This reference covers state totals, largest reported fuels, national history and source accounting, with fuel rankings restricted to numeric source rows, exact cells and downloads. The data describe electricity generated in a location, not electricity consumed there, installed capacity, utility bills or a customer’s electricity supply contract. Separate estimates of small solar generation are not added to this EIA923-based series.
Key Takeaways
- 40.93% of U.S. net generation came from natural gas in the selected final 2025 industry series. (EIA, state generation workbook, Sep 2026).
- 17.75% came from nuclear and 16.61% from coal in the same 2025 series. (EIA, state generation workbook, Sep 2026).
- Texas generated 600.98 TWh in 2025, the largest all-source state total. (EIA, state generation workbook, Sep 2026).
- Natural gas was the largest reported fuel in 28 of the 51 state and DC generation mixes. (EIA, state generation workbook, Sep 2026).
- Coal was the largest reported fuel in nine states, while wind led in five. (EIA, state generation workbook, Sep 2026).
- Texas generated 290.25 TWh from gas but ranked fifteenth for its reported gas share, 48.30%. (EIA, state generation workbook, Sep 2026).
- Rhode Island’s 88.26% gas share led the 50 of 51 state/DC geographies with numeric gas rows. (EIA, state generation workbook, Sep 2026).
- West Virginia’s 86.67% coal share led the 47 of 51 state/DC geographies with numeric coal rows. (EIA, state generation workbook, Sep 2026).
- Texas led reported wind volume, while Iowa led reported wind share at 58.48%. (EIA, state generation workbook, Sep 2026).
- Texas led reported solar volume, while Nevada led the reported solar share at 30.79%. (EIA, state generation workbook, Sep 2026).
- National gas share increased 8.20 percentage points between 2015 and 2025 in this source edition. (EIA, state generation workbook, Sep 2026).
- National coal share decreased 16.55 percentage points over the same endpoints. (EIA, state generation workbook, Sep 2026).
- 25 of the 51 state and DC mixes had one reported fuel supplying more than half of generation. (EIA, state generation workbook, Sep 2026).
- Source totals are preserved even where summed fuel rows differ by up to 2 MWh. (EIA, state generation workbook, Sep 2026).
What does the U.S. electricity generation mix look like?
Natural gas was the largest reported national source in 2025. Nuclear and coal followed, while wind, the reported solar category and conventional hydro supplied additional large volumes. The denominator is the workbook’s published Total Electric Power Industry net generation, not national electricity consumption.
Figure 1 in text
NATIONAL OUTPUT · 2025 The national generation mix in 2025 Reported fuel output, TWh, using the published all-source total. 0 1,000 1,800 Natural gas
1,811.00 Nuclear
785.30 Coal
734.93 Wind
461.58 Solar (reported)
300.30 Conventional hydro
241.30 Wood fuels
31.18 Petroleum
19.24 Geothermal
15.63 Other biomass
14.11 Other gases
9.53 Other
6.28 Pumped storage
-5.36 TWh of net generation EIA final 2025 industry series, Sep 2026. All 13 national reported fuel rows. Negative pumped storage remains signed; separate small solar estimates are excluded.
Source and data: EIA state generation workbook
Permanent link to this figure’s text and data · Sources and context
How much electricity came from the principal sources?
The national comparison keeps each producer fuel label and calculates its share of the published all-source total. One TWh equals one million MWh, so the article’s readable volumes are direct unit conversions from the source fields.
- The national published all-source total was 4,425.03 TWh in 2025. (EIA, state generation workbook, Sep 2026).
- Natural gas contributed 1,811.00 TWh, or 40.93%. (EIA, state generation workbook, Sep 2026).
- Nuclear contributed 785.30 TWh, or 17.75%. (EIA, state generation workbook, Sep 2026).
- Coal contributed 734.93 TWh, or 16.61%. (EIA, state generation workbook, Sep 2026).
These values concern electrical output after the source’s net generation accounting. They are not the heat content of fuel delivered to plants. A gas volume in cubic feet, a coal weight in tons and electricity output in MWh cannot be substituted for one another. Conversion to fuel input would require heat content, technology and efficiency evidence beyond this workbook. A writer describing national electricity output can use these figures directly, while retaining the series and data year in the sentence. (EIA, state generation workbook, Sep 2026). For the separate question of federal coal royalty dollars, use the federal coal royalty revenue reference.
How much came from wind, solar and conventional hydro?
The workbook reports wind, solar thermal and photovoltaic generation, and conventional hydro as separate categories. AMR preserves that taxonomy instead of folding them into a new renewable definition whose scope would be unclear.
- Wind produced 461.58 TWh nationally, 10.43% of the published total. (EIA, state generation workbook, Sep 2026).
- The reported Solar Thermal and Photovoltaic category supplied 300.30 TWh, or 6.79%. (EIA, state generation workbook, Sep 2026).
- Conventional hydro supplied 241.30 TWh, or 5.45%. (EIA, state generation workbook, Sep 2026).
- Geothermal contributed 15.63 TWh, or 0.35%. (EIA, state generation workbook, Sep 2026).
The solar figure belongs to this generating-facility series. EIA publishes separate estimates for small solar photovoltaic systems through other survey inputs and estimation methods. Those estimates are not silently appended here, because adding one fuel’s extra output without rebuilding the all-source denominator would change the meaning of every share. The EIA reporting instructions define the survey’s grid connection and one MW plant threshold; the technical notes separately explain small solar estimation. (EIA, state generation workbook, Sep 2026).
Why are the smaller and negative categories retained?
Small categories still contribute to the accounting, and a negative net value is information rather than an error to erase. The national source includes separate other gases, petroleum, biomass, wood, other and pumped storage rows.
- Petroleum produced 19.24 TWh nationally in 2025. (EIA, state generation workbook, Sep 2026).
- Wood and Wood Derived Fuels contributed 31.18 TWh. (EIA, state generation workbook, Sep 2026).
- Other Gases contributed 9.53 TWh and Other Biomass 14.11 TWh. (EIA, state generation workbook, Sep 2026).
- The national Pumped Storage row reports negative 5,361,958 MWh, a negative 0.121173% share of the published total. (EIA, state generation workbook, Sep 2026).
Net generation subtracts electricity used to operate the generating facility from gross production. Pumped storage can consume more electricity in pumping and operation than it returns through generation over an annual period. Its negative value should not be converted to zero to make a chart easier to read. Other signed source values are also retained without guessing their individual causes. The accounting table allows a reader to inspect all reported rows, including the categories whose shares disappear when a display rounds every value to whole percentages. (EIA, state generation workbook, Sep 2026).
Which source supplies the largest reported share in each state?
Natural gas was the largest reported fuel in 28 state and DC generation mixes. Coal led in nine, wind in five, conventional hydro in four, nuclear in four and petroleum in one. The classification compares each geography’s reported fuel volumes using its own source total.
Figure 2 in text
COMPLETE NAMED INVENTORY · 2025 The largest reported source, state by state All 50 states and DC, grouped by their largest reported fuel in 2025. Natural gas
28 of 51 Alabama, Alaska, Arizona, Arkansas, California, Colorado, Connecticut, Delaware, District of Columbia, Florida, Georgia, Louisiana, Maine, Maryland, Massachusetts, Michigan, Mississippi, Nevada, New Jersey, New York, North Carolina, Ohio, Oklahoma, Pennsylvania, Rhode Island, Texas, Virginia, Wisconsin Coal
9 of 51 Indiana, Kentucky, Missouri, Montana, Nebraska, North Dakota, Utah, West Virginia, Wyoming Wind
5 of 51 Iowa, Kansas, Minnesota, New Mexico, South Dakota Conventional hydro
4 of 51 Idaho, Oregon, Vermont, Washington Nuclear
4 of 51 Illinois, New Hampshire, South Carolina, Tennessee Petroleum
1 of 51 Hawaii Source: EIA, Sep 2026. Largest reported fuel, not necessarily a majority. Each state and DC appears once; no geographic area weighting.
Source and data: EIA state generation workbook
Permanent link to this figure’s text and data · Sources and context
How many state mixes were led by each fuel?
Counting the largest reported fuel gives each state or DC one entry. It answers how widespread the largest source is across geography averages; it does not measure that fuel’s share of national output or the number of customers using it.
- Natural gas led 28 of the 51 state and DC mixes in 2025. (EIA, state generation workbook, Sep 2026).
- Coal led nine states; wind led Iowa, Kansas, Minnesota, New Mexico and South Dakota. (EIA, state generation workbook, Sep 2026).
- Conventional hydro led Idaho, Oregon, Vermont and Washington. (EIA, state generation workbook, Sep 2026).
- Nuclear led Illinois, New Hampshire, South Carolina and Tennessee; petroleum led Hawaii. (EIA, state generation workbook, Sep 2026).
The complete table identifies the largest reported fuel, its MWh and its fraction of the same state total. No omitted fuel row is filled with zero to create this classification. The relevant question is explicitly the largest reported row in the source. Gas appearing first in more states does not imply that it supplied a majority in each one. Nor does a count of states weight Texas and DC by their very different generation totals. Use a national volume calculation when the story concerns output rather than geographic prevalence. (EIA, state generation workbook, Sep 2026).
Does the largest source always supply a majority?
No. A fuel can be the largest individual category while other categories together supply more electricity. The distinction is particularly useful when a state is described as relying on one source even though its reported mix is more divided.
- 25 of the 51 state and DC mixes had a largest reported fuel share above 50%; the other 26 did not. (EIA, state generation workbook, Sep 2026).
- Texas’s largest reported source was gas at 48.30% of its 2025 total. (EIA, state generation workbook, Sep 2026).
- California’s largest reported source was gas at 36.22%. (EIA, state generation workbook, Sep 2026).
- Vermont’s largest reported source was conventional hydro at 49.73%. (EIA, state generation workbook, Sep 2026).
A statement such as “gas was the largest source” is supported for Texas, but “most Texas generation came from gas” would be incorrect if most means more than half. The table makes that distinction testable. A majority threshold is also not a reliability score or an estimate of supply security. Those questions involve the timing of generation, availability, network connections and other variables absent from these annual totals. The classification here is a descriptive annual output comparison. (EIA, state generation workbook, Sep 2026).
Which states generated the most electricity overall?
Total output ranks a geography’s scale rather than its fuel preference. The national contribution of a large generator can be important even when the fuel is a smaller fraction of that state’s own mix than elsewhere.
- Texas generated 600.98 TWh in 2025, the largest state total. (EIA, state generation workbook, Sep 2026).
- Florida generated 268.26 TWh and Pennsylvania 245.72 TWh. (EIA, state generation workbook, Sep 2026).
- California generated 205.11 TWh, followed by Illinois at 191.82 TWh. (EIA, state generation workbook, Sep 2026).
- DC’s published all-source generation total was 0.20 TWh. (EIA, state generation workbook, Sep 2026).
State generation is not a state’s electricity sales or demand. Power can be traded across state lines, and generation facilities can serve markets larger than their physical location. This dataset does not include the necessary flow and sales records to reconcile those exchanges. The totals therefore cannot identify how much electricity a state imported, exported or consumed. A newspaper comparing output can quote the figures; a story about dependence on outside electricity needs a separate, matched state electricity balance. (EIA, state generation workbook, Sep 2026).
Why do gas and coal volume rankings differ from share rankings?
Texas led the reported state volumes for both gas and coal. Rhode Island led the reported gas share, while West Virginia led the reported coal share. The rankings differ because volume measures physical output and share divides that output by the particular state’s generation total.
Figure 3 in text
VOLUME VERSUS STATE SHARE · 2025 Large output is not the highest share Different 2025 leaders for reported gas and coal volume versus share. Natural gas · 50 eligible state/DC geographies Output, 0 to 300 TWh State share, 0 to 100%
Texas 290.25 48.30%
Rhode Island 8.66 88.26% Coal · 47 eligible state/DC geographies Output, 0 to 300 TWh State share, 0 to 100%
Texas 68.89 11.46%
West Virginia 46.27 86.67% Source: EIA, Sep 2026. Numeric reported fuel rows only. MWh denominators vary by state. Hawaii has no gas row; omitted fuel rows are not assigned zero.
Additional phone figure text: Natural gas · 50 eligible geographies · Share · Coal · 47 eligible geographies
Source and data: EIA state generation workbook
Permanent link to this figure’s text and data · Sources and context
Who led gas volume and who led gas share?
For gas, the comparison contains numeric fuel observations for 50 of the 51 state/DC geographies. Hawaii has no gas row in the selected 2025 producer series, so it is not assigned a fabricated zero or included in the fuel ranking.
- Texas generated 290.25 TWh from gas and ranked first for volume, but fifteenth for share at 48.30%. (EIA, state generation workbook, Sep 2026).
- Florida generated 195.70 TWh from gas and ranked second by volume, with a 72.95% share. (EIA, state generation workbook, Sep 2026).
- Rhode Island’s gas share was 88.26%, highest among the 50 numeric reporting geographies. (EIA, state generation workbook, Sep 2026).
- Rhode Island generated 8.66 TWh from gas, ranking thirty sixth by volume. (EIA, state generation workbook, Sep 2026).
The source supports “Rhode Island had the highest reported gas share” within this defined comparison. It does not support converting an omitted fuel row into a verified zero. Stating the eligible comparison set is more useful than hiding the distinction in an unexplained footnote. The full ranking carries its reporting geography count beside each result, preserves the national comparator without a state rank and keeps the MWh denominator visible for independent reproduction. (EIA, state generation workbook, Sep 2026).
Who led coal volume and who led coal share?
The coal ranking contains 47 of the 51 state/DC geographies with numeric source rows. The source includes some literal zero rows; those are observations and remain in the comparison. Omitted rows have a different status and are left out.
- Texas generated 68.89 TWh from coal, ranking first by volume and twenty fifth by share at 11.46%. (EIA, state generation workbook, Sep 2026).
- Indiana generated 49.22 TWh from coal, with a 46.05% state share. (EIA, state generation workbook, Sep 2026).
- West Virginia generated 46.27 TWh from coal, ranking fourth by volume and first by share at 86.67%. (EIA, state generation workbook, Sep 2026).
- Kentucky’s reported coal share was 65.43%, followed by Missouri at 64.30%. (EIA, state generation workbook, Sep 2026).
DC, Hawaii, Oregon and Vermont have no coal row in the selected 2025 producer series. This is an observed source omission, not a conclusion that these places used no coal-generated electricity or had no coal plants under every possible definition. Imports, contractual supply and generation are different questions. The source does explicitly report zero coal generation for Connecticut, which is retained as zero. A complete, honest comparison preserves both kinds of record instead of making an apparently fuller table by changing their meanings. (EIA, state generation workbook, Sep 2026).
How should the ranks and denominators be quoted?
Every fuel rank should name its measure and eligible set. A position in a volume ranking cannot be reused as a position in a share ranking, and a fuel-specific source set cannot be presented as an unqualified ranking of all states.
- Pennsylvania ranked third by gas volume at 143.12 TWh, but seventh by gas share at 58.25%. (EIA, state generation workbook, Sep 2026).
- California ranked fifth by gas volume at 74.29 TWh, but twenty fifth by gas share at 36.22%. (EIA, state generation workbook, Sep 2026).
- Ohio’s gas share was 54.21%, while its coal share was 24.60%. (EIA, state generation workbook, Sep 2026).
- Shares equal reported fuel MWh divided by the published same-year, same-geography Total row, multiplied by 100. (EIA, state generation workbook, Sep 2026).
The ranking uses competition positions, one plus the count of strictly larger values. Literal ties share a rank. Public percentages are rounded for readability, while the download retains six decimal shares and whole source MWh. A high displayed share is not a separate physical amount that can be added to another state’s share. To aggregate generation, add compatible volumes and use a matched total denominator rather than taking an unweighted mean of percentages from states of very different sizes. (EIA, state generation workbook, Sep 2026).
How do nuclear, wind and solar comparisons change the picture?
The volume-versus-share distinction also changes nuclear, wind and solar comparisons. Illinois led reported nuclear volume, while New Hampshire led nuclear share. Texas led reported wind and solar volumes, while Iowa led wind share and Nevada led the reported solar share.
Figure 4 in text
OTHER FUELS · SAME DENOMINATOR RULE Nuclear, wind and solar shift the leaders Each fuel’s largest output state versus its highest reported share state. Nuclear · 28 eligible Volume leader: Illinois 50.92% Share leader: New Hampshire 57.08% Wind · 44 eligible Volume leader: Texas 21.65% Share leader: Iowa 58.48% Solar (reported) · 50 eligible Volume leader: Texas 10.99% Share leader: Nevada 30.79% 0% 20% 40% 60% Source: EIA, Sep 2026. Share of each state’s published total. Solar is the reported generating- facility category; separate small solar estimates are excluded.
Source and data: EIA state generation workbook
Permanent link to this figure’s text and data · Sources and context
How do nuclear volume and share compare?
Nuclear has numeric source rows for 28 of the 51 state/DC geographies in the selected 2025 industry series. Illinois had the largest reported nuclear volume, while New Hampshire had a larger nuclear fraction of its smaller total.
- Illinois generated 97.67 TWh from nuclear, ranking first by reported state volume. (EIA, state generation workbook, Sep 2026).
- Illinois’s nuclear share was 50.92%, the third highest in the eligible set. (EIA, state generation workbook, Sep 2026).
- New Hampshire’s nuclear generation was 10.91 TWh, but its 57.08% share was highest. (EIA, state generation workbook, Sep 2026).
- South Carolina’s nuclear share was 54.52%, second highest, with 56.67 TWh of output. (EIA, state generation workbook, Sep 2026).
A share does not reveal installed reactor capacity or operating performance per unit of capacity. Capacity factor requires generation, available capacity and time alignment. This workbook alone does not provide all of those matching inputs for a defensible capacity factor calculation. The comparison is therefore kept to observed output and its share. The source’s nuclear category is also kept separate from a custom low-carbon grouping, because that would introduce a new classification and potential emissions claim not required by the underlying question. (EIA, state generation workbook, Sep 2026).
Why can Texas lead wind volume while Iowa leads wind share?
Texas’s large all-source generation total means that a very large wind volume can still represent a smaller fraction than wind does in another state. The eligible comparison contains 44 of the 51 state/DC geographies with numeric reported wind rows.
- Texas generated 130.14 TWh from wind, the largest reported state amount. (EIA, state generation workbook, Sep 2026).
- Texas’s wind share was 21.65%, ranking twelfth in the eligible set. (EIA, state generation workbook, Sep 2026).
- Iowa generated 43.57 TWh from wind, second by volume but first by share at 58.48%. (EIA, state generation workbook, Sep 2026).
- South Dakota’s wind share was 58.21%, while Kansas’s was 47.14%. (EIA, state generation workbook, Sep 2026).
The complete table shows both quantities because the phrases wind producer and wind dependent can otherwise be used as if they meant the same thing. Even the latter phrase needs care: the annual generation share does not establish the electricity mix delivered to consumers or the source available at a particular hour. Annual totals also do not identify curtailment, peak demand contributions or backup requirements. Those are legitimate reporting questions, but this source does not settle them through a single state share. (EIA, state generation workbook, Sep 2026).
What does the reported solar category show?
The solar ranking has numeric source rows for 50 of the 51 state/DC geographies. It uses the workbook’s Solar Thermal and Photovoltaic category consistently, rather than mixing generating-facility output with separately estimated small systems in selected states.
- Texas generated 66.06 TWh in the reported solar category, the largest state volume. (EIA, state generation workbook, Sep 2026).
- California generated 54.89 TWh, with a 26.76% reported share. (EIA, state generation workbook, Sep 2026).
- Nevada’s 30.79% share was highest in the eligible set, with 13.91 TWh. (EIA, state generation workbook, Sep 2026).
- DC’s reported solar share was 18.35%, ranking third by share but forty eighth by volume. (EIA, state generation workbook, Sep 2026).
DC’s contrasting positions are an especially clear warning against interpreting a high share as a large national contribution. Its denominator is small. A rooftop solar story needs the relevant small solar series and a consistent combined denominator, not this facility series alone. EIA’s technical notes explain that small solar generation is estimated using capacity and other survey information. AMR does not treat unadded estimates as measured zero; it specifies the narrower source scope for every solar comparison here. (EIA, state generation workbook, Sep 2026).
How did gas and coal generation change from 2015 to 2025?
National gas share rose from 32.72% in 2015 to 40.93% in 2025. Coal share fell from 33.16% to 16.61%. Those are changes in fractions of a changing total, so the accompanying output volumes and state comparisons are essential to their interpretation.
Figure 5 in text
COMPLETE NATIONAL HISTORY Gas and coal shares across eleven years National reported shares, 2015 to 2025, one historical workbook edition. Gas: 32.72% to 40.93% Coal: 33.16% to 16.61% 0% 10% 20% 30% 40% 50% 2015 2020 2025 Share of each year’s national Total Source: EIA, Sep 2026. One historical source edition, 2015 through 2025. Descriptive generation shares; no causal substitution claim.
Source and data: EIA state generation workbook
Permanent link to this figure’s text and data · Sources and context
How much did national output and shares change?
The fixed endpoint comparison uses 2015 and 2025 observations from the same retrieved historical edition. Each endpoint share divides by its own year’s national all-source total, avoiding a constant-denominator assumption. Changes are calculated before rounding; displayed endpoints and changes are rounded independently.
- National gas output increased from 1,334.67 TWh in 2015 to 1,811.00 TWh in 2025, an increase of 35.69%. (EIA, state generation workbook, Sep 2026).
- National gas share increased 8.203604 percentage points over those endpoints. (EIA, state generation workbook, Sep 2026).
- National coal output decreased from 1,352.40 TWh to 734.93 TWh, a decrease of 45.66%. (EIA, state generation workbook, Sep 2026).
- National coal share decreased 16.549021 percentage points, from 33.157468% to 16.608446%. (EIA, state generation workbook, Sep 2026).
A percentage point change subtracts two percentages. It is not the same as the percentage change in output, and it is not the percentage change in the share itself. The dataset provides the endpoint MWh, shares and share-point changes as separate fields. The national all-source total rose from 4,078.71 TWh in 2015 to 4,425.03 TWh in 2025. That changing denominator is one reason a source’s physical output trend and its share trend can tell different stories. (EIA, state generation workbook, Sep 2026).
Did all state gas shares move in the national direction?
No. The matched gas comparison includes 50 of the 51 state/DC geographies with numeric gas rows at both endpoints. Most had a higher gas share, but ten had a lower share, including Texas despite its higher physical gas generation.
- 40 of the 50 matched gas shares increased between 2015 and 2025; ten decreased. (EIA, state generation workbook, Sep 2026).
- Pennsylvania’s gas share increased from 27.711940% to 58.246683%, a rise of 30.534744 percentage points. (EIA, state generation workbook, Sep 2026).
- Ohio’s gas share increased from 23.740475% to 54.210729%, a rise of 30.470254 percentage points. (EIA, state generation workbook, Sep 2026).
- Texas’s gas share fell from 52.849602% to 48.295220%, while its gas output rose from 237.73 TWh to 290.25 TWh. (EIA, state generation workbook, Sep 2026).
Texas is a useful counterexample to the assumption that more gas generation always means a larger gas share. Its all-source denominator changed enough for the share to fall. This observation does not identify which technologies displaced which, nor does the national comparison establish a causal substitution between gas and coal. Power plant additions, retirements, dispatch, demand and other factors require separate evidence. The matched panel supports a transparent description of observed output and fractions, with no explanatory variable invented to complete the story. (EIA, state generation workbook, Sep 2026).
What does the matched coal comparison establish?
The coal endpoint comparison includes 47 of the 51 state/DC geographies with numeric rows in both years. Its eligible set is defined by actual source records, while literal zero observations remain in the data rather than being dropped as if they were missing.
- 46 of the 47 matched coal shares declined from 2015 to 2025; one was unchanged. (EIA, state generation workbook, Sep 2026).
- Indiana’s coal share fell from 75.207877% to 46.047576%, a decline of 29.160301 percentage points. (EIA, state generation workbook, Sep 2026).
- Kentucky’s coal share fell from 86.925111% to 65.427039%, a decline of 21.498072 percentage points. (EIA, state generation workbook, Sep 2026).
- West Virginia’s coal share fell from 94.136302% to 86.668186%, a decline of 7.468117 percentage points. (EIA, state generation workbook, Sep 2026).
This describes the matched reported series, not a census of coal plant closures or a count of workers affected. The workbook has no employee, ownership, royalty or retirement fields, so those outcomes cannot be derived from a decline in generation share. The year-by-year national table is also retained instead of presenting endpoints as if the path between them were continuous and uniform. Using one historical vintage makes the arithmetic reproducible, while later producer revisions would require a new pinned edition and explicit comparison. (EIA, state generation workbook, Sep 2026).
What do the source boundaries and accounting details change?
The total industry category, reporting threshold, numeric statuses and original national total determine what these statistics mean. Keeping them explicit prevents a complete-looking table from silently inventing zeros, duplicating producer categories or turning generation into a claim about consumption, capacity or affordability.
Figure 6 in text
SOURCE STATUS · 2025 Read the record before ranking Positive, zero, omitted and negative are different source observations. Positive output Texas · Natural gas
290,245,569 MWh Use the numeric row in output and share comparisons. Literal zero Connecticut · Coal
0 MWh Retain as a numeric reported observation. Omitted source row Hawaii · Natural gas
No row in this selected 2025 source Do not assign zero or include in the gas rank. Negative net value U.S. · Pumped storage
−5,361,958 MWh Keep the sign in generation accounting. Source: EIA, Sep 2026, selected Total Electric Power Industry rows. Omission finding concerns this source only. Negative net generation is not missing data.
Source and data: EIA state generation workbook
Permanent link to this figure’s text and data · Sources and context
Which producers and solar systems belong in this series?
The analysis selects only Total Electric Power Industry rows. The workbook separately lists utilities, independent generators and several combined heat and power groups, so adding every producer category would duplicate the total rather than extend its coverage.
- The selected national industry total is 4,425,028,526 MWh in 2025. (EIA, state generation workbook, Sep 2026).
- The source separately reports 137,728,117 MWh for Combined Heat and Power, Industrial Power and 15,339,977 MWh for Combined Heat and Power, Commercial Power. (EIA, state generation workbook, Sep 2026).
The EIA923 reporting instructions specify grid-connected plants with at least one MW of total generator nameplate capacity. (EIA923 reporting instructions).
- Separate small solar estimates are not added to the selected workbook’s fuel rows or denominator. (EIA, state generation workbook, Sep 2026).
Total Electric Power Industry should not be casually renamed the electric power sector, because industrial and commercial combined heat and power belong in the selected broader total. The workbook’s separate producer rows make that inclusion inspectable. The official survey instructions also distinguish generation from fuel consumption and describe eligible facilities. For context on account spending rather than output, use the separate residential electricity bill reference, which uses EIA861 revenue and customer records rather than these generating-plant data. (EIA, state generation workbook, Sep 2026).
How are zero, omitted and negative values treated?
Source status is preserved before any ranking or arithmetic. A number can be positive, explicitly zero or negative; a row that the workbook does not supply is a different kind of evidence and is not materialized as a numeric observation.
- Connecticut’s selected 2025 coal row explicitly reports 0 MWh, retained as a measured source value. (EIA, state generation workbook, Sep 2026).
- Hawaii has no selected 2025 Natural Gas row in the reviewed workbook. (EIA, state generation workbook, Sep 2026).
- The national Pumped Storage row explicitly reports negative 5,361,958 MWh. (EIA, state generation workbook, Sep 2026).
- Across the 572 geography-year reconciliations, the largest absolute reported-fuel sum minus published Total difference is 2 MWh. (EIA, state generation workbook, Sep 2026).
The source-row ledger defines the scope of the omission finding: one selected producer category, one annual edition, one geography and one fuel label. It does not establish a universal absence outside that source. Negative values are included in fuel-sum checks rather than clipped, and every residual is kept in MWh. Percentages are calculated against the published total even when rounded fuel values sum slightly differently. There is no invented Other adjustment to force an exact 100% display. (EIA, state generation workbook, Sep 2026).
How can writers reuse the findings without changing the question?
A reusable sentence names the fuel, geography, year, physical output or share and the comparison population. Specific finding cards below connect those facts to exact fragments, matching CSVs and EIA’s original workbook.
- A gas-share ranking compares 50 numeric reported geographies in 2025, while the coal-share ranking compares 47. (EIA, state generation workbook, Sep 2026).
- The complete total and largest reported fuel comparison contains all 50 states and DC. (EIA, state generation workbook, Sep 2026).
- The 2015 to 2025 endpoint comparisons use matched numeric observations from one source vintage. (EIA, state generation workbook, Sep 2026).
One MWh is 1,000 kWh; one TWh is one million MWh, while MW describes capacity rather than annual energy. (EIA, generation and capacity definitions).
Do not recast a generation share as the fuel mix in a household’s bill, a cause of local prices or a measure of reliability. The source has no state import balance, hourly operation, emissions or causal price model. For household hardship, the separate energy insecurity reference uses a different survey population and denominator. Exact source links, readable HTML tables and complete downloads allow another writer to verify the narrower statement that this dataset can actually support. (EIA, state generation workbook, Sep 2026).
Methodology and Sources
AMR selected Total Electric Power Industry rows for 2015 through 2025 from EIA’s September 2026 final annual workbook. All 52 geographies have a positive published Total in each year. Every share divides the reported fuel MWh by the same geography-year Total. Fuel rankings use only numeric reported rows, including literal zero observations; omitted rows are never assigned zero. Competition ranks compare volume and share separately. Endpoint changes require numeric fuel rows in both years. Negative source values remain signed, including pumped storage. Reconciliation retains reported fuel sums, published totals and their residuals. Public TWh and percentages are rounded, while downloads preserve whole MWh and six decimal derived shares. One current vintage supports offline reproduction. EIA reporting instructions and technical notes define the generating-facility scope and separate small solar estimates. (EIA, state generation workbook, Sep 2026).
Glossary
- Net generation: Gross electricity output minus the electricity used to operate the generating facility, allowing a signed negative value in some records.
- MWh: A megawatthour, a unit of electrical energy equal to 1,000 kWh.
- TWh: A terawatthour, equal to one million MWh.
- MW: A megawatt, a unit of power or capacity, not an annual electricity quantity.
- Published Total: The all-source generation denominator supplied by EIA for the selected geography, year and producer category.
- Reported fuel share: A numeric source fuel row divided by the matched published Total, multiplied by 100.
- Largest reported fuel: The source fuel with the greatest reported MWh in a geography-year, which need not exceed half of the total.
- Pumped storage: Storage using electricity to move water before later generation; annual net output can be negative.
- Combined heat and power: A facility producing electricity and useful thermal output from a common energy source, with separate producer categories in the workbook.
- Eligible comparison set: The explicitly counted geographies with numeric source observations needed for the particular ranking or matched endpoint calculation.
Sources
- EIA: Net generation by state, producer type and energy source, final annual historical workbook
- EIA: State annual electricity source catalog and release dates
- EIA: EIA923 detailed data, final 2025 annual release and survey description
- EIA: Form EIA923 reporting instructions, respondent scope and signed net generation
- EIA: Electric Power Monthly technical notes, generating facility data and separate small solar estimates
- EIA: Electricity generation, net generation and capacity definitions