Original research
Net metering statistics: who is on the tariff, and what sits outside it
Enrolment across 11 years of federal utility filings, the three states that closed the tariff to new customers, and the 2,031 MW nobody counts.
Written by HyreSolar Research team Research and analysis
Audited by HyreSolar Research team Data audit and fact check
The finding
The tariff, in six figures
Net metering is the arrangement under which a household exports surplus generation to the grid and is credited for it. Almost every rooftop solar statistic in circulation is really a statistic about the households on this one tariff. Here is its size.
EIA counts installations behind a utility meter, not signed tariff agreements. One household with one system is one enrolment; a household that added a second array is not counted twice. Figures labelled HyreSolar analysis are computed by us from the EIA source files named below. EIA publishes the inputs; it does not publish these ratios or rankings.
Three arrangements that all get called net metering
Retail net energy metering
The original bargain, and the one the phrase properly describes. Exported kilowatt hours are credited against imported kilowatt hours at the retail rate, so a unit sent to the grid in the afternoon cancels a unit drawn back at night. The meter runs both ways and the bill is settled on the net.
This is what EIA's Net Metering schedule collects, and it is the population every figure in the first half of this page describes.
Net billing
Exports are bought at a separate, lower export rate while imports are still charged at retail. The two flows are priced independently rather than netted, which breaks the one-for-one exchange and moves the economics decisively towards using generation on site, which is to say, towards a battery.
Utilities file net billing customers on the same schedule, so a state that has moved to net billing does not vanish from the count. That is why California still reports 2,090,983 enrolments after moving off retail netting.
Everything else, filed separately
Buy-all-sell-all arrangements, utility-run distributed generation programmes, self-supply tariffs with no export at all, and interconnections outside any retail credit scheme go on the Non-Net-Metered Distributed Generation schedule instead. That schedule collects megawatts and no installation count.
The distinction is not academic. It is the difference between a state that appears to have almost no rooftop solar and a state that has rooftop solar filed somewhere else, and it accounts for 2,031 MW nationally.
Enrolment across 11 years
Net metering enrolment rose from 642,276 US households in 2014 to 5,077,543 in 2024, a multiple of 7.9. Over the same period the share of residential electricity accounts enrolled went from 0.50% to 3.55%, and the capacity behind those meters from 3,453 MW to 34,226 MW.
The cumulative line is uninteresting because it only ever goes up. The flow underneath it is where the tariff's history shows. New enrolments peaked at 761,174 in 2023 and fell to 527,942 in 2024: a drop of 30.6%, the steepest in the series, of which 49% is one state.
There were only two earlier falls, in 2017 and 2018, of 15.8% and 6.4%. 2024 is larger than both together. It is also worth stating what the fall is not: 527,942 new enrolments still exceeds every year in the series before 2022. The rate of new sign-ups fell hard; the programme did not shrink.
The three vertical markers on the chart are the years in which a state closed retail net metering to new residential customers. Two of the three sit on visible bends.
The three states that closed net metering to new customers
The closure dates below are regulatory dates, not statistics: the only figures on this page we did not compute ourselves. Each is sourced below. Existing customers were grandfathered in all three cases; what closed was the door for new applicants.
-
12 October 2015
Hawaii: the first state to end retail net meteringThe Hawaii Public Utilities Commission closed the Hawaiian Electric companies' net energy metering programme to new participants and approved Customer Self-Supply and Customer Grid-Supply as the replacement, one exporting at an avoided-cost credit and one not exporting at all (Hawaii PUC). Hawaii still has the highest household solar penetration in the country at 22.04%, and, on the 2024 filing, a battery attachment rate of 18.2%.
-
15 November 2017
Utah, Rocky Mountain Power Schedule 135 closesSchedule 135, Net Metering Service, closed to new applicants under a settlement approved by the Utah Public Service Commission; customers with an interconnection application on file before that date were grandfathered to 1 January 2036. New applicants moved first to the Schedule 136 Transition Program, itself closed on 31 October 2020, and now take Schedule 137 Net Billing Service (DSIRE).
-
15 April 2023
California: the Net Billing Tariff takes effectCPUC Decision 22-12-056, adopted 15 December 2022, replaced retail net energy metering with the Net Billing Tariff for the three large investor-owned utilities. Applications completed before the effective date kept the old terms; everything after took export rates far below retail (CPUC). Municipal utilities set their own rules and are not covered by the decision. California accounts for 49% of the national fall in new enrolments the following year.
Between them these three jurisdictions hold 45% of all US residential net metering enrolments. Retail net metering is closed to new applicants across that share of the installed base.
Three closures, three different aftermaths
A closure is a natural experiment, and the filings let you measure it, with one large caveat, which is that EIA reports data years and tariffs change on calendar dates, so the closing year always contains some months of both regimes.
Hawaii fell hardest and stayed down. New enrolments were 10,966 in 2015 and 7,916 in 2016, a fall of 28%. In 2018 the state posted a net loss of 922 enrolments: the cumulative count reported by its utilities actually fell. 2024 additions were 4,776, still below the 2015 figure a decade on.
Utah fell and then recovered most of the way. Enrolments went from 11,170 in 2017 to 6,279 in 2018, down 44%, and were back to 8,025 by 2024, below the pre-closure peak, but not obviously a market in retreat.
California is the largest case and the freshest. Enrolments fell from 278,068 in 2023 to 164,524 in 2024, down 41%, and 2024 is the only full data year under the new tariff so far. A single post-closure year is not a trend, and anyone telling you where California settles is guessing.
What the three share is direction, not magnitude. Ending retail netting reduced new enrolment in every case; how much and for how long varied by a factor of several, and the successor tariff's terms (whether exports are paid at all, and whether storage is credited) look far more important than the closure itself.
Closure aftermath, measured
| State | Closed to new customers | Enrolments, closing year | Enrolments, year after | Change | Enrolments, 2024 | Enrolled now |
|---|---|---|---|---|---|---|
| Hawaii | 12 October 2015 | 10,966 | 7,916 | -28% | 4,776 | 98,418 |
| Utah | 15 November 2017 | 11,170 | 6,279 | -44% | 8,025 | 81,115 |
| California | 15 April 2023 | 278,068 | 164,524 | -41% | 164,524 | 2,090,983 |
HyreSolar analysis of EIA-861 for every column except the closure date, which is the regulatory record cited in the timeline above.
"Enrolled now" is the cumulative count at the end of 2024, which keeps rising in all three states because grandfathered customers stay on the tariff. A closed programme and a shrinking programme are different things.
The twelve largest net metering populations
Where the enrolments are, and how fast each state got there. The multiple is 2024 enrolment divided by 2014 enrolment in the same state.
| State | Enrolled 2014 | Enrolled 2024 | Multiple | Added in 2024 | Share of US enrolments |
|---|---|---|---|---|---|
| California | 324,529 | 2,090,983 | 6.4× | 164,524 | 41.2% |
| Arizona | 51,282 | 317,033 | 6.2× | 21,892 | 6.2% |
| Florida | 6,778 | 289,774 | 42.8× | 39,146 | 5.7% |
| New York | 25,637 | 226,979 | 8.9× | 23,177 | 4.5% |
| New Jersey | 28,473 | 199,822 | 7.0× | 17,401 | 3.9% |
| Massachusetts | 19,246 | 189,611 | 9.9× | 30,323 | 3.7% |
| Colorado | 25,197 | 179,751 | 7.1× | 19,057 | 3.5% |
| Nevada | 4,198 | 134,257 | 32.0× | 19,182 | 2.6% |
| Texas | 9,993 | 133,501 | 13.4× | 11,317 | 2.6% |
| Maryland | 10,675 | 108,428 | 10.2× | 8,361 | 2.1% |
| Connecticut | 7,195 | 103,824 | 14.4× | 13,142 | 2.0% |
| Illinois | 749 | 103,145 | 137.7× | 30,401 | 2.0% |
HyreSolar analysis of EIA-861 annual files, 2014 and 2024. Sorted by 2024 enrolment.
The full 51-jurisdiction comparison, on eight measures rather than one, is on solar statistics by state.
The capacity that is not on the tariff at all
Form EIA-861 has a second distributed generation schedule, and it is the reason several state solar figures in wide circulation are simply wrong.
What the net metering schedule holds
5,077,543 residential installations and 34,226 MW. Both an installation count and a capacity, which is what makes household penetration computable at all.
This is the schedule every rooftop solar statistic you have read is built on, usually without saying so.
What the other schedule holds
2,031 MW of residential photovoltaic capacity, reported by 40 states. 5.6% of all reported residential capacity in the country.
Capacity only. No installation count, no household count, no way to convert it into systems, which is why the gap can be measured but not closed.
Texas: the single largest measurement gap in US solar data
Texas reports 984 MW of residential photovoltaic capacity inside net metering and 1,748 MW outside it. 64% of Texas's residential solar fleet is therefore invisible to the dataset every state solar ranking uses, including our own household penetration figure of 1.06%.
The cause is the structure of the Texas retail market. There is no single statewide net metering mandate; competitive retailers offer their own buyback products, and much of the state's rooftop solar interconnects under arrangements that are not net metering as the form defines it. Those megawatts land on the second schedule.
Texas alone accounts for 86% of the entire national non-net-metered residential total. If a state solar table shows Texas looking small, the net metering schedule is almost always the reason, and the correct response is to say which schedule you used, not to patch the number.
Tennessee, and the states that look empty and are not
Tennessee has 86 residential net metering enrolments across 3,166,103 residential electricity accounts, 0.0027%, last of 51 jurisdictions, and one of the most misread numbers in American energy data.
The second schedule shows why. Tennessee reports 18.0 MW of residential photovoltaic capacity outside net metering against 0.6 MW inside it: 96.8% of the state's residential fleet is filed off the schedule everyone reads. The Tennessee Valley Authority runs distributed generation through its own programmes rather than a state net metering tariff, and its distributors file accordingly.
Alabama, also largely TVA-served, shows the same pattern at 80.8% outside. In 3 states the majority of reported residential photovoltaic capacity sits outside net metering: Tennessee (96.8%), Alabama (80.8%), Texas (64.0%).
Writing that Tennessee "has almost no rooftop solar" is wrong. Writing that it has almost no net-metered rooftop solar is right, and it is the more interesting sentence, because it says something about tariff design rather than about roofs.
The 7 states where at least a tenth of residential photovoltaic capacity sits outside net metering. Below that threshold the omission is real but rarely changes a ranking position.
Where the gap is, geographically
Mapped, the gap is concentrated rather than diffuse. Most states that file the schedule at all report a small residual, California reports 11.5 MW against 12,702 MW net-metered, 0.1% of its fleet. A handful report enough to change what their solar statistics mean, and 11 jurisdictions file nothing on it at all.
A blank tile is not a zero. It is a state that did not report on this schedule, which is a different fact, and treating the two as the same is how a coverage gap becomes a false precision.
Residential photovoltaic capacity outside net metering, all 40 reporting states
| State | Outside net metering (MW) | Inside net metering (MW) | Total reported (MW) | Share outside |
|---|---|---|---|---|
| Texas | 1,748.2 | 984.0 | 2,732.2 | 64.0% |
| Hawaii | 90.3 | 531.7 | 622.0 | 14.5% |
| Massachusetts | 45.7 | 1,304.6 | 1,350.3 | 3.4% |
| Rhode Island | 43.2 | 106.0 | 149.2 | 29.0% |
| Tennessee | 18.0 | 0.6 | 18.6 | 96.8% |
| California | 11.5 | 12,701.6 | 12,713.1 | 0.1% |
| Iowa | 11.3 | 136.5 | 147.8 | 7.6% |
| North Carolina | 11.0 | 424.4 | 435.4 | 2.5% |
| Indiana | 9.2 | 96.0 | 105.2 | 8.8% |
| Oregon | 8.0 | 333.5 | 341.5 | 2.4% |
| Illinois | 6.9 | 798.2 | 805.1 | 0.9% |
| Connecticut | 5.1 | 800.6 | 805.7 | 0.6% |
| Alabama | 3.7 | 0.9 | 4.6 | 80.8% |
| Georgia | 2.5 | 128.8 | 131.3 | 1.9% |
| Colorado | 2.5 | 927.5 | 930.0 | 0.3% |
| Kentucky | 2.4 | 77.7 | 80.1 | 2.9% |
| Michigan | 2.2 | 155.4 | 157.6 | 1.4% |
| Mississippi | 2.1 | 9.8 | 11.9 | 17.9% |
| Wisconsin | 1.5 | 134.6 | 136.1 | 1.1% |
| Florida | 1.0 | 2,572.1 | 2,573.1 | 0.0% |
| Nebraska | 0.7 | 22.5 | 23.2 | 3.1% |
| Washington | 0.6 | 428.0 | 428.6 | 0.1% |
| South Carolina | 0.5 | 302.0 | 302.5 | 0.2% |
| Arizona | 0.4 | 2,277.7 | 2,278.1 | 0.0% |
| New Jersey | 0.4 | 1,380.7 | 1,381.1 | 0.0% |
| Louisiana | 0.3 | 187.8 | 188.1 | 0.2% |
| Minnesota | 0.3 | 190.0 | 190.3 | 0.1% |
| South Dakota | 0.3 | 3.6 | 3.9 | 6.5% |
| Utah | 0.2 | 481.0 | 481.2 | 0.1% |
| Virginia | 0.2 | 513.8 | 514.0 | 0.0% |
| Kansas | 0.2 | 66.1 | 66.3 | 0.2% |
| Ohio | 0.1 | 204.6 | 204.7 | 0.1% |
| North Dakota | 0.1 | 1.0 | 1.1 | 10.4% |
| Montana | 0.1 | 67.0 | 67.1 | 0.1% |
| New York | 0.1 | 1,572.5 | 1,572.6 | 0.0% |
| Missouri | 0.1 | 274.8 | 274.9 | 0.0% |
| Alaska | 0.0 | 14.2 | 14.2 | 0.3% |
| Nevada | 0.0 | 1,020.8 | 1,020.8 | 0.0% |
| Oklahoma | 0.0 | 123.6 | 123.6 | 0.0% |
| West Virginia | 0.0 | 31.5 | 31.5 | 0.0% |
HyreSolar analysis of EIA-861 2024, Net Metering and Non-Net-Metered Distributed Generation schedules, residential sector. Sorted by capacity outside net metering.
The 11 jurisdictions absent from this table filed nothing on the second schedule. That is not evidence that they have no non-net-metered residential solar; it is absence of a filing, and it is shown as absence.
Methodology
What is counted as an enrolment
One residential net-metered photovoltaic installation, as reported by the utility whose meter it sits behind, on the Net Metering schedule of Form EIA-861 for the 2024 data year. Leased systems and owned systems each count once; a household is not counted twice for adding capacity to an existing array.
We use "enrolment" rather than "customer" deliberately. EIA collects installations, and the mapping from installations to billing accounts is not one-to-one in every filing. Nothing on this page should be read as a count of signed tariff agreements.
The denominator for any share is residential electricity accounts from Sales to Ultimate Customers, same year, same filers, summed across Parts A, B and D so that restructured markets are not double-counted.
How annual enrolments are derived
Form EIA-861 is a stock survey. It asks a utility how many net-metered customers it has at year end, never how many it connected during the year. Every annual enrolment figure on this page (national and state) is the difference between two consecutive cumulative counts.
That arithmetic is exact but not innocent. If a utility reclassifies customers between schedules, fails to file, or merges with another, the difference registers as an enrolment or a loss that no installer ever performed. Hawaii's negative year two years after its closure is the clearest example in the dataset, and we report it as what it is rather than smoothing it away.
The three regulatory dates
The closure dates are the only figures on this page not computed from EIA files, and each is linked to its primary or authoritative secondary source in the timeline: the Hawaii Public Utilities Commission's own release, the DSIRE record of the Rocky Mountain Power schedules, and the California Public Utilities Commission's net energy metering and net billing page.
Every consequence we draw from those dates (enrolments before, after, and now) is computed from EIA-861 and is our analysis, not the regulator's. The regulators published dates; they did not publish these comparisons.
The second schedule, and why the gap cannot be closed
Non-net-metered distributed generation is collected as capacity in megawatts with no installation count attached. There is therefore no arithmetic that converts 2,031 MW into households, and no honest corrected penetration figure to publish.
We considered dividing that capacity by an assumed average system size to impute installations, and rejected it: average size varies by more than a factor of two across states in this same dataset, so the imputation would be dominated by the assumption. Instead the boundary is stated everywhere, every enrolment and penetration figure in this library is explicitly net-metered.
The two schedules also cannot be summed into a fleet total that means anything more than "capacity reported under either arrangement", because capacity is filed on AC and DC bases interchangeably and EIA does not normalise.
Limitations
A tariff change inside a data year is smeared across it. All three closures took effect part-way through a calendar year, so the "closing year" column mixes months of both regimes. Comparisons of the closing year with the following year understate the immediate effect and are not a regression discontinuity.
One post-closure year is not a trend. California has exactly one full data year under net billing. We report the change and decline to extrapolate it.
Closures we may not have caught. This page dates three state-level closures of retail net metering to new residential customers. Individual utilities elsewhere have hit programme caps or moved to net billing without a statewide decision, and EIA-861 does not carry a tariff-type flag that would let us detect that from the data.
Nothing here is compensation. The dataset records who is enrolled, not what they are paid. EIA's energy-sold-back column covers only a minority of installations and reports implausible per-system values for several large states, so we exclude it entirely rather than publish an export-value figure we cannot defend.
Accounts, not households. A master-metered apartment building is one residential account covering many homes, so enrolment shares are somewhat overstated where that housing stock is large.
Reproducing and correcting this
The source is public: EIA's Form EIA-861 annual archives, retrieved 2 September 2026. Every computed figure above is generated from those workbooks by script and inserted into this page as a variable, which is why the prose, the tables and the charts cannot drift apart.
If a regulatory date here is wrong, or a fourth state has closed its tariff and we have missed it, tell us. We will correct the page at this URL with a dated note.
Questions
How many people are on net metering in the US?
Which states have ended net metering?
Is net metering going away everywhere?
What happens to my net metering if my state changes the rules?
Why is Texas so low in net metering statistics?
What is the difference between net metering and net billing?
How much solar is not covered by net metering statistics?
How current are these figures?
Written and audited by
HyreSolar Research
Primary-source research, data analysis and fact checking
We are a research desk, not a sales floor. We read the statute, the tariff, the code section, the federal filing or the manufacturer data sheet ourselves, and we publish the figure with the document it came from and the date we retrieved it. Where a number cannot be traced to a primary source, we publish the shorter page and say what we could not verify. That rule has cost us whole sections, and it is the reason the rest can be trusted.
- 160
- primary sources read and cited
- 220
- figures with a retrieval date
- 115
- federal and state government sources
- 66
- researched pages published
How this desk works
- Primary sources only. Statutes from the legislature’s own publishing system, federal data from the agency that collects it, code text from the adopted edition, manufacturer claims from the data sheet. We do not cite an article that cites a source; we go and read the source.
- Every figure carries its provenance. A named document and the date we retrieved it, so you can check it and so you know how old it is. Retrieval dates are not decoration: an EIA rate from May is a different fact from an EIA rate from August.
- We publish what we could not verify. Every research page carries a section naming the things we tried to establish and could not, and why. A paywalled standard, a state website that refused the request, a manufacturer that publishes no figure at all.
- We separate measurement from modelling from our own reasoning, and label which is which on the page. A laboratory measurement, an assumption inside a modelling tool and our own inference are three different kinds of claim and they are never presented as one.
- We do not sell solar, and we take no payment for placement, ranking or a favourable mention. Nobody buys a position on this site.
Data as of EIA-861 2024 final release. Authorship on this site is organisational: the analysis belongs to the desk rather than to a named individual, and we do not publish credentials we do not hold. Our editorial policy sets out how we source, date and correct what we publish.
Sources & retrieval dates
- US EIA, Form EIA-861 — Net Metering (annual files, 2014–2024) — Utility-level net-metered installations, capacity and PV-paired batteries by state and customer sector. Retrieved 2 September 2026.
- US EIA, Form EIA-861 — Sales to Ultimate Customers (annual files, 2014–2024) — Residential revenue, sales and customer counts by utility and state. The basis for price, bill and consumption figures. Retrieved 2 September 2026.
- US EIA, Average Price of Electricity to Ultimate Customers by End-Use Sector — EIA’s published price series, used to validate our derivation. Agreement across all 357 overlapping state-years is within 0.005¢/kWh. Retrieved 2 September 2026.
- Hawaii Public Utilities Commission, PUC Reforms Energy Programs to Support Future Sustainable Growth in Hawaii Rooftop Solar Market — Decision and Order No. 33258, 12 October 2015: closed the Hawaiian Electric net energy metering programme to new participants and approved the Customer Self-Supply and Customer Grid-Supply tariffs. Retrieved 2 September 2026.
- DSIRE, Rocky Mountain Power, Net Billing (Utah) — Record of Schedule 135 closing to new applicants on 15 November 2017, the Schedule 136 Transition Program closing on 31 October 2020, the Schedule 137 Net Billing Service that succeeded them, and the grandfathering periods. Retrieved 2 September 2026.
- California Public Utilities Commission, Net Energy Metering and Net Billing — Decision 22-12-056, adopted 15 December 2022, effective for new applications from 15 April 2023: replaced retail net energy metering with the Net Billing Tariff for the three large investor-owned utilities. Retrieved 2 September 2026.
Net metering or net billing where you are?
The tariff decides whether a battery pays for itself. Work the arithmetic for a specific export rate rather than a national average.
HyreSolar is an independent analysis and matching service. We are not an installer, lender or utility. When a reader asks to be introduced, installers may pay us a referral fee. That fee never buys ranking, scores or placement in research. Our editorial policy sets out the rules.