Original research
Demand response statistics: the flexible fleet nobody counts
Residential demand response enrolment across all 51 US jurisdictions, and the difference between a demand response programme and a virtual power plant.
Written by HyreSolar Research team Research and analysis
Audited by HyreSolar Research team Data audit and fact check
The headline
Two terms that are not interchangeable
Most of what is written about household flexibility slides between these two words inside a single paragraph. The slide is the reason the numbers below are surprising, so the definitions come before the data rather than after it.
- Demand response
- A programme run by a utility, in which a household agrees that its consumption can be reduced or shifted at the utility’s request in exchange for a bill credit or a lower rate. The federal definition, at 18 CFR § 35.28(b)(4), is “a reduction in the consumption of electric energy by customers from their expected consumption in response to an increase in the price of electric energy or to incentive payments designed to induce lower consumption of electric energy”. Note what it does not say: nothing about aggregation, nothing about markets, nothing about batteries, nothing about software.
- Virtual power plant
- An aggregation of distributed resources (batteries, thermostats, water heaters, electric vehicles, sometimes generators) coordinated so that the group behaves like a single dispatchable power station. The US Department of Energy’s Pathways to Commercial Liftoff: Virtual Power Plants describes VPPs in exactly those terms: aggregations of distributed energy resources that can balance supply and demand and provide grid services in the way a conventional plant does. The defining feature is the aggregation, not the customer relationship.
- Where they overlap
- A demand response programme whose enrolled load is bid into a wholesale market by an aggregator is both. FERC’s Order No. 2222, issued on 17 September 2020, is the rule that opened organised wholesale markets to those aggregations, and it explicitly lists demand response among the resources that may take part.
- Where they do not
- A cooperative that switches its members’ water heaters off for ninety minutes on a January morning is running demand response and is not running a virtual power plant: there is no aggregation, no market bid and no dispatchable capacity product. Equally, a battery sitting in a garage under nobody’s control is neither. The 10,658,027 households below are enrolled in demand response. It would be wrong to describe them as being in virtual power plants, and this page never does.
- What EIA actually counts
- Form EIA-861’s Demand Response schedule asks each utility for the number of customers enrolled in its residential demand response programmes, plus the energy savings those programmes delivered. It is an enrolment census, not a dispatch record. A household counted here has signed up; the schedule does not say how often it was called on.
The residential demand response fleet in six figures
Every figure here is computed from Form EIA-861, the mandatory annual census of American electric utilities. It is a census rather than a survey, so there is no sampling error in these counts, only the reporting quality of the utilities themselves.
The size of the thing, put next to the thing everyone writes about
Residential storage is the story of the decade in distributed energy, and it deserves to be. It is also, measured in households, small. 297,550 American homes have a photovoltaic system with a paired battery. Set that against 10,658,027 homes enrolled in a demand response programme and the ratio is 36 to one.
That ratio is not a reason to be unimpressed by batteries. A battery can export; a load-control switch cannot. A battery is available in a night-time outage; a water heater relay is not. The two resources do different jobs and a megawatt from one is not a megawatt from the other. But when the trade press describes the household flexibility fleet, it counts the 297,550 and not the 10,658,027, and the resulting picture of the American home as a grid resource is off by more than an order of magnitude.
The reason for the omission is not mysterious. Demand response is old. Direct load control of electric water heaters and central air conditioners was widespread in American utilities decades before anyone used the phrase “distributed energy resource”, and by the time the current wave of interest arrived it had already stopped being news. Nothing is being launched, so nothing gets covered. The switches are simply there, on 7.45% of American electricity accounts, being counted once a year in a federal workbook that almost nobody opens.
The Department of Energy’s Pathways to Commercial Liftoff work sets a national ambition of 80–160 GW of virtual power plant capacity by 2030. Whatever one makes of that target, it is worth noticing that the enrolment base it would have to be built from is not a greenfield. In 32 of the 51 jurisdictions, more households are already enrolled in demand response than have a rooftop solar system of any kind.
The state spread is the story
National enrolment of 7.45% is an average of jurisdictions that have almost nothing in common. Delaware enrols 56.1% of its residential accounts. Tennessee reports 0. The median state sits at 6.0% (Oklahoma, if you want a name for the middle) with the upper quartile at 7.7% and the lower at 2.0%.
10 jurisdictions enrol at least one household in ten. 27 enrol at least one in twenty. 9 sit below one in a hundred. That distribution has no relationship to sunshine, to electricity price, or to how much rooftop solar a state has, and it is not obviously about climate either: Minnesota and Maryland sit second and third, and they do not share a season.
What the leaders do share is a long institutional history of direct load control on a single dominant programme. Enrolment at these levels is not the result of a consumer marketing campaign; it is the result of a tariff in which a large share of a utility’s residential base is signed up as a matter of course, usually attached to an electric water heater or a whole-house air-conditioning cycling switch. Where that history exists the rate is enormous. Where it does not, the rate is near zero, and no amount of recent interest in flexibility has moved it.
The sixteen highest enrolment rates
Read this chart alongside the battery figures further down and the shape of the finding becomes visible. The three states at the top of it hold 2,648,267 enrolled households between them and 625 PV-paired home batteries: a ratio of roughly 4,237 to one.
Three states where the contrast is extreme
The three states below are not cherry-picked outliers on a noisy measure. They are ranks 1, 2 and 3 on enrolment, and each of them has a battery fleet small enough to count by hand.
Delaware, 56.1% enrolled, 3 batteries
265,511 of Delaware’s 473,508 residential electricity accounts are enrolled in demand response. That is rank 1 in the country and more than one household in two.
In the same state, in the same data year, utilities report 3 residential photovoltaic systems with a paired battery. The state has 14,087 rooftop solar systems in total, on 2.98% of households, so the battery count is not a rounding artefact of a missing solar fleet: the solar is there, the storage is not.
Expressed as a ratio the number is almost absurd: 88,504 demand-response households for every household battery. If you were sizing a residential flexibility programme in Delaware on the strength of the battery fleet, you would be looking at the wrong asset class by five orders of magnitude.
Minnesota, 48.0% enrolled, 88 batteries
1,239,273 households, rank 2, against 88 PV-paired batteries and 22,621 rooftop systems. Minnesota has the second-largest residential demand response fleet in absolute terms of any state in the country, behind nobody except in rate.
It is worth stating what this does and does not mean. Enrolment is not dispatch. A Minnesota household on a controlled water-heater tariff has agreed to an interruption; the schedule does not record how many hours of interruption actually happened. What the figure establishes is the size of the contractual base, which is the thing the virtual-power-plant conversation is usually trying to estimate and usually estimates from battery sales.
Maryland, 46.6% enrolled, 534 batteries
1,143,483 households enrolled, rank 3, and 534 paired batteries across 108,428 rooftop systems. Maryland also runs one of the larger utility efficiency programmes in the country, saving 446,715 MWh of residential electricity in 2024.
The three states together account for 2,648,267 enrolled households. Every home battery in all three, added up, comes to 625. Any national estimate of household grid flexibility built from storage data misses this entirely, and misses it hardest in exactly the states where the flexibility is largest.
Enrolment and batteries are unrelated measures
Plotted against one another across every state that reports both, demand response enrolment and household battery ownership show no useful relationship. The states with the most batteries per household (Hawaii, California, Utah) are ordinary or below average on enrolment. The states with the highest enrolment have almost no batteries.
This matters for anyone modelling residential flexibility. The two resources are not substitutes that a state chooses between, and they are not complements that arrive together. They are set by completely different things: battery attachment tracks export tariff design, and enrolment tracks whether a dominant utility ran a load-control programme in the 1980s. Neither is a proxy for the other, and a model that uses storage adoption to estimate a state’s flexible load will get Delaware, Minnesota and Maryland badly wrong.
In 44 of the 45 states that report both figures, more households are enrolled in demand response than own a solar battery. The single exception is Tennessee, which reports no residential demand response programme at all.
The energy figure is a trap, and here is how it springs
EIA reports that residential demand response programmes saved 247,973 MWh of electricity in 2024. Divided across 10,658,027 enrolled households that is 23.3 kWh each, per year, 0.22% of what an average American household uses. As a share of all residential electricity sold in the country it is 0.017%.
Read casually, those numbers say demand response is pointless. Read correctly, they say demand response is not an energy resource and should never have been measured as one. The purpose of a load-control programme is to remove megawatts at the moment the system peaks, not megawatthours over a year, and the value of a curtailment is set almost entirely by which hour it lands in. Summed across every filing utility, reported 2024 summer peaks come to 1,104,571 MW against 970,925 MW in winter: these are non-coincident peaks that each utility experiences on its own worst day, so the total is well above the simultaneous national peak and should be read as a seasonal ratio rather than a system figure.
So the honest statement of the finding is narrower than the headline invites. 10,658,027 households have contracted flexibility. The annual energy those contracts displace is trivial. Both facts are true and neither cancels the other, and a page that quoted only the first would be selling something.
EIA does publish potential and actual peak demand savings in megawatts on the same schedule. Those columns are not in our extraction and therefore appear nowhere on this page, we do not report figures we have not processed and validated ourselves.
Demand response against utility efficiency programmes
The same federal filing carries a second, much larger demand-side programme that gets even less attention, and the comparison is instructive because it runs the opposite way.
Demand response: many households, almost no energy
10,658,027 households enrolled, 7.45% of accounts.
247,973 MWh saved, 0.017% of residential sales.
23.3 kWh per enrolled household per year.
Value is in capacity: a megawatt withheld in the peak hour. The energy total is close to meaningless as a measure of worth.
Efficiency programmes: fewer headlines, real energy
10,200,084 MWh of residential electricity saved by utility energy-efficiency programmes in 2024.
That is 0.69% of all residential electricity sold, and 41 times the energy demand response saved.
71.3 kWh per US household per year, averaged across every account in the country rather than only participants.
Value is in energy: kilowatthours that were never sold. Nobody calls this a virtual power plant, and correctly so: it is not dispatchable and it cannot be called on.
Efficiency programmes are as unevenly distributed as enrolment
Utility energy-efficiency savings vary between states by a factor large enough that a national average tells you very little. Illinois saves 444 kWh per residential account per year through utility programmes; North Dakota saves 0.1. In absolute terms Illinois alone accounts for 2,394,336 MWh, more than any other state and 23.5% of the national total.
The dispersion has an obvious institutional cause (energy efficiency resource standards are state law, and a state either has one with teeth or does not) and it is one of the few measures in this dataset where the policy explanation is uncontroversial. We note it because efficiency savings and demand response enrolment are frequently reported together as “demand-side management”, and adding them produces a number that means nothing: one is energy, the other is capacity, and the units only look compatible.
Residential demand response enrolment, all 51 jurisdictions
| State | Rank | Households enrolled | Enrolment rate | Residential accounts | PV-paired batteries | Rooftop systems | Efficiency savings (MWh) |
|---|---|---|---|---|---|---|---|
| Delaware | 1 | 265,511 | 56.1% | 473,508 | 3 | 14,087 | 4,530 |
| Minnesota | 2 | 1,239,273 | 48.0% | 2,581,180 | 88 | 22,621 | 208,926 |
| Maryland | 3 | 1,143,483 | 46.6% | 2,451,753 | 534 | 108,428 | 446,715 |
| Washington | 4 | 602,575 | 17.9% | 3,368,978 | 2,579 | 59,235 | 208,686 |
| Michigan | 5 | 727,142 | 16.1% | 4,516,367 | 4,316 | 23,456 | 373,013 |
| North Carolina | 6 | 684,492 | 13.5% | 5,067,762 | 248 | 57,393 | 618,414 |
| North Dakota | 7 | 45,516 | 11.4% | 397,839 | 0 | 81 | 58 |
| Florida | 8 | 1,148,168 | 11.0% | 10,443,373 | 3,120 | 289,774 | 201,149 |
| Iowa | 9 | 151,656 | 10.5% | 1,448,818 | 22 | 16,392 | 66,152 |
| Oregon | 10 | 192,807 | 10.3% | 1,868,012 | 862 | 52,600 | 54,063 |
| Utah | 11 | 119,238 | 9.4% | 1,272,861 | 8,642 | 81,115 | 196,873 |
| Colorado | 12 | 227,456 | 8.9% | 2,557,732 | 1,298 | 179,751 | 107,578 |
| District of Columbia | 13 | 24,416 | 7.7% | 318,780 | — | 17,780 | 5,779 |
| Kentucky | 14 | 156,115 | 7.5% | 2,077,215 | 476 | 10,069 | 25,339 |
| Illinois | 15 | 398,254 | 7.4% | 5,392,477 | 28 | 103,145 | 2,394,336 |
| Indiana | 16 | 224,908 | 7.4% | 3,056,216 | 629 | 11,045 | 232,792 |
| South Dakota | 17 | 32,036 | 7.4% | 433,548 | 0 | 469 | 3,828 |
| New Mexico | 18 | 66,644 | 7.1% | 936,098 | 112 | 61,209 | 91,948 |
| Nebraska | 19 | 63,724 | 7.0% | 905,415 | 10 | 2,807 | 9,611 |
| Vermont | 20 | 22,781 | 7.0% | 323,960 | — | 9,193 | 20,227 |
| Hawaii | 21 | 29,939 | 6.7% | 446,512 | 17,953 | 98,418 | 23,231 |
| Nevada | 22 | 84,725 | 6.5% | 1,311,755 | 3,143 | 134,257 | 64,867 |
| South Carolina | 23 | 162,850 | 6.3% | 2,581,882 | 753 | 40,055 | 205,798 |
| Georgia | 24 | 286,553 | 6.0% | 4,815,501 | 395 | 18,612 | 140,701 |
| Oklahoma | 25 | 112,758 | 6.0% | 1,878,818 | 318 | 16,755 | 143,180 |
| Texas | 26 | 754,484 | 6.0% | 12,547,863 | 3,634 | 133,501 | 400,655 |
| Arizona | 27 | 179,914 | 5.7% | 3,134,733 | 10,345 | 317,033 | 676,939 |
| Kansas | 28 | 60,259 | 4.6% | 1,317,095 | 2 | 8,583 | 12,255 |
| California | 29 | 594,271 | 4.2% | 14,217,180 | 221,123 | 2,090,983 | 1,211,755 |
| Missouri | 30 | 110,883 | 3.8% | 2,933,914 | 760 | 30,022 | 85,897 |
| Virginia | 31 | 116,331 | 3.2% | 3,654,481 | 697 | 69,553 | 163,381 |
| Ohio | 32 | 123,237 | 2.4% | 5,142,108 | 1,647 | 27,520 | 17,152 |
| New York | 33 | 173,989 | 2.3% | 7,420,214 | 4,006 | 226,979 | 574,241 |
| Wisconsin | 34 | 61,174 | 2.2% | 2,843,874 | 287 | 18,581 | 60,473 |
| Massachusetts | 35 | 61,857 | 2.1% | 2,924,535 | 2,528 | 189,611 | 85,864 |
| Arkansas | 36 | 29,229 | 2.0% | 1,476,944 | 45 | 19,146 | 139,348 |
| Connecticut | 37 | 31,329 | 2.0% | 1,552,746 | 677 | 103,824 | 30,068 |
| Idaho | 38 | 17,451 | 2.0% | 868,141 | 465 | 22,899 | 47,034 |
| Alabama | 39 | 40,918 | 1.7% | 2,398,172 | — | 76 | 25,468 |
| New Hampshire | 40 | 8,204 | 1.3% | 652,608 | 718 | 22,503 | 19,389 |
| Mississippi | 41 | 11,362 | 0.8% | 1,340,459 | 11 | 1,789 | 50,286 |
| Pennsylvania | 42 | 42,843 | 0.8% | 5,534,580 | 1,298 | 81,375 | 307,357 |
| Alaska | 43 | 2,227 | 0.7% | 298,397 | — | 2,706 | — |
| Louisiana | 44 | 10,301 | 0.5% | 2,147,745 | 2 | 34,417 | 100,735 |
| Montana | 45 | 2,749 | 0.5% | 557,432 | 205 | 9,151 | 9,511 |
| New Jersey | 46 | 9,859 | 0.3% | 3,735,638 | 1,605 | 199,822 | 218,338 |
| West Virginia | 47 | 1,987 | 0.2% | 866,456 | 603 | 3,535 | 1,647 |
| Maine | 48 | 149 | 0.0% | 742,549 | 55 | 14,867 | 11,837 |
| Tennessee | 49 | 0 | 0.0% | 3,166,103 | 34 | 86 | 54,440 |
| Rhode Island | — | — | — | 457,383 | 1,155 | 17,321 | 36,856 |
| Wyoming | — | — | — | 286,475 | 119 | 2,913 | 11,363 |
Sorted by enrolment rate. HyreSolar analysis of EIA-861 2024: Demand Response, Net Metering, Energy Efficiency and Sales to Ultimate Customers schedules, joined on state.
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.
What the two demand-side programmes actually deliver
| Measure | Demand response | Utility efficiency programmes | For scale |
|---|---|---|---|
| Households involved | 10,658,027 enrolled | Not reported as an enrolment count | 143,144,185 US residential accounts |
| Residential energy saved | 247,973 MWh | 10,200,084 MWh | 1,482,830,612 MWh sold to households |
| As a share of residential sales | 0.017% | 0.69% | 100% |
| Per household | 23.3 kWh per enrolled home | 71.3 kWh per US home | 10,359 kWh used per home |
| What the resource is | Capacity (load withheld on request | Energy) consumption that never happened | , |
| Dispatchable | Yes, by the utility | No | — |
HyreSolar analysis of EIA-861 2024. Residential sales are reconstructed as 143,144,185 accounts × 10,359 kWh; the reconstruction reproduces EIA’s own published efficiency share of sales to 0.69% against a reported 0.69%, which is the check that the denominator is right.
The two columns must not be added together. One is measured in capacity that was available, the other in energy that was not consumed, and the shared unit on the second row is a coincidence of reporting rather than a common quantity.
Enrolment against storage in the ten highest-enrolment states
| State | Enrolment rate | Households enrolled | PV-paired batteries | Enrolled per battery | Battery attachment on solar |
|---|---|---|---|---|---|
| Delaware | 56.1% | 265,511 | 3 | 88,504 | 0.02% |
| Minnesota | 48.0% | 1,239,273 | 88 | 14,083 | 0.39% |
| Maryland | 46.6% | 1,143,483 | 534 | 2,141 | 0.49% |
| Washington | 17.9% | 602,575 | 2,579 | 234 | 4.35% |
| Michigan | 16.1% | 727,142 | 4,316 | 168 | 18.40% |
| North Carolina | 13.5% | 684,492 | 248 | 2,760 | 0.43% |
| North Dakota | 11.4% | 45,516 | 0 | — | 0.00% |
| Florida | 11.0% | 1,148,168 | 3,120 | 368 | 1.08% |
| Iowa | 10.5% | 151,656 | 22 | 6,893 | 0.13% |
| Oregon | 10.3% | 192,807 | 862 | 224 | 1.64% |
HyreSolar analysis of EIA-861 2024. “Enrolled per battery” is households in a demand response programme divided by households with a PV-paired battery, our derivation, not an EIA figure.
Battery attachment in the final column is the share of that state’s solar systems with a paired battery, which is a different denominator again. It is included to show that these states are not simply states without solar. Full storage analysis is on home solar battery statistics.
How to read a demand response statistic without being misled
Five errors we encountered repeatedly while assembling this page, three of which we made ourselves before checking.
- Do not describe enrolled households as being “in a virtual power plant”
Enrolment in a utility load-control programme is a bilateral arrangement. Whether it is aggregated into a market-facing resource is a separate question that this dataset does not answer for any of the 10,658,027 households.
- Do not read the energy savings as a measure of programme value
247,973 MWh sounds like failure and is not. Demand response is bought for the megawatts it removes in the peak hour; the annual energy total is close to irrelevant to its worth and is reported here only because it is what EIA collects.
- Do not add demand response savings to efficiency savings
Both appear in MWh on adjacent schedules, which makes the addition tempting. One is capacity availability expressed awkwardly in energy; the other is energy. The sum has no interpretation.
- Do check the denominator before comparing enrolment rates
Our rate is enrolled households ÷ residential electricity accounts. Measured instead against the 147,612,180 residential meters on the Advanced Meters schedule, national enrolment is 7.22% rather than 7.45%. Neither is wrong; they answer slightly different questions and the difference is real.
- Do treat a zero as a reporting fact, not a physical one
Tennessee reports 0 enrolled residential customers and Maine reports 149. Rhode Island and Wyoming report nothing at all. A zero in a federal filing means no filer reported a programme; it is not proof that no household anywhere in the state is on a controlled tariff.
Methodology
Source and extraction
US Energy Information Administration, Form EIA-861 for data year 2024, final release, downloaded as the published ZIP archives on 2 September 2026. Filing is mandatory for US electric utilities, so this is a census and carries no sampling error.
Enrolment and energy savings come from the Demand Response schedule, residential sector columns, summed by state across every filing utility. Battery and rooftop system counts come from the Net Metering schedule. Residential account counts come from Sales to Ultimate Customers, restricted to filing Parts A, B and D. Efficiency savings come from the Energy Efficiency schedule, residential sector. Meter counts come from Advanced Meters.
Every column is located by matching the schedule’s multi-row header rather than by fixed position, because EIA moves columns between years. The extraction is scripted and re-runnable; no figure on this page was transcribed by hand.
Definitions used on this page
Enrolled household: one residential customer that a utility reported as enrolled in a demand response programme on the 2024 Demand Response schedule. Utilities file one row each, so summing does not double-count a household across programmes within a utility.
Enrolment rate, enrolled residential customers ÷ residential electricity accounts, same state, same year, same source form. Reported to one decimal place.
Energy savings: the MWh a utility attributes to its residential demand response programmes for the year, as filed. Attribution methods are the utility’s own and EIA does not standardise them.
Efficiency savings, incremental annual MWh savings attributed to utility-run residential energy-efficiency programmes, from the Energy Efficiency schedule.
PV-paired battery: a residential battery reported as paired with a net-metered photovoltaic system. Standalone home batteries are filed separately and are not in any figure here.
How the derived figures are built
The battery ratio. 10,658,027 ÷ 297,550 = 36. Two national counts from two schedules of the same form, same data year.
Residential sales. EIA does not publish a single national residential MWh figure in a form convenient to this page, so it is reconstructed as 143,144,185 accounts × 10,359 kWh = 1,482,830,612 MWh. The check is that 10,200,084 MWh of efficiency savings against that denominator gives 0.69%, against the 0.69% computed directly from the underlying sales column. Agreement to that tolerance is what licenses the 0.017% figure built on the same denominator.
Per-household energy. 247,973 MWh × 1,000 ÷ 10,658,027 enrolled = 23.3 kWh. Efficiency is divided by all 143,144,185 accounts rather than by participants, because the schedule reports no participant count.
Ranks. Precomputed once in the dataset across the 49 jurisdictions reporting a rate, so no section of this page can disagree with another about which state is third.
Limitations
Enrolment is not dispatch. Nothing here says how often a programme was called, for how long, or whether the household opted out on the day. A large enrolment base is a contractual fact, not a demonstrated capacity.
Peak savings are not published here. The Demand Response schedule carries potential and actual peak demand savings in megawatts. Those columns are outside our extraction, so this page reports enrolment and energy only. Any megawatt figure you see attributed to us is not ours.
Programme composition is not in the data. EIA’s schedule carries a grid-connected water heater flag but does not break enrolment down by device or programme design, so we cannot say what share of the 10,658,027 is water heating, air conditioning, thermostats or time-varying rates. The characterisation of the fleet as predominantly water-heater and air-conditioner control is drawn from the long-standing structure of US direct load control programmes, not from a column in this file, and we flag it as such.
Reporting quality varies. Rhode Island and Wyoming report no residential figure and are excluded from every rate, average and rank rather than being treated as zero.
Accounts are not homes. A master-metered apartment building is one residential account covering many households, so every rate on this page is slightly overstated where that stock is large.
One year only. These are 2024 figures. We have not built a demand response time series and make no claim about the direction of travel.
Corrections
Every number on this page is generated from the source workbooks by script rather than typed into the prose, which removes the most common class of error but not all of them. If you find a figure you believe is wrong, tell us what it is and where you think the extraction went astray, and we will correct it with a dated note on the page.
Questions
How many US households are in demand response programmes?
Is demand response the same thing as a virtual power plant?
Which state has the highest demand response enrolment?
How much electricity does demand response actually save?
Why are there more demand response households than solar batteries?
Does high demand response enrolment mean a state is good at distributed energy?
Where does this data come from and how current is it?
Can I cite 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.
- US EIA, Form EIA-861, Demand Response, Energy Efficiency and Advanced Meters schedules (2024) — Residential customers enrolled, residential energy savings, utility efficiency programme savings and residential meter counts by utility and state. Retrieved 2 September 2026.
- FERC, 18 CFR § 35.28(b)(4), definition of demand response — The federal regulatory definition quoted on this page. Retrieved 2 September 2026.
- FERC Order No. 2222 explainer, participation of distributed energy resource aggregations in RTO/ISO markets — Issued 17 September 2020. The rule that opened organised wholesale markets to aggregations including demand response. Retrieved 2 September 2026.
- US DOE, Pathways to Commercial Liftoff: Virtual Power Plants — The definition of a virtual power plant used on this page, and the 80–160 GW by 2030 deployment ambition cited in the text. Retrieved 2 September 2026.
Solar, storage, or a tariff change?
Demand response is something your utility offers you. Solar and storage are decisions about your own roof, and the arithmetic depends on your rate.
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.