Demand Charges Are Eating Your Power Bill — Can Solar (and Storage) Fix It?

Short answer: mostly no — not solar by itself. A demand charge is set by your single highest sustained power draw in the billing period (commonly the highest average over a ~15-minute interval), and that peak usually doesn't line up with peak sun. The lever that actually moves a demand charge is a battery, which can discharge to flatten that peak no matter what the sun is doing. In a joint NREL/Berkeley Lab study, standalone commercial solar cut demand charges by a median of just ~7%, while solar paired with storage cut them by a median of ~42% (NREL/LBNL, McLaren & Mullendore). If demand charges are eating your bill, the panels are the cheap part of the energy story — the battery is what touches the demand part.

Below is what's actually on your bill, why solar alone rarely helps the demand line, what a battery changes, and the 2026 federal tax math that — between Section 48E and an active legal fight over the construction deadline — makes the timing genuinely urgent for commercial operators.

First, what's actually on your bill

A commercial and industrial (C&I) electricity bill is really two bills stapled together:

• Energy charges ($/kWh) — you pay for total energy consumed over the month. This is the part most people picture.
• Demand charges ($/kW) — you pay for your single highest sustained draw of power, typically the highest average usage in a short interval (commonly 15 minutes) during the billing period. It doesn't matter that the peak lasted only 15 minutes; that one interval can set the charge for the whole month.

Demand charges are not a niche line item. Per NREL's survey of more than 10,000 utility tariffs, roughly 5 million U.S. commercial customers are subject to demand charges at or above the level where battery storage starts to pencil out (NREL). And they're not small: NREL reports demand charges can account for roughly 30%–70% of a commercial customer's electricity bill — a wide range that varies a lot by utility, tariff, and how spiky your load is (NREL). Not every C&I customer is at 70%; some are far below 30%. But if you run a spiky operation, this is often the bill line you actually feel.

Who feels it most? Operations with sharp, short load spikes:

• Warehouses — banks of forklifts charging at once, dock equipment, conveyor startups.
• Manufacturing — large motors and presses spiking on startup.
• Cold storage — compressor cycling and morning pull-down, when everything kicks on at once.

These are exactly the load profiles where one bad 15-minute window can dominate the bill.

Why solar alone barely touches your demand charge

Here's the mechanism: your demand charge is set by one bad 15 minutes, and solar only helps if it happens to be producing during that window.

If your monthly peak lands at 6–7 a.m. (cold-storage pull-down), at 6 p.m. (an evening shift starting up), or under a passing cloud that knocks out your array for a few minutes, solar simply isn't generating then — so it can't shave that peak. Berkeley Lab's work on this is blunt: demand-charge savings from solar are "generally negligible" when the charge is based on peak demand at any time of day, and become "more significant" only when the tariff defines the peak as a designated daytime window (for example, maximum demand during roughly a 12–4 p.m. period) (LBNL/NREL, Darghouth et al., LBNL-1007030).

That's why the median demand-charge reduction from standalone commercial PV came in at only ~7% in the NREL/LBNL synergies study — modeled for customers with PV sized to generate about 50% of annual load, under a common non-coincident demand charge — and why reductions were below ~15% in roughly 90% of the cases analyzed (NREL/LBNL).

One caveat: that ~7% is a median across NREL's simulations, not a promise. Your specific building could do better or worse depending entirely on when your peak occurs.

What solar alone does do well

Solar still attacks the energy ($/kWh) half of your bill, and that's real money. It shaves total consumption during daylight hours, it's the cheaper technology per kW of capacity, and it generates a predictable hedge against rising utility energy rates for decades.

So set expectations correctly: buy solar for energy savings, not for demand-charge relief. Be skeptical of any pitch claiming panels alone will gut your demand charges.

Enter the battery: peak shaving

A battery changes the game because it breaks the dependence on the sun being out at the right moment.

Peak shaving, in plain English: the battery charges when it's cheap or convenient (off-peak hours, or off your solar during the day), then discharges during your peak 15 minutes to cover part of the spike — so the meter never records as high a draw. Because the battery can fire precisely when your peak occurs, it directly attacks the interval that sets your demand charge.

That's why the combination beats the sum of its parts. In the NREL/LBNL study, solar-plus-storage delivered a median ~42% demand-charge reduction versus ~7% for solar alone — and the researchers found that combined savings are almost always greater than the sum of what either technology achieves separately (NREL/LBNL). The solar can recharge the battery and cover energy use; the battery handles the peak. They cover each other's weaknesses.

A battery also adds resilience — backup power that keeps refrigeration running or a line moving through an outage. For cold storage (spoilage risk) and manufacturing (downtime cost), that's not a footnote; it can be a core reason to do the project.

Solar vs. battery vs. both — at a glance

Figures are median results across NREL/LBNL simulations (source); a specific site can land well above or below them. Costs are relative, not quotes.

When does solar actually help your demand charge?

There are real cases where solar pulls its weight on the demand line — just narrower than the pitch suggests:

• Tariffs with a daytime peak-demand window. If your utility defines the demand charge by your max draw in, say, a noon–6 p.m. window, solar production overlaps it and can meaningfully help (LBNL).
• Broad midday-peaking loads. Operations whose biggest sustained draw genuinely happens in the middle of the day.
• Panel orientation at the margin. West/southwest-facing arrays push production later into the afternoon, which Berkeley Lab found can be marginally more effective at nudging coincidence with a late-day peak (LBNL).

The single most useful thing you can do before spending a dollar: pull your actual interval (15-minute) data and find out when your monthly peak occurs. That one fact largely determines whether solar helps your demand charge at all — or whether you need a battery to touch it.

The 2026 federal money — and the clock

Here's where commercial and residential diverged sharply. The residential federal credit (Section 25D) is dead for systems placed in service after December 31, 2025 — the IRS states the credit "will not be allowed for any expenditures made after December 31, 2025," treating an expenditure as made when the original installation is completed (IRS OBBB FAQ). This article is commercial-only for exactly that reason.

The commercial Section 48E Clean Electricity Investment Credit, by contrast, survived:

• 6% base rate, increasing 5× to 30% for facilities meeting prevailing-wage and registered-apprenticeship (PWA) requirements. The IRS confirms a key exception: a qualified facility with maximum net output of less than 1 MW (AC) qualifies for the full 30% rate without meeting PWA — but state both rates and don't just assume 30% (IRS — Section 48E; IRS — PWA FAQ).
• +10 percentage points for meeting domestic-content requirements, and +10 percentage points if located in an energy community — two stackable adders (IRS — Section 48E). Whether a specific project clears those thresholds is a detailed determination for your tax advisor.
• Energy storage is independently eligible. A standalone battery qualifies for 48E on its own — Notice 2025-42 expressly carves energy storage technology out of the wind/solar credit-termination rule (§ 2.02, citing § 48E(e)(4)(C)) (IRS — Notice 2025-42). That matters here: the battery, your best demand-charge tool, keeps its credit.
• 5-year MACRS depreciation. Solar property and energy storage technology are treated as 5-year property for accelerated depreciation, for property placed in service after Dec. 31, 2024 (IRS — Cost Recovery for Clean Energy).

One note that doubles as a caution: the live IRS 48E page still displays phase-out boilerplate saying the credit phases out "for the later of 2032 or when U.S. greenhouse gas emissions from electricity are 25% of 2022 emissions or lower" (IRS — Section 48E). For wind and solar that language has been superseded by OBBBA's begin-construction termination. Read the operative guidance, not just the credit landing page.

There's also a depreciation wrinkle — claiming 48E reduces your depreciable basis by 50% of the credit (the "ITC basis haircut," IRC § 50(c)), and OBBBA is commonly cited as restoring 100% bonus depreciation. But these are tax-treatment rules, not promised dollar outcomes — confirm them with a CPA (IRS — Pub. 946).

The begin-construction deadline (and the legal curveball)

The reason timing is urgent: under OBBBA, wind and solar projects that begin construction after July 4, 2026 must be placed in service by December 31, 2027 to keep the 48E credit — while projects that begin construction before July 5, 2026 generally have until December 31, 2030 under the continuity safe harbor (placed in service no more than four calendar years after the year construction began) (IRS — Notice 2025-42). That's a big swing in runway depending on which side of one date you fall on.

How you prove you began construction is, as of mid-2026, genuinely unsettled — and you should know that going in:

• Notice 2025-42 made the Physical Work Test the sole method (physical work of a significant nature plus a continuous program of construction) and largely eliminated the old 5% cost safe harbor, leaving it only for low-output solar facilities (maximum net output ≤ 1.5 MW AC) (IRS — Notice 2025-42).
• Then, on June 6, 2026, the U.S. District Court for D.C. vacated Notice 2025-42 in its entirety and remanded it to the IRS — temporarily restoring the broader 5% safe harbor — but an appeal is expected, and certainty is unlikely before the July 4, 2026 deadline (Foley Hoag LLP).

The outcome isn't settled yet. The practical guidance: treat the July 4, 2026 deadline as real, and confirm your safe-harbor strategy with tax counsel before relying on either method. For a deeper walk-through of the deadline mechanics, see our companion piece on the July 4, 2026 commercial solar deadline.

How to actually run the numbers

A checklist before you commit capital:

1. Pull 12 months of interval (15-minute) data and find out when your monthly peak occurs. This single fact tells you whether solar helps your demand charge at all.
2. Get your tariff's demand-charge structure — the $/kW rate and how the peak window is defined (any-time-of-day vs. a daytime window). The window definition is the difference between "negligible" and "significant" solar savings.
3. Model solar and battery separately. Solar for energy ($/kWh) savings; battery for peak ($/kW) shaving. Don't let one number hide the other.
4. Have a CPA confirm the tax picture before you assume a payback — the exact 48E rate you qualify for, the basis haircut, and bonus-depreciation eligibility.

And a caveat on payback: battery economics are highly site- and tariff-specific. Any payback figure should be presented as a CPA-confirmed range or estimate tied to your real data — never a flat promise. The right battery size and solar-to-storage ratio depend on your interval data, not on a rule of thumb.

Get the real numbers for your facility

If demand charges are eating your bill, the model that matters is one built from your interval data. PanelPerks connects warehouse, manufacturing, and cold-storage operators with vetted commercial installers who model demand charges from real 15-minute data and the current 48E timeline — including the July 4, 2026 deadline and the live safe-harbor uncertainty. There's no charge to get matched, and you can see how the process works before you talk to anyone.

For the broader policy picture across every incentive in play this year — what survived, what's paused, and what's dead — start with our 2026 solar incentives guide.

---

Verified as of June 2026. Federal energy-credit timing rules are in active litigation (IRS Notice 2025-42 was vacated June 6, 2026, with an appeal expected) — confirm any tax position with your CPA or tax counsel before relying on it.