Privacy choices

Choose what we may use.

The site works without optional tracking. Your choice is saved on this device and can be changed here anytime.

Necessary

Runs the site, protects forms, and remembers this privacy choice.

Always on

Energy statistics

Number of Solar Farms by State: Solar and Storage Statistics 2026

State counts for operating solar plants and battery storage, with generators, capacity and planned additions kept separate. Download the calculations behind the tables.

The short answer

California has the most operating solar plants, 983, while Texas has the most solar nameplate capacity, 34,977.4 MW AC, in PorchRocket's analysis of September 2026 EIA-860M data. Nationally, the analysis counts 7,659 solar plants and 1,071 battery plants. These EIA plant counts include qualifying rooftop systems; the separate USPVDB ground-mounted counts narrow the solar-farm definition. Asset Search supplies reviewed site inventories for a buyer's territory.

Find solar sites

What matters most

  • PorchRocket counts 7,659 operating solar plants and 171,966.0 MW AC in September 2026 EIA-860M data.
  • Operating battery storage totals 1,071 plants and 55,574.5 MW in the same September 2026 analysis.
  • USPVDB v4.0 contains 5,976 projects classified exclusively as ground-mounted systems in PorchRocket's calculation.
  • Solar and batteries share 477 operating EIA plant IDs, so adding the two plant totals would count those IDs twice.
  • The September 2026 planned inventory contains 122,225.2 MW AC of solar and 64,234.6 MW of batteries across all reported planned dates.

Updated November 2026. California leads the operating solar-plant count, while Texas leads solar capacity: 983 plants in California versus 250 in Texas, with Texas reporting 34,977.4 MW AC. Those are different ways to size a market. A service crew needs locations; an equipment supplier may care more about installed capacity. The figures come from PorchRocket’s calculation of the September 2026 EIA-860M inventory.

This reference separates EIA plants and generators from mapped ground-mounted projects. It also reports battery storage and planned additions. Asset Search by PorchRocket uses a buyer’s territory and acceptance rules to produce reviewed site inventories. The downloadable state dataset supports the statistics here; the tables cover the 50 states and District of Columbia, excluding territories.

Key statistics

PorchRocket’s calculations distinguish operating EIA plants from mapped USPVDB projects, with September 2026 inventory figures and an April 2026 ground-mounted snapshot.

  • PorchRocket’s analysis of September 2026 EIA-860M counts 7,659 operating U.S. solar plants containing 8,346 solar generators.
  • PorchRocket’s analysis of September 2026 EIA-860M totals 171,966.0 MW AC of operating U.S. solar nameplate capacity.
  • PorchRocket’s analysis of September 2026 EIA-860M counts 1,071 operating U.S. battery plants containing 1,125 battery generators.
  • PorchRocket’s analysis of September 2026 EIA-860M totals 55,574.5 MW of operating U.S. battery nameplate power.
  • PorchRocket’s analysis of September 2026 EIA-860M totals 152,869.8 MWh of reported operating U.S. battery energy capacity.
  • PorchRocket’s analysis of September 2026 EIA-860M places California first in operating solar-plant count, at 983 plants.
  • PorchRocket’s analysis of September 2026 EIA-860M places Texas first in operating solar nameplate capacity, at 34,977.4 MW AC.
  • PorchRocket’s analysis of September 2026 EIA-860M places California first in operating battery-plant count, at 277 plants.
  • PorchRocket’s analysis of September 2026 EIA-860M places Texas first in operating battery nameplate power, at 19,194.6 MW.
  • PorchRocket’s analysis of September 2026 EIA-860M counts 983 operating solar plants and 25,609.3 MW AC in California.
  • PorchRocket’s analysis of September 2026 EIA-860M counts 769 operating solar plants and 7,076.1 MW AC in North Carolina.
  • PorchRocket’s analysis of September 2026 EIA-860M counts 742 operating solar plants and 3,209.1 MW AC in New York.
  • PorchRocket’s analysis of September 2026 EIA-860M counts 547 operating solar plants and 2,175.4 MW AC in Minnesota.
  • PorchRocket’s analysis of September 2026 EIA-860M counts 261 operating battery plants and 19,194.6 MW in Texas.
  • PorchRocket’s analysis of September 2026 EIA-860M counts 277 operating battery plants and 16,267.1 MW in California.
  • PorchRocket’s analysis of September 2026 EIA-860M counts 56 operating battery plants and 6,877.8 MW in Arizona.
  • PorchRocket’s analysis of September 2026 EIA-860M finds 477 U.S. plant IDs with both operating solar and operating batteries.
  • PorchRocket’s analysis of September 2026 EIA-860M identifies 1,101 planned U.S. solar generators totaling 122,225.2 MW AC across all reported planned dates.
  • PorchRocket’s analysis of September 2026 EIA-860M identifies 483 planned U.S. battery generators totaling 64,234.6 MW across all reported planned dates.
  • PorchRocket’s analysis of final 2025 EIA-860 counts 7,317 operating U.S. solar plants with 154,251.1 MW AC at year-end 2025.
  • PorchRocket’s analysis of final 2025 EIA-860 counts 933 operating U.S. battery plants with 43,134.6 MW at year-end 2025.
  • PorchRocket’s analysis of April 2026 USPVDB v4.0 counts 6,611 mapped U.S. photovoltaic projects across all system types.
  • PorchRocket’s analysis of April 2026 USPVDB v4.0 counts 5,976 U.S. projects classified only as ground-mounted systems.
  • PorchRocket’s analysis of April 2026 USPVDB v4.0 counts 407 U.S. projects classified only as rooftop systems.

What counts as a solar farm in EIA data?

EIA counts reporting power plants and generators, so its solar inventory includes qualifying installations beyond ground-mounted farms and uses a plant-level 1 MW reporting threshold.

The EIA-860 instructions define the reporting population around a grid-connected plant’s combined generator nameplate capacity. Because a generator below 1 MW can belong to a qualifying plant, dropping every generator below that size would change the reporting population and remove records the survey intends to collect.

The counting unit matters just as much. A plant is one distinct EIA plant identifier. A generator is the combination of plant identifier and generator identifier. A solar generator record can represent a grouped installation; it does not mean one panel, one inverter or one fenced compound. For an account list, start with plants, then investigate how the site and its purchasing responsibility are arranged.

This analysis selects solar by primary energy source code SUN. Photovoltaic and solar-thermal technologies both qualify. Batteries use prime mover code BA. Other storage technologies stay outside that category. Both operating totals require status OP. Read the status field.

Use these labels when requesting a count:

Measure Counting rule Decision it supports
Operating solar plants Distinct EIA plant IDs with an OP solar generator Reported locations to research
Operating solar generators Distinct plant and generator ID pairs Capacity and online-year analysis
Ground-mounted PV projects USPVDB project IDs with system type exactly ground A narrower mapped farm inventory
Operating battery plants Distinct EIA plant IDs with an OP battery generator Storage-site research
Nameplate MW Sum of qualifying generators’ reported AC power Installed power comparison

Source for EIA definitions: EIA-860 instructions; mapped-project definitions: USPVDB data documentation.

The commercial building statistics reference is the broader hub for building-based markets. Keep its building counts separate from these plants.

Number of solar farms by state

California’s 983 operating plants are the largest EIA solar-plant count in September 2026; Texas leads the same inventory’s nameplate capacity.

The EIA columns below are PorchRocket’s calculations from the September 2026 raw generator workbook. The final column comes from the USPVDB v4.0 raw CSV (April 14, 2026). It keeps only projects whose system type is exactly ground. Each table uses its own reporting population and date.

State EIA solar plants, Sep 2026 EIA solar generators EIA solar MW AC USPVDB ground-only projects, Apr 2026
Alabama 15 15 986.3 11
Alaska 2 18 7.7 2
Arizona 128 171 9,095.0 90
Arkansas 74 74 2,796.1 63
California 983 1,161 25,609.3 677
Colorado 187 193 3,398.2 151
Connecticut 95 102 501.5 71
Delaware 28 30 139.0 15
District of Columbia 17 18 35.2 5
Florida 227 227 13,492.2 188
Georgia 154 157 5,495.6 148
Hawaii 35 57 578.0 32
Idaho 14 15 826.6 12
Illinois 479 487 5,876.1 215
Indiana 126 127 5,293.4 106
Iowa 37 37 700.6 28
Kansas 39 39 434.4 36
Kentucky 21 21 1,375.7 14
Louisiana 21 21 1,555.9 13
Maine 177 178 967.9 140
Maryland 217 221 1,406.5 134
Massachusetts 557 572 1,505.2 493
Michigan 79 80 3,310.1 63
Minnesota 547 728 2,175.4 521
Mississippi 26 26 1,724.5 22
Missouri 38 39 1,897.8 28
Montana 9 10 181.0 9
Nebraska 28 29 159.8 18
Nevada 70 107 5,479.9 57
New Hampshire 3 3 8.4 1
New Jersey 358 420 1,309.4 173
New Mexico 113 116 3,300.8 86
New York 742 762 3,209.1 611
North Carolina 769 798 7,076.1 756
North Dakota 0 0 0.0 0
Ohio 79 82 4,969.5 60
Oklahoma 23 23 1,184.9 15
Oregon 152 152 1,462.2 139
Pennsylvania 81 84 1,227.3 57
Rhode Island 77 85 451.5 69
South Carolina 129 130 1,999.0 118
South Dakota 4 4 217.3 3
Tennessee 54 56 1,243.6 29
Texas 250 264 34,977.4 176
Utah 59 62 3,454.4 52
Vermont 54 54 155.5 53
Virginia 145 150 5,308.3 119
Washington 10 10 348.4 7
West Virginia 6 6 259.3 4
Wisconsin 119 123 2,556.7 84
Wyoming 2 2 242.0 2
United States 7,659 8,346 171,966.0 5,976

For the search “number of solar farms by state,” choose the column before quoting a number. Use EIA for the more recent operating inventory, or USPVDB’s ground-only column for a narrower system-type definition; differing coverage and observation dates prevent interpreting their difference as a count of missing farms.

Massachusetts has 557 operating solar plants and 1,505.2 MW AC; Texas has 250 plants and 34,977.4 MW AC. North Carolina leads the narrower ground-only mapped count with 756 projects. I would sort by locations for service outreach, then apply the crew’s size and travel requirements.

A zero means no qualifying operating record survived the filters. Use the solar farm location guide to move from totals to researched sites.

Battery storage sites and capacity by state

Texas leads operating battery power at 19,194.6 MW, while California leads battery-plant count with 277 in PorchRocket’s September 2026 EIA analysis.

All rows below use the September 2026 EIA-860M workbook. The filter requires status OP with battery prime mover BA. Power is nameplate MW; energy is the sum of reported MWh. The national energy sum covers 1,124 of the 1,125 operating battery generators because the remaining record has no reported energy value. That generator is in California, so its missing MWh also affects the state’s energy total.

State Battery plants Battery generators Nameplate MW Reported energy MWh
Alabama 1 1 1.0 2.0
Alaska 6 6 131.7 188.9
Arizona 56 61 6,877.8 24,979.6
Arkansas 5 6 28.2 73.0
California 277 306 16,267.1 55,539.1
Colorado 18 18 670.4 2,414.7
Connecticut 2 2 1.6 6.3
Delaware 0 0 0.0 0.0
District of Columbia 0 0 0.0 0.0
Florida 26 26 1,202.2 3,093.1
Georgia 9 10 863.7 2,764.4
Hawaii 15 15 514.9 1,623.2
Idaho 10 12 659.0 2,632.0
Illinois 7 8 82.3 54.1
Indiana 9 9 587.3 1,501.6
Iowa 6 6 109.7 125.6
Kansas 0 0 0.0 0.0
Kentucky 0 0 0.0 0.0
Louisiana 0 0 0.0 0.0
Maine 8 9 239.0 436.0
Maryland 5 5 12.3 22.1
Massachusetts 131 133 738.7 1,598.8
Michigan 4 4 120.3 457.5
Minnesota 5 5 21.0 63.6
Mississippi 4 4 151.5 601.5
Missouri 0 0 0.0 0.0
Montana 0 0 0.0 0.0
Nebraska 6 6 9.0 25.5
Nevada 15 18 2,104.0 6,366.0
New Hampshire 1 1 14.0 14.0
New Jersey 12 12 130.1 136.6
New Mexico 21 21 1,360.2 5,179.0
New York 53 57 249.7 746.8
North Carolina 40 40 372.1 884.0
North Dakota 0 0 0.0 0.0
Ohio 5 6 28.8 29.2
Oklahoma 5 5 432.0 1,513.0
Oregon 6 6 523.0 2,056.0
Pennsylvania 3 3 34.6 28.7
Rhode Island 1 1 3.0 10.0
South Carolina 3 3 88.0 278.0
South Dakota 1 1 0.8 0.8
Tennessee 0 0 0.0 0.0
Texas 261 265 19,194.6 31,594.8
Utah 7 7 1,070.0 3,320.0
Vermont 8 8 20.9 53.3
Virginia 5 5 230.6 896.7
Washington 3 3 4.0 9.8
West Virginia 2 2 110.5 313.1
Wisconsin 9 9 314.9 1,237.4
Wyoming 0 0 0.0 0.0
United States 1,071 1,125 55,574.5 152,869.8

The Department of Energy’s storage explanation distinguishes power capacity from energy capacity. MW describes the discharge rate; MWh describes the amount stored. Keep both fields in a storage inventory. A ranking by reported MWh need not match a ranking by MW, and neither field is a count of cabinets or containers.

Solar and batteries share 477 operating plant IDs in this calculation. The union is therefore 7,659 + 1,071 - 477 = 8,253 distinct plant IDs with operating solar, batteries or both. Keep the technology categories for capacity analysis, but remove repeated plant IDs before treating their sum as a prospect count.

Matching plant IDs establish a reporting relationship. They do not establish electrical coupling, a shared maintenance contract or a single person authorized to purchase service. Likewise, the 594 battery plant IDs without operating solar under the same identifier should not automatically be labeled physically separate from all solar projects. That requires a site-level check.

For a storage campaign, I would ask the researcher to preserve the plant and generator identifiers, then add the observed compound and access location. The battery storage site location guide covers that handoff. The data center power statistics page covers another energy-related market; its facilities should remain a separate category in the sales database.

How many solar farms are in Texas?

Texas has 250 operating EIA solar plants with 264 generators and 34,977.4 MW AC in September 2026; USPVDB’s April ground-only count is 176 projects.

Both figures reproduce from the downloadable data, but they answer different requests. For a current inventory brief, I would write “250 operating solar plants in EIA-860M, September 2026.” For a map of ground-mounted PV projects, I would write “176 ground-only USPVDB projects, April 2026.” Keep the dataset name attached to the count.

Texas also has 261 operating battery plants and 19,194.6 MW of battery nameplate power in the September file. Its planned inventory contains 37,298.8 MW AC of solar and 25,301.5 MW of batteries across all reported planned operation dates. Those planned capacities are separate from the operating figures and should stay out of an installed-base sales target.

For a Texas outreach brief, define the counties your team can serve and the site types it accepts. Preserve the source date and reporting status. Keep unresolved locations in a separate queue until someone answers the missing question. Assign that check before dispatching an estimator.

Solar and battery inventory changes since 2025

Operating solar nameplate capacity increased 11.5% and battery nameplate power increased 28.8% between final 2025 EIA data and the September 2026 preliminary inventory.

PorchRocket applied identical geography and technology filters, retaining OP status in both the final 2025 EIA-860 workbooks and the September 2026 EIA-860M workbook. The differences below describe the change between those inventories.

Measure Final 2025 September 2026 Change in inventory Percent change
Solar plants 7,317 7,659 +342 +4.7%
Solar generators 7,990 8,346 +356 +4.5%
Solar nameplate MW 154,251.1 171,966.0 +17,714.9 +11.5%
Battery plants 933 1,071 +138 +14.8%
Battery generators 981 1,125 +144 +14.7%
Battery nameplate MW 43,134.6 55,574.5 +12,439.9 +28.8%

Inventory change is different from gross construction. Because a comparison can reflect additions and retirements alongside status changes or revised capacities, the September workbook’s units with a 2026 operating year need not sum to the difference between the two capacity snapshots. The next table answers the operating-year question directly.

Keep the annual file as a fixed baseline. Refresh the same monthly fields under the same filters. A revised definition can otherwise look like growth. For a service plan, retain plant counts alongside MW; neither measure establishes how many contracts became available.

Solar and battery generators by year online

Units commissioned in 2026 account for 17,645.6 MW of solar and 12,452.4 MW of batteries in September’s operating inventory, spread across 339 and 145 generators respectively.

These cohorts group the Operating Year field in September 2026 EIA-860M. They include only generators still marked OP in that snapshot. They are not historical totals of everything built during each year.

Operating-year cohort Solar generators Solar MW AC Battery generators Battery MW
2015 and earlier 1,679 13,745.2 21 219.0
2016 to 2020 2,997 34,809.5 172 1,223.0
2021 697 13,773.7 124 3,005.8
2022 501 11,301.5 128 4,124.5
2023 625 19,813.4 134 7,114.5
2024 775 31,167.9 202 11,064.8
2025 733 29,709.2 199 16,370.5
January to September 2026 339 17,645.6 145 12,452.4

Generators are the right counting unit for this table. A plant can contain units with different operating years; counting the entire plant in each cohort would repeat it. Preserve that distinction when someone asks for “solar farms built in 2025.” This table supplies operating generator cohorts, not unique new-farm counts.

An online year does not establish component condition or the date a battery module was replaced. Request service records before proposing replacement work. The rooftop unit statistics reference applies similar counting discipline to another equipment market. The asset terminology glossary separates installed base from replacement demand.

Planned solar and battery additions

The September 2026 planned inventory contains 1,101 solar generators totaling 122,225.2 MW AC and 483 battery generators totaling 64,234.6 MW across reported planned dates.

PorchRocket calculated the following tables from the Planned sheet of EIA’s September 2026 workbook. They exclude the separate Canceled or Postponed sheet. A planned generator remains planned even if its reported operation date has passed.

Planned additions by reported operation date

Reported planned operation Solar generators Solar MW AC Battery generators Battery MW
September 2026 or earlier 7 318.1 3 105.3
October to December 2026 349 23,375.0 120 9,544.5
2027 465 44,052.2 212 26,847.5
2028 and later 280 54,479.9 148 27,737.3

The elapsed-date bucket still contains planned records. Do not promote those rows to operating capacity yourself. Use a later release or documented site check. Keep 2027 projects on a watch list and recheck their status before assigning travel or estimating hours.

Construction stage of planned generators

Development stage Solar generators Solar MW AC Battery generators Battery MW
Pre-construction (P, L, T) 490 75,475.5 244 40,936.2
Under construction (U, V) 514 35,947.9 210 20,315.6
Construction finished, not commercial (TS) 97 10,801.8 29 2,982.8

The stage codes come from the workbook’s status descriptions. P means approvals have not been initiated, L means approvals are pending, and T means approvals have been received without construction starting. U and V describe construction progress. TS means construction has finished but commercial operation has not started.

I would separate stages in the outreach calendar. Ask which stage makes your service relevant and filter on that criterion. Capacity alone puts early plans beside finished construction.

States with the most planned solar capacity

State Planned solar generators Planned solar MW AC Planned battery generators Planned battery MW
Texas 167 37,298.8 152 25,301.5
California 88 7,816.3 89 9,935.9
Arizona 25 6,168.6 26 6,558.0
Oregon 17 4,540.2 9 2,940.0
Michigan 25 3,886.4 3 420.0
Georgia 16 3,818.0 3 533.5
New York 88 3,745.3 33 783.7
Nevada 14 3,726.0 10 2,919.5
Indiana 20 3,553.0 1 20.0
Arkansas 15 3,504.1 4 1,100.0

This selection ranks states by total planned solar nameplate MW across all reported dates, then shows their battery plans beside them. The CSV includes planned plant and generator totals for every state, with capacity and reported operation-year cuts. Planned plant totals count identifiers with planned units; they can include plants that already have other operating units.

An added unit at an existing plant is not automatically another purchasing account. Keep both identifiers so a refresh can resolve the relationship.

What USPVDB adds to the solar farm count

USPVDB v4.0 supplies mapped photovoltaic projects and system types, letting PorchRocket identify 5,976 exclusively ground-mounted projects separately from EIA’s broader operating inventory.

The USGS data documentation describes visually checked locations and array boundaries. Its help guide explains that known projects below 1 MW DC are excluded and that some larger rooftop and parking-canopy arrays can appear. That is why this article filters the system-type field rather than assuming every mapped project is a farm.

PorchRocket’s calculation of the v4.0 CSV archive finds 6,611 projects across all system types, including 407 classified only as rooftop and 148 classified only as canopy. Mixed categories remain outside the ground-only column, even when ground is one of their listed types. Keeping an exact-category rule makes the count reproducible.

Use the map data to check where an array sits and to distinguish its footprint from a general plant location. Preserve the image timestamp. A release date and an image date are different fields, and a recently published dataset does not mean every location was photographed recently.

EIA plant identifiers and USPVDB project records describe different inventories. Subtracting their totals cannot establish an omission rate. Review unmatched records individually; a similar name is insufficient. The aerial asset inventory guide describes that review process.

How to size a solar and storage service territory

Start with operating locations in the service area, then estimate account gross profit and sales hours before treating a state capacity ranking as revenue potential.

Say a solar vegetation and exterior-service crew near Raleigh researches 240 candidate records within its chosen territory. Assume a 240-hour sales budget, $47 per sales hour, six hours per site visit including travel and follow-up, and the conversion rates shown below. Reviewed targets reduce office research from 40 to 12 minutes per record; the crew uses the released hours to visit more fitting sites.

Assumption or calculation Unscreened records Reviewed targets
Records researched 240 240
Office research minutes per record 40 12
Office research hours 160 48
Office labor at $47/hour $7,520 $2,256
Site visits, six hours each including follow-up 12 20
Visit hours 72 120
Hours left from a 240-hour sales budget 8 72
Visits fitting scope and buyer requirements 50% 80%
Qualified visits 6 16
Proposals at 50% of qualified visits 3 8
Expected accounts at 20% of proposals 0.6 1.6

Source: PorchRocket’s planning model, reproduced in the downloadable dataset. Qualified means the site fits the crew’s scope and buyer requirements. The same proposal and win rates apply to both columns.

The research calculation is 240 x (40 - 12) / 60 = 112 hours, carrying $5,264 of labor value at $47 an hour. Those hours can be reassigned to selling. They become a cash saving only if the business avoids the corresponding expense. The table leaves time for other sales work in both cases.

For customer lifetime value, assume each account pays $18,700 annually at a 32% gross margin for five paid years, with no discounting. Annual gross profit is $18,700 x 32% = $5,984. Five-year gross profit is $5,984 x 5 = $29,920 per account, before acquisition costs.

The sales funnel produces 0.6 expected accounts from unscreened records and 1.6 from reviewed targets, about 2.7 times as many within the hours budget. These are expected values across repeated campaigns. The increase comes from more qualified visits; the assumed proposal and win rates stay fixed. The additional account represents $29,920 in five-year gross profit before acquisition costs.

Change the qualified share before approving the spending plan:

Scenario Qualified share of 20 visits Expected accounts won Five-year gross profit
Low 40% 0.8 $23,936
Base 65% 1.3 $38,896
High 80% 1.6 $47,872

Source: the same PorchRocket model; each row uses 20 visits, a 50% proposal rate, a 20% win rate and $29,920 of five-year gross profit per account. The base case uses a 65% qualified share.

The decision rule is to compare incremental gross profit with the quoted search cost and other acquisition expenses, including extra travel or estimating time that the research-hour comparison does not capture. Customer acquisition cost is all acquisition spending divided by accounts won. Lead-list ROI is account gross profit minus list and outreach costs, divided by those acquisition costs. Enter the scope-specific quote in the lead-list ROI calculator, then reduce the conversion assumptions until the decision changes.

The fastest way to find buildings for solar outreach

Ground Control by PorchRocket finds commercial buildings for solar outreach by business type, building size and visible installed solar in the area you choose. Commercial buildings cost $10 to $20 each, with the file delivered within 48 hours after payment and wrong matches credited within 30 days. Owner information and decision-maker research are add-ons. These are target properties for outreach. Use the solar property lists or Ground Control for building-based campaigns; request Asset Search for custom investigations of solar farms or battery compounds.

Keep public opportunities in their own pipeline. Bid Control by PorchRocket finds solar work for cities, counties and schools, keeping the papers and addenda alongside deadlines. A public project’s solicitation determines the next step; its appearance in an installed-base table does not establish an open contract.

How to build a solar and storage site list, step by step

Use EIA identifiers and dated site evidence to build a territory list, then qualify purchasing responsibility before sending an estimator to a solar or battery location.

  1. Choose the service and territory. Write the counties or travel boundary and accepted system types. Separate installed-site service from selling new solar installations on buildings. Specify what the research team should exclude.
  2. Download a dated baseline. Start with the September 2026 EIA operating file. Preserve plant and generator identifiers. Remove repeated plant IDs from an account-level view and store the source edition beside each record.
  3. Commission the missing site investigation. Tell Asset Search what counts and where to search. Request parcel and owner information, dated views, and fits, rejects and unresolved results in a paid pilot credited toward the larger order. Agree on the field check that resolves each uncertain category.
  4. Check system type and location. Review USPVDB’s mapped PV context against the site’s dated evidence. Keep the reporting plant separate from the observed compound until their relationship is established. Do not match on name alone.
  5. Identify the purchasing path. Ask who approves the scope and manages access. Check whether an existing contract controls the work. Record the next agreed action.
  6. Visit the sites that fit. Confirm access with the authorized contact. Capture measurements for the estimate. Keep observed features separate from the technical records required to specify electrical or battery work.
  7. Track acquisition cost and account value. Record research hours and proposals, then signed accounts. Calculate gross profit after delivery costs. Replace the model’s assumptions with your own rates.
  8. Refresh on the buying schedule. Recheck a service territory quarterly. Revisit planned projects when their status changes. Keep prior evidence dates and route unresolved cases to the assigned check.

I would buy the pilot before a statewide investigation. Check whether its evidence answers the service manager’s question and its file fits the team’s working system.

Methodology

PorchRocket computed these tables from EIA-860 (final 2025) and EIA-860M (preliminary September 2026), plus USPVDB v4.0, using distinct counting units.

The build date is November 5, 2026. EIA’s monthly release page lists September 2026 as the latest inventory, published October 23, 2026. The annual baseline is final 2025. USPVDB is version 4.0 dated April 14, 2026. The EIA reporting periods and USPVDB edition date identify each table; individual map-image dates remain separate.

The script reads downloaded raw files and records input URLs with SHA-256 checksums. It runs with python -I and writes the eight-column CSV: stat, value, unit, year, geography, source_name, source_url and method. It also generates the numerical tables and money model.

The filters and calculations are:

  • Geography: retain the 50 states and District of Columbia; exclude Puerto Rico and other territories from national totals.
  • Operating status: retain OP only in both EIA editions. Standby, out-of-service, retired and planned rows do not enter operating counts.
  • Technology: select primary source SUN for solar and prime mover BA for batteries. Keep reported generators below 1 MW at qualifying plants.
  • Plant and generator counts: count distinct plant IDs for plants, and distinct plant-plus-generator identifiers for generators. Assert that selected generator identifiers are unique.
  • Capacity: sum Nameplate Capacity (MW) as AC power. For annual battery energy, join the energy-storage workbook on plant and generator identifiers; for monthly energy, sum reported MWh directly and preserve the reported-value coverage count.
  • Online cohorts: group September operating generators by Operating Year. Older cohorts contain surviving operating units rather than every historical installation.
  • Plans: use the Planned sheet, retaining reported dates and stages; group capacity and identifiers by state, operation year and stage. Keep canceled or postponed records separate.
  • Mapped farms: count unique USPVDB case IDs where p_sys_type is exactly ground; mixed-system records are excluded from that column.

The script reconciles state and national plant counts, then generator counts and MW. It also checks the solar-and-battery union arithmetic. MW, MWh and percentages display to one decimal. Model dollars round to whole dollars. Calculations use unrounded values.

EIA’s monthly inventory is preliminary, and its reporting population does not cover every small solar or battery installation. USPVDB has different mapping dates and system coverage, so its project totals cannot be used to measure EIA omissions. Plant identifiers and imagery establish an inventory baseline; current site conditions and purchasing responsibility require their own evidence, as does electrical coupling.

How to cite these statistics

Credit PorchRocket’s analysis and name the underlying release when quoting a state count, because September 2026 EIA plants and April 2026 USPVDB projects differ.

Suggested citation: PorchRocket, “Number of Solar Farms by State: Solar and Storage Statistics 2026,” 2026, https://porchrocket.com/solar-and-storage-sites-by-state

Reuse these statistics with a link back to this page. Download the solar and storage statistics CSV to retain the units, geography, source URLs and row-level method. Use the year field with the value rather than replacing all observation dates with the page’s publication year.

Label a chart with its counting unit and observation date. Keep planned capacity separate from installed capacity, and distinguish battery MW from MWh.

Asset Search returns locations and dated evidence for the service area you choose, with people reviewing every candidate against rules agreed with your experts. Tell Asset Search what to find and where in the contiguous United States: ground-mounted arrays, large rooftop systems, battery compounds or a custom combination. The return includes locations with parcels and owner information. Each record carries dated views and observed facts plus a fits, rejects or unresolved answer; unresolved sites name the field check needed. Delivery is CSV, Excel or GIS-ready files. Per accepted result pricing fits a service territory whose acceptance rules are clear; a small paid pilot is credited toward the full order, and quarterly monitoring suits a territory book. Start with the sites your crew can serve and the evidence your estimator needs.

Questions people ask

Quick answers

Which state has the most solar farms?

California leads the broader EIA solar-plant count with 983 operating plants in PorchRocket's September 2026 calculation. North Carolina leads the narrower USPVDB ground-only count with 756 mapped projects in April 2026. Texas leads solar nameplate capacity with 34,977.4 MW AC. Retain the dataset and date when quoting a count.

How many solar farms are in Texas?

Texas has 250 operating EIA solar plants containing 264 generators in PorchRocket's September 2026 analysis. The narrower April 2026 USPVDB count is 176 exclusively ground-mounted projects. The counts use different dates and definitions, so choose one measure and retain its label when quoting the figure.

How many battery storage sites are in the United States?

PorchRocket counts 1,071 operating battery plants containing 1,125 battery generators in September 2026 EIA-860M. Their nameplate power totals 55,574.5 MW. A plant is a distinct EIA plant ID, so this count describes reported plants rather than individual battery containers, electrical cabinets or service contracts.

Which state has the most battery storage capacity in 2026?

Texas leads battery nameplate power at 19,194.6 MW in PorchRocket's September 2026 EIA-860M calculation. California follows at 16,267.1 MW and has more battery plants. Energy capacity is a separate measure: compare the table's reported MWh column when the question concerns the amount of stored energy.

Does the solar farm count include rooftop solar?

Yes, the EIA solar-plant column can include qualifying rooftop installations at reporting plants. It is broader than a ground-mounted farm count. The USPVDB ground-only column excludes rooftop, canopy, floating and mixed-system categories. Neither column is a census of household solar installations.

What is the difference between battery MW and MWh?

MW measures power, or the rate at which a battery can discharge electricity. MWh measures energy, or how much it can store. This page reports nameplate power and reported energy separately. Confirm both ratings and the operating conditions for a particular project before estimating its discharge duration.

Are planned solar and battery projects already operating?

No. Planned rows remain outside the operating totals, including projects whose reported planned dates have already passed. The September 2026 file separates planned generators from operating units. Construction stage and reported operation date should travel with a planned-project count, especially when using it to schedule outreach.

Can I download and reuse the state statistics?

Yes. Download the CSV linked in the citation section and credit PorchRocket's analysis with the source edition and reporting date. Each row includes its measure, value, unit, geography, source URL and method. Keep EIA plants, EIA generators and USPVDB projects separate when combining the data with another inventory.

Research file

Sources used in this guide

Sources are linked at the claim they support and collected here for auditability. Vendor features and prices can change; verify them before purchasing.

  1. EIA-860M release calendar and monthly inventoryU.S. Energy Information Administration: Confirms September 2026 as the latest available inventory, released October 23, 2026; checked November 5, 2026.
  2. EIA-860 annual data and workbook definitionsU.S. Energy Information Administration: Confirms the final 2025 annual edition and explains the generator and storage workbooks; checked November 5, 2026.
  3. Final 2025 EIA-860 raw workbooksU.S. Energy Information Administration: Annual operating baseline from Schedule 3 generator and energy-storage workbooks; downloaded and checked November 5, 2026.
  4. September 2026 EIA-860M raw inventoryU.S. Energy Information Administration: Latest monthly inventory available November 5, 2026; released October 23, 2026; source of current state counts and plans.
  5. EIA-860 reporting instructionsU.S. Energy Information Administration: Plant reporting threshold, AC capacity, generator grouping and operating-status definitions; checked November 5, 2026.
  6. USPVDB data and methodsU.S. Geological Survey: Version 4.0, April 14, 2026; mapped photovoltaic facilities, imagery verification and coverage limits; checked November 5, 2026.
  7. USPVDB version 4.0 raw CSV archiveU.S. Geological Survey: Downloaded archive contains uspvdb_v4_0_20260414.csv; used for project counts by exact system type and state.
  8. USPVDB help and field definitionsU.S. Geological Survey: Project identifier, AC and DC capacity, image timestamp and exclusion of known systems below 1 MW DC; checked November 5, 2026.
  9. Solar energy and storage basicsU.S. Department of Energy: Distinction between storage power and energy, and between co-located and separate storage systems; checked November 5, 2026.
Your next best move

Map the solar sites your team can serve.

Tell us what to find and where. Asset Search returns reviewed locations, owner information and dated evidence in your format, starting with a scoped pilot.

Find solar sites