Net Metering 101: How It Works in 2026

Every solar quote rests on one assumption: what is one exported kilowatt-hour worth? Get that number right and the payback projection is honest. Get it wrong — or let an installer quietly assume rules from five years ago — and you can overpay by thousands of dollars while believing the math works.

It is also the number most often modeled wrong. And in 2026 it matters more than ever: the federal 30% residential tax credit expired on December 31, 2025, so the export rule is now the export math.^[16]

What net metering actually is

A grid-connected solar system creates three energy flows, and confusing them causes most bad solar math:

• Solar you use immediately. Power goes from your panels straight into your home, so you never buy that kilowatt-hour at retail. This is self-consumption.
• Solar you export. When production exceeds demand — usually midday — surplus flows through a bidirectional meter to the grid.
• Power you import. At night and in winter you buy from the utility at retail.

Net metering ties flows two and three together. Each exported kilowatt-hour earns a bill credit at the full retail rate — the same price you pay when you import — so a credit earned at noon can be spent at 9 PM, dollar for dollar. Think of it as a rolling store credit at the full retail price.

Net billing breaks that link. Imports still bill at retail, but exports earn a separate, lower rate — wholesale price, "avoided cost," or a regulator-set value. It is the difference between a store credit and selling back at wholesale; in most markets net billing values run 20–50% of retail^[3].

So the real question is never "does my state have net metering?" It is three: what is one exported kilowatt-hour worth, how long does the credit last, and which charges survive on my bill no matter how much I export?

The three compensation models

Nearly every program in the country is a variant of three structures.^[3]

Export-value ranges from SurgePV's installer reference guide tracking US policy through 2026^[3]; full-retail state list cross-checked against TheGreenWatt's April 2026 survey^[13]. Always confirm against your utility's current tariff sheet.

Net metering is the most generous, because exports are worth exactly what imports cost; it rewards oversizing, with a design target of 100–110% of annual usage. Net billing is stricter: with midday exports earning a fraction of retail, design shifts to daytime load plus battery charging — typically 60–80% self-consumption without storage — and oversizing destroys return^[3].

Buyback and feed-in structures pay a fixed price for exports, sometimes for a 10–20 year term. European feed-in tariffs are the classic examples; the closest US analogues are municipal programs and negotiated buyback plans in deregulated markets. Compare the buyback to your retail rate — if it is lower, and it almost always is, design for self-consumption^[3].

Watch the credit mechanics, not just the headline rate. Monthly rollover carries credits forward indefinitely and favors the customer. Annual true-up cashes out surplus at avoided cost, so seasonal overproduction earns pennies. Credit expiry forfeits unused credits outright^[3].

Why 2026 changes the stakes

The residential clean energy credit under Section 25D dropped to 0% for systems placed in service after December 31, 2025. The operative date is placed in service, not contract signed. Section 48E still supports commercial and third-party-owned projects, which is why lease and PPA offers remain viable — but a homeowner buying outright gets no federal credit in 2026^[16].

Strip out the subsidy and your return depends on two things: the retail rate you avoid and the export rate you earn. The first is set by your utility's rate schedule. The second is set by policy — and that policy is in motion everywhere. The N.C. Clean Energy Technology Center counted 253 distributed solar policy actions in Q1 2026 alone, with 61 addressing net metering; 44 states plus DC and Puerto Rico took some action that quarter^[6].

The direction of travel is one-way. Programs move toward net billing as solar penetration rises, because utilities argue full-retail credits shift grid costs onto customers without panels^[3][5]. That argument has now been tested in court in the largest market in the country — and it won.

How export compensation sets your payback period

Here is the whole article in one table. Take one identical 8 kW system producing 11,000 kWh a year, on a home that uses 45% of that production on site and exports the other 6,000 kWh. Only the export rule changes.

Illustrative modeling using a $0.17/kWh retail rate, near the 2026 US residential average of 17.91¢/kWh^[17], and export values from published 2026 ranges^[3][13]. Payback effects scale with local system cost — see our solar panel cost guide for current $/watt figures.

Notice what the table does not say. A weak export rate does not make solar worthless — it makes unused exports worthless. The 5,000 kWh that system sends into the house is still worth about $850 a year at retail under any regime. Export policy only governs the surplus, which is why the fix is to shrink the surplus by consuming or storing more of your own production.

Non-bypassable charges: the quiet haircut

Even under "full retail" net metering you rarely receive the entire retail rate on an export. Most states deduct non-bypassable charges — the parts of your bill every customer pays regardless of usage, and which cannot be avoided by generating your own power. In California they fund public-purpose programs such as energy efficiency and low-income rate assistance; similar charges exist elsewhere under different names^[14].

The export credit is the full retail rate minus the non-bypassable portion — usually a few cents per kilowatt-hour, small enough that it rarely changes a decision but large enough that an "exact 1:1" claim is not literally true^[14]. NBCs are also a permanent floor on your bill: even a system that nets to zero energy still owes fixed customer and non-bypassable charges. That is why "your electric bill will be $0" is essentially never accurate, and why the honest comparison is combined post-solar cost versus the utility-only path, not versus zero^[10].

California: the net billing template

California matters out of proportion to its size because it is the model other states copy — and the clearest illustration of how much net billing changes the picture.

Since April 15, 2023, new interconnection applicants in PG&E, SCE and SDG&E territory take service on the Net Billing Tariff under CPUC Decision 22-12-056^[1]. It is formally no longer net metering — the commission dropped the label deliberately, because exports are no longer credited at retail. The old NEM tariffs remain open only to grandfathered customers^[1].

Sources: CPUC Customer Generation program page^[1]; SurgePV California NEM 3.0 guide^[2]; pv magazine's March 2026 coverage describes the reduction as "approximately 75% to 80%"^[4].

The ACC is not a single number. It varies by hour, month and territory, and it is lowest exactly when solar produces most: on PG&E's approximate 2026 schedule, summer off-peak export values run roughly $0.04–$0.08/kWh while summer 4–9 PM peak values run $0.12–$0.16/kWh, with overnight super off-peak at $0.01–$0.03/kWh^[2].

Solar produces hardest between 10 AM and 3 PM, exactly when the ACC sits closest to its floor. The expensive hours arrive as the sun goes down. That mismatch is the economic problem net billing creates — and the reason batteries went from optional to standard equipment.

Grandfathering and the April 15, 2026 deadline

NEM 2.0 grandfathering was available to customers who filed a complete interconnection application before the April 2023 cutover, and it lasts 20 years from the Permission to Operate date; adding battery storage does not disturb it, and the status follows the property on sale^[2][15]. Grandfathered systems had to achieve Permission to Operate by April 15, 2026. Applications filed before the cutoff that had not gone live by then lose the protection and fall to the Net Billing Tariff^[15].

The legal challenge is over — net billing won

On March 9, 2026, the California First District Court of Appeal affirmed the CPUC's decision in full in Center for Biological Diversity v. Public Utilities Commission (Case No. A167721), deferring to the commission's assessment of the costs and benefits of distributed generation. Press coverage described the ruling as a blow to rooftop solar^[4][5]. The California Supreme Court's August 2025 opinion had addressed only how appellate courts apply the judicial review standards in Public Utilities Code sections 1757 and 1757.1; it did not reach the merits^[5].

What this means for you: the Net Billing Tariff is settled law for now, and the appellate decision was listed among the top five distributed solar policy actions of Q1 2026^[6]. Do not plan around a reversal. For the full state picture, see our California Solar Guide 2026.

State-by-state: export compensation in 2026

This is the reference table. It is a starting point, not a quote — the utility's tariff sheet always governs.

Primary sources: DSIRE for each state's governing rule^[18]; CPUC for California^[1][2]; Texas Utilities Code §39.916(j) and PUCT Rule §25.242 for Texas^[7][8]; Florida Administrative Code Rule 25-6.065 for Florida^[10]; SurgePV state guides for Arizona and New York^[11][12]; TheGreenWatt's April 2026 survey for the remaining states^[13]. Retail figures cross-referenced to EIA-derived averages for mid-2026^[17]. Programs change quarterly — verify before you sign.

The no-mandate states: Texas and Florida

Two of the three biggest solar states have no statewide net metering mandate at all. They are also the two most misunderstood.

Texas

The PUCT's net-metering rule applies only to price-to-beat retail providers and integrated investor-owned utilities that have not unbundled, and expressly does not apply to municipal utilities, river authorities or electric cooperatives. Its single-meter "running backward" option for renewable generators of 50 kW or less is closed to any contract executed after December 31, 2008^[7].

What governs instead is state statute. Under Texas Utilities Code §39.916(j), a distributed generation owner in a competitive area must sell surplus electricity to the retail electric provider serving their load, at a value agreed between the owner and that provider — which may be based on the market clearing price at delivery, or be a bill credit carried forward^[8].

In plain English: your export rate is a contract term you negotiate, not a right the state grants you. That is why outcomes vary so widely — Austin Energy publishes a Value of Solar rate near $0.10/kWh for systems up to 20 kW, while in Oncor and CenterPoint territory your buyback depends entirely on which REP you chose. Switching providers at contract renewal is a legitimate way to improve your solar economics^[9]. See our Texas Solar Guide 2026.

Florida

Florida is often called a full-retail net metering state. That is half right, and the wrong half matters. Florida Public Service Commission Rule 25-6.065 applies to investor-owned utilities — FPL, Duke Energy Florida, Tampa Electric and Florida Public Utilities — and provides monthly net metering: exports offset imports within the billing period at retail, and excess credits carry forward up to 12 months. Three caveats matter^[10]:

• Year-end true-up is not at retail. Unused credits pay out at the utility's as-available or avoided-cost rate, so design around realistic annual consumption rather than maximum surplus.
• Sizing limits exist. FPL requires estimated annual production below 115% of annual consumption for net-metering approval.
• Municipal and cooperative utilities are outside the rule. OUC grandfathered legacy full-retail accounts through June 30, 2045, but accounts interconnected after June 30, 2025 follow a transition: full-retail grace period through fall 2026, then a community solar energy rate through June 30, 2030, then a levelized fuel rate. SECO Energy credits excess at wholesale; KUA pays avoided cost to non-grandfathered exports.

So in FPL, Duke and Tampa Electric territory a battery is mostly a resilience purchase — monthly retail netting already handles energy shifting through the bill. In OUC, SECO or KUA territory the battery also has a financial job^[10].

Arizona and New York: two opposite lessons

These two states show how much the model, not the sunshine, determines the outcome.

Arizona: abundant sun, modest export value

Arizona has some of the best solar resource in the country — Phoenix sees roughly 6.0–6.5 peak sun hours daily — and no statewide net metering mandate^[11]. APS credits exports using the Resource Comparison Proxy, a rate that steps down over time, with monthly netting and a year-end true-up that pays excess at the RCP rather than forfeiting it. SRP sits outside Arizona Corporation Commission jurisdiction and left retail-rate net metering back in 2015; its residential solar rate plan includes a demand charge on peak 15-minute usage, which can materially reduce solar-only savings — one reason storage sized to shave peak demand performs well in SRP territory. TEP pays below retail on a time-of-use structure with a 3–8 PM weekday peak^[11].

Published 2026 figures differ between trackers: SurgePV places APS's RCP around $0.076–$0.10/kWh and SRP's export price around $0.06–$0.08/kWh, while TheGreenWatt's survey lists Arizona utilities at roughly $0.03–$0.05/kWh^[11][13]. That disagreement is itself the point — Arizona's rates step down annually and vary by utility, so the only reliable number is the one on your current tariff sheet.

New York: the strongest survivor

New York is the most important counterexample to the idea that net metering is universally dying. Residential systems up to about 25 kW generally still qualify for standard retail-rate net metering, and a parallel tariff, VDER (Value of Distributed Energy Resources), compensates exports through a "Value Stack" that prices energy, capacity, environmental attributes and locational congestion relief separately^[12].

VDER is not automatically worse. In constrained parts of New York City the demand reduction and locational relief components can push compensation above retail during peak periods, in the $0.20–$0.35/kWh range, while upstate it often falls below retail. New York also stacks incentives: NY-Sun pays upfront per-watt rebates of roughly $0.15–$0.40/W by territory and program block, and the state offers a 25% residential solar tax credit capped at $5,000 with a five-year carryforward^[12]. In a year when the federal credit is gone nationally, state-level support like this separates one market from another.

How batteries change the math

Under net metering, a battery is a backup-power purchase. Under net billing, it is also a financial instrument. Instead of exporting midday solar for a few cents, you store it and discharge in the evening, offsetting grid imports at the retail peak rate — earning the retail rate on stored energy instead of the avoided-cost export rate^[2][3]. The size of that opportunity equals the spread between what exports earn and what peak imports cost, and the wider the spread, the faster storage pays back^[3].

Export ranges and rate structures per the sources cited above; retail rate levels cross-referenced to EIA-derived state averages for mid-2026^[17]. Your utility's tariff sheet and rate plan determine the actual spread.

This reframes the battery question. It is not "do I want backup power?" It is "is my export rate low enough that storing my own solar beats selling it?" In California under the Net Billing Tariff the answer is emphatically yes. In a full-retail net metering state, usually no — the grid is already storing your energy for free, at full value. Our Best Solar Batteries 2026 guide covers equipment choices.

One caution: a battery adds substantial upfront cost, and the spread economics above are projections, not promises. Model it with your actual tariff and backup goals — and be skeptical of any proposal claiming both full-retail net metering and a battery that pays for itself through arbitrage. Those two claims are contradictory^[10].

What to check before you sign anything

Five things determine whether a savings projection is real, and all five are verifiable in writing.

1. Confirm your actual utility and tariff. Not the state — the utility. A PG&E customer and an SMUD customer face entirely different economics; in Florida, FPL territory and OUC territory change the battery case^[10][12].
2. Get the export rate in writing, from the tariff sheet. "Retail rate" is not an answer under a net billing regime^[3].
3. Ask for hourly or time-of-use modeling, not annual netting. A flat export assumption can substantially overstate savings under avoided-cost regimes^[2][3].
4. Check rollover and true-up mechanics. The same nominal rate pays very differently depending on whether credits roll monthly, true up annually, or expire^[3][10].
5. Model both solar-only and solar-plus-storage. Under net billing the honest comparison is solar versus solar plus storage, not solar versus nothing^[2].

One more note on financing: because Section 48E still supports commercial and third-party-owned projects, lease and PPA providers can pass through value a cash buyer cannot claim in 2026. We cover that comparison in our PPA vs Lease vs Buy guide, and the ownership question more broadly in Is Solar Worth It in 2026. For a first-pass estimate, start with our solar savings calculator — every assumption, including the export rate, is adjustable.

Frequently asked questions

What is the difference between net metering and net billing?

Net metering credits each exported kilowatt-hour at the full retail rate, so a credit earned at noon offsets a kilowatt-hour bought at 9 PM dollar for dollar. Net billing charges imports at retail but pays a separate, lower rate for exports — wholesale, avoided cost, or a regulator-set value, typically 20–50% of retail. In practice it turns design from "cover my annual usage" into "cover my daytime usage," and makes batteries far more valuable^[3].

Which states still offer full retail net metering in 2026?

Roughly 20, concentrated in the Northeast, Mid-Atlantic and Upper Midwest — New York, New Jersey, Massachusetts, Connecticut, Maryland, Illinois, Colorado, Oregon, Vermont, New Hampshire, Maine, Rhode Island, Pennsylvania, Ohio, Delaware, Iowa, Minnesota, Montana, West Virginia and Wisconsin. About nine states have moved to net billing or a reduced rate. Confirm your state in DSIRE^[13][18].

Did California's NEM 3.0 survive the lawsuit?

Yes. On March 9, 2026, the California First District Court of Appeal affirmed the CPUC's net billing decision in full in Center for Biological Diversity v. Public Utilities Commission (A167721). The California Supreme Court's August 2025 opinion had addressed the standard of judicial review, not the merits. The Net Billing Tariff remains in effect for all new PG&E, SCE and SDG&E interconnections^[4][5][6].

Am I grandfathered under California's NEM 2.0?

You are grandfathered if you filed a complete interconnection application before the April 2023 cutover and your system achieved Permission to Operate by April 15, 2026. Protection lasts 20 years from your PTO date, transfers with the property if you sell, and is not broken by adding a battery^[2][15].

What if my state has no net metering mandate?

Then your export rate may be a contract term rather than a statutory right. Under Texas Utilities Code §39.916(j), a competitive-market customer must sell surplus power to their chosen retail electric provider at a value agreed between the two. Florida's PSC rule covers only investor-owned utilities; municipal utilities and cooperatives set their own terms and some pay wholesale or avoided cost^[8][10].

What are non-bypassable charges and how much do they cost me?

They are the parts of your bill every customer pays regardless of usage, typically funding public-purpose programs such as energy efficiency and low-income assistance. Under a retail net metering tariff your export credit is the full retail rate minus the non-bypassable portion, usually a few cents per kilowatt-hour. They are also why solar can never reduce your bill to exactly zero^[10][14].

Do I need a battery if my export rate is low?

Usually yes, if you want a strong financial return — but the reasoning matters more than the answer. A low export rate does not hurt the solar you consume yourself; it hurts the surplus you send to the grid. A battery converts low-value exports into high-value evening self-consumption, worth the spread between the export rate and your peak retail rate. Under full-retail net metering that spread is essentially zero^[2][3].

Bottom line

Export compensation is not a footnote to a solar decision. In 2026, with the federal residential credit gone, it is close to the whole decision^[16].

• Net metering is a rolling credit at the full retail price. Net billing is selling back at wholesale. Buyback and feed-in tariffs are a fixed price set by contract or regulation^[3].
• The model matters more than the sunshine. Arizona has better sun than New York and weaker export economics. Texas has no mandate, so your REP contract decides your rate^[8][11][12].
• Check the mechanics, not just the headline rate. Rollover, true-up and credit expiry decide what surplus production is worth^[3][10].
• Self-consumption is the value you always keep. No export rule touches the kilowatt-hours your own house uses^[2].
• Get the number in writing from the tariff sheet. Any projection that cannot point to a current publication is a guess.

Policy here moves quarterly — 253 distributed solar policy actions in one quarter of 2026 was the norm, not the exception^[6]. Verify your state's rule at DSIRE before you sign^[18].