load profile. The upside was a plausible six-year payback. The risk was $40,000 of storage sitting at 100% state of charge doing nothing. I kept asking myself whether the theoretical payback was worth defending that line item at budget review. Sometimes it was — usually when the building had a real outage history or a tenant with an SLA.

If you genuinely need backup, you need to be honest about what you're backing up. A single battery cabinet at a 400-amp service doesn't run your HVAC. It runs the server closet and the emergency lighting. That's a perfectly good reason to buy one. It's just not the same reason the brochure gave you.

Financing structure matters more here than people expect. Cash purchase captures the full federal credit but requires tax appetite. Leases and power purchase agreements move the system off your balance sheet and can make sense for nonprofits, some tenant-occupied properties, and entities that can't use the credit. Vivint Solar built its business on that leasing model before the Sunrun acquisition, and it's still a legitimate path — just read the escalator clause. A 2.9% annual escalator over 25 years is not a rounding error.

And if anyone promises you a guaranteed zero electricity bill, end the meeting. That promise doesn't exist. What exists is a modeled offset, and models have assumptions in them.

Scenario 3: You already have an array, and the problem isn't kilowatt-hours

This is most of my week. Two sub-problems come up over and over.

Birds under the panels

Solar panel bird guard installation is unglamorous and it's the single most-skipped line item on residential and light-commercial jobs. Pigeons and sparrows nest in the gap between the module and the roof deck. The nest material blocks airflow under the array, droppings etch the glass and corrode frame edges, and the wiring in that gap becomes a chew toy. We've pulled nests out of junction boxes.

Where it goes counterintuitive: installing critter guard clips and mesh costs meaningfully less at the time of the original install than as a retrofit, because the crew isn't working around live modules on a finished roof. Commonly quoted ranges run roughly $3 to $8 per linear foot, which puts a typical residential perimeter somewhere in the $500 to $1,500 band; commercial arrays scale with perimeter length and get real money fast. Get it quoted as a line item, not folded into "misc."

Two things I'd push back on. First, don't DIY a hardware-store mesh kit onto a finished array — you can void the roof warranty and, worse, the module clamping warranty, and mesh installed without drainage consideration traps water. Second, if you're in an area with a nesting season, that season is your deadline. March arrives whether or not your vendor's schedule does.

Surge protection, and what a joule rating actually means

Let's define it, because the surge protector joule rating meaning gets mangled in marketing copy constantly. A joule is one watt-second. On a surge protective device, the joule rating describes how much energy the device can absorb — per protected mode — before its metal-oxide varistors sacrifice themselves. Higher means it can take more hits before it dies. That's it. It does not tell you how well it protects.

I'm not an electrical engineer, so I won't pretend to spec your gear. What I can tell you is what to make your electrician or integrator show you, because joule count is the number consumers shop and it's the least useful one on the datasheet:

  • UL 1449 listing and Type (1, 2 or 3) — Type 1 goes on the line side of the main, Type 2 on the load side, Type 3 at the point of use. Different jobs.
  • Voltage protection rating (VPR) — the let-through voltage. A device with a big joule number and a lazy clamp still lets damaging voltage reach your equipment.
  • Surge current capacity in kA per mode — this is the number that actually corresponds to the joule figure people think they're buying.
  • MCOV — must exceed your system's normal operating voltage with margin, or you'll be replacing it every storm season.

The PV side is a separate device. DC-side SPDs are rated under IEC 61643-31 in kA with an 8/20 microsecond waveform, and they are not interchangeable with an AC panel device. Most of the lightning-adjacent failures I see land on inverter control boards and monitoring gateways — the cheap gear behind the expensive gear. Also: SPDs are sacrificial by design. If the status indicator has gone dark, the device already did its job and is now a plastic brick.

Scenario 4: You're still on "how much is a solar panel cost"

Don't price panels. Price the system in dollars per watt installed, because the module is maybe a fifth of your cost and the rest is racking, inverters, labor, permitting and interconnection.

For rough orientation only — these move, so treat them as a starting bracket and not a quote: publicly reported installed costs in the 2024–2025 window have generally run around $1.80 to $2.80 per watt for commercial rooftop and roughly $2.50 to $4.00 per watt for residential, before incentives. Hawaii runs higher on both, for shipping and labor reasons. Get three current quotes in writing and throw my numbers away.

The line items that blow up budgets and rarely appear in an online calculator: utility interconnection upgrades (a transformer change can be five figures on its own), roof condition and membrane age, structural reinforcement, and anything involving making a 1970s service panel accept a second source.

So which one are you?

Answer these five and the scenario picks itself:

  1. What do you pay per kWh, and what does your utility pay for export? If the gap is wide, you're a self-consumption project, not an export project.
  2. How many times has the power actually gone out in the last two years? If you can't remember, a battery is a want, not a need.
  3. Does the array already exist? If yes, you're in maintenance territory and the questions are about birds, surge protection and monitoring, not panel brands.
  4. What's your roof's remaining life? Under ten years, fix the roof first. I've watched people panel a failing membrane and pay twice.
  5. Can you use the tax credit? If not, the financing structures in Scenario 2 matter more than the equipment in Scenario 1.

One last thing, and it's the thing that's earned the most trust from me over the years: a decent vendor will tell you which of these four you are, even when the answer costs them the sale. We've talked commercial clients out of storage and into a generator plus a service contract, and we've told people their array didn't need what they called about. A supplier who can name their own limits — who says "this isn't our lane, here's whose it is" — is telling you something about how they'll handle the call at 9 p.m. on a Sunday. That's the vendor worth keeping.

Next: NMC vs LFP Battery Storage: A Procurement Comparison (and Where LG Energy Solution Fits)