What a 100 kW Commercial Solar System Costs in California (2026)

The Real Cost Per Watt

NREL's last commercial rooftop benchmark, a 200 kWdc system modeled in Q1 2022, came in at a modeled market price of $1.84 per watt (2021 dollars, national average). NREL notes that actual costs vary significantly by region and installer, and a 2022 model is dated.

The more recent primary figures are LBNL's reported prices for systems installed in 2023. LBNL's Tracking the Sun (2024 Edition) put 2023 installed prices for small non-residential systems (100 kWdc or less) at $2.50 to $4.30 per watt and for large non-residential systems (over 100 kWdc) at $1.70 to $3.10 per watt, 20th to 80th percentile, national.

A 100 kW rooftop job sits right at the boundary between those bands, so its price can fall in either. Composition or membrane roofs with adequate existing electrical service land toward the bottom. Tile roofs, standing-seam metal, older switchgear that needs a service upgrade, or a utility that requires a full interconnection study push the number toward the top.

Get at least three itemized bids and compare them against these bands rather than against a single rule-of-thumb figure.

Where the Money Actually Goes

Hardware is not the majority of the invoice. NREL's commercial benchmark builds its price from module and inverter costs, structural and electrical balance-of-system, installation labor, sales tax, permitting, inspection and interconnection, EPC and developer overhead, contingency and profit — and the categories other than modules and inverters make up most of it.

In plain terms: on a commercial rooftop system, engineering, permitting, interconnection and margin cost about as much as the equipment, or more.

That's worth knowing when a quote looks high on a per-watt basis relative to a residential job. It generally isn't hardware markup; it's the fixed cost of engineering, permitting, and interconnecting a commercial-scale system. Ask any installer quoting a 100 kW job for an itemized breakdown along those lines. If they can't produce one, that's a flag on its own.

The Tax Credit Deadline That Actually Matters Now

Section 48E pays 30% of the qualified investment for a facility with a maximum net output under 1 MW AC, and 100 kW is comfortably under that line. Below that line the 30% rate does not depend on the prevailing-wage and apprenticeship requirements, per 26 U.S.C. §48E(a)(2).

What changed is the deadline. The cutoff comes from the statute: §48E(e)(4) ends the credit for solar property placed in service after December 31, 2027, and Public Law 119-21 (signed July 4, 2025) applies that rule to facilities whose construction begins after July 4, 2026, twelve months after enactment.

Projects that began construction by that date are outside the December 31, 2027 cutoff. For any project not yet underway, the system has to be placed in service by December 31, 2027 or the credit is zero; ask a tax professional what placed-in-service evidence you will need.

That leaves roughly 15 to 16 months from today to design, permit, interconnect, and commission a system.

IRS Notice 2025-42 had let a solar facility of 1.5 MW AC or less, such as a 100 kW system, show that construction began either through physical work of a significant nature or through a 5% cost safe harbor; the U.S. District Court for the District of Columbia vacated that notice in full on June 6, 2026, so ask a tax professional which method applies to your facts.

For a project starting today with no cost yet incurred, the number that matters is the placed-in-service deadline, not the safe harbor. Under 26 U.S.C. §168(e)(3)(B)(viii), §48E qualified property and energy storage technology are 5-year property, and §168(k) provides a 100% first-year allowance for qualified property acquired after January 19, 2025.

Under §50(c)(3), the depreciable basis is reduced by half of the credit, leaving about 85% of the qualified cost depreciable when the credit is 30%. Get the exact interaction of the credit, the basis reduction, and any state-level depreciation treatment confirmed by a CPA before it goes into a payback model.

Annual Production: What 100 kW Actually Generates

No primary source publishes a single production figure for a 100 kW system in California; output depends on tilt, azimuth, shading and local weather. Coastal and fog-belt locations produce less than Central Valley and desert Southern California locations, which get higher irradiance and fewer overcast days.

Before signing a contract, run the address through NREL's PVWatts calculator (pvwatts.nrel.gov), which accounts for actual tilt, azimuth, shading, and local weather data. Any proposal for a system this size should include a PVWatts-derived, or equivalent, production estimate with the specific assumptions shown, not just an annual kWh number with no methodology behind it.

If a proposal only gives a single number and no assumptions, ask for the modeling.

Demand Charges: Why Solar Alone Won't Fix Them

Demand charges are set in each utility's commercial tariff, and this page does not repeat the secondary, installer-blog figures that circulate for them. Pull the current B-19/B-20, TOU-8, or AL-TOU rate sheets directly from pge.com, sce.com, or sdge.com before building a payback model that leans on a specific demand-charge figure.

The mechanism matters more than the exact dollar figure: fully offsetting demand charges usually requires a battery timed to discharge during the peak window, not just solar generation. And the program that used to subsidize commercial batteries for exactly this purpose, SGIP, is not currently funding new applicants.

As of September 22, 2026, both non-residential SGIP storage categories, Large-Scale Storage and Non-Residential Storage Equity, show closed at Step 5, at listed rates of $0.25/Wh and $0.85/Wh respectively. That means a battery sized to chase demand-charge savings is currently a fully out-of-pocket cost on top of the ITC-eligible solar system, not a subsidized add-on.

Net Billing: What You Get for Exported Power

California's commercial customers at PG&E, SCE, and SDG&E sit on the CPUC's Net Billing Tariff, the successor to net metering that most people still call NEM 3.0 even though that name technically describes the residential version.

Export compensation is based on avoided cost rather than the retail rate you pay to buy power, and the CPUC says it is usually lower than the retail rate, though it can rise above it on late summer evenings.

This is standard CPUC policy, laid out on the commission's own Net Energy Metering and Net Billing program page, and it applies to commercial and residential customers alike at all three utilities.

The practical upshot: a 100 kW system sized to export a large surplus back to the grid usually earns less for that power than it is worth used on-site. Right-sizing the system to your actual daytime load, rather than maximizing kW to fit the roof, usually produces a better payback than over-building and relying on export credits.

How to Check Payback Under Current Rules

Run the numbers straight, on your own quote. Start with the installed price. The 30% Section 48E credit comes off that price for a system under 1 MW AC, so a bid of $250,000, for example, would carry a $75,000 credit. That's the ITC in isolation.

Some installers quote lower net-cost figures by also stacking the 100% bonus depreciation benefit described above on top of the credit; the value of that depends entirely on the buyer's marginal tax rate and isn't a figure that applies the same way to every business. Don't take a stacked net-cost number at face value.

Ask the installer to itemize exactly which incentives and depreciation elections produce it, and confirm the depreciation math with your own CPA.

For the energy side, multiply the PVWatts production estimate for your address by the share you will use on-site and your tariff's energy prices, and value exports at the export credits instead. Divide the net cost by that yearly value for a simple payback.

No primary source publishes a typical payback period for a 100 kW system, and demand-charge-heavy load profiles change the answer unless the system is paired with a battery, which, per the SGIP closure above, currently carries no rebate.

The binding constraint for anyone starting today is the calendar: construction has to finish and the system has to be placed in service by December 31, 2027, to keep the 30% credit at all.

Vet the Contractor Like It's a Six-Figure Bet, Because It Is

California's residential solar dealer network has taken serious hits. Freedom Forever filed Chapter 11 in April 2026 and was converted to Chapter 7 liquidation by court order on August 7, 2026, leaving its homeowners with no operating company behind their systems.

Sunnova filed Chapter 11 in June 2025; Solaris Assets completed an acquisition of its assets and servicing platform that September, and SunStrong Management now runs day-to-day operations on the legacy portfolio. Neither entity is originating new agreements. SunPower filed Chapter 11 in August 2024.

All three were primarily residential, dealer-network businesses, so their direct relevance to a commercial 100 kW EPC contract is lower, but the underlying lesson applies at least as hard on a six-figure commercial job: verify the contractor's CSLB license class and standing directly at cslb.ca.gov, confirm bonding, and ask how long they've operated as a commercial EPC specifically, not just how long the parent brand has existed.

A firm that can't produce a current CSLB license number and a bond certificate on request shouldn't get anywhere near a six-figure contract, commercial or residential.