Energy Data Literacy

How to Read the EIA Winter Fuels Outlook

The page block by block, four fuels, and the checks that tell you whether a number is current.

Last updated: October 23, 2026 · Educational analysis, not financial advice

The Winter Fuels Outlook is the EIA’s once-a-year forecast of what an average household will use and spend on its main heating fuel from November to March. The 2026-27 edition, released October 6, 2026, has heating oil homes at $2,115 (up 21 percent), electric homes at $1,196 (up 4 percent), propane homes at $1,246 (down 3 percent) and natural gas homes at $640 (down 9 percent). The numbers are only useful if you know what they count.

Two facts about the page change how much weight those figures can bear. Last winter’s heating oil forecast finished 26 percent above where it started in October, so a first-release number is a starting point, not a verdict. And as checked on October 22 and again on October 23, 2026, some narrative text and one scenario table on the live page still carried the previous year’s release, which a reader skimming for a quote could easily pick up. The tables themselves hold together: price times quantity reproduces each spending figure, which is the check to run before quoting any of them. For what the numbers mean in a home, see this site’s guides to heating oil prices, how much heating oil a house uses and the cost of each fuel per MMBtu.

What the outlook is, and what its authors say it is not

The outlook is a supplement to the Short-Term Energy Outlook (STEO), the EIA’s monthly forecast, which is covered in this site’s STEO guide. The October STEO supplies the retail price and weather forecasts; the winter supplement converts them to household consumption and spending using the EIA’s Residential Energy Consumption Survey (RECS), specifically the 2020 edition. Each winter the EIA re-runs the tables as each new STEO appears and records the successive values in a tracker.

The agency is direct about the limits. The averages are best interpreted as a broad guide to compare the upcoming winter with the previous one. Differences in the housing populations complicate comparisons across fuels, and readers should not assume that switching to another fuel with lower energy expenditures would necessarily lower their own. That one warning, which is on the page, disqualifies the most common use of the headline table: ranking fuels by dollars per household and concluding which is cheapest to use. A better comparison is per unit of heat, which is why this site handles it in a separate analysis.

Four scope rules that apply to every table

1. Main heating fuel only: a home that heats with oil and runs an electric space heater counts in the oil table and the electric use is outside it. 2. All uses of that fuel: water heating, cooking and drying are inside the main-fuel totals. 3. Nominal dollars, with about 2 percent inflation assumed. 4. Averages across households that already use the fuel, so the sample differs by fuel.

The page, section by section

The EIA page has a stable order. The table lists each block, what it holds and what it is for.

Table 1. Layout of the 2026-27 outlook page, as checked October 23, 2026

BlockWhat it holdsUse it to
Forecast updates trackerSuccessive releases of the previous winter, monthly spending by fuel, degree daysSee how much a forecast moved before it was final
What is the outlookScope, fuels covered, regions covered, two weather side casesConfirm what the forecast does and does not include
Forecast summaryOne paragraph per fuel with the base-case directionGet the headline and the stated drivers
Weather assumptionsHeating degree days by region and month; definition of the unitCheck whether a forecast winter is colder or milder than the last
Household characteristicsHouseholds by main heating fuel and region, with changeLearn which fuel serves how many homes and where
Energy expendituresSpending, price and consumption by month, with regional detailRead the identity price times quantity and compare regions
Wholesale to retailHow fast each fuel passes wholesale moves to the homeJudge whether a market move has reached bills yet
Consumption method and inflationRECS basis, degree-day method, nominal valuesKnow what the numbers are built from
Comparisons across fuelsCaveats on ranking fuels by spendingAvoid the cross-fuel mistake
Winter resourcesLinks to weekly EIA data seriesFind the data that updates faster than the outlook

The order runs from the question (what changed) to the evidence (weather, households, spending) to the machinery (pass-through, method). Reading it that way, rather than from the top headline, makes it easier to spot a figure that does not belong with its neighbors, a skill that matters for the stale text discussed further below.

The identity that makes the tables checkable

Every fuel block publishes three columns that multiply: price, consumption and spending. That means any block can be audited with a calculator. Spending should equal price times consumption, within rounding, and the percentage change in spending should be close to the compound of the other two. Here are the four U.S. blocks for 2026-27.

Table 2. Price times quantity for each fuel, U.S. base case, November to March 2026-27

FuelPriceQuantityPrice times quantityPublished spendingCompound of the two percent changes
Heating oil$5.26 per gallon (+34%)402 gallons (-9%)$2,115$2,115 (+21%)21.9%
Electricity18.23 cents per kWh (+3%)6,562 kWh (+1%)$1,196$1,196 (+4%)4.0%
Propane$2.52 per gallon (-2%)494 gallons (-1%)$1,245$1,246 (-3%)-3.0%
Natural gas$13.44 per Mcf (-9%)48 Mcf (0%)$645$640 (-9%)-9.0%

Source: EIA, Winter Fuels Outlook 2026-27, U.S. column for each fuel (checked October 23, 2026). The last two computed columns are this page's arithmetic.

Heating oil passes: 402 gallons at $5.26 is $2,115 against $2,115 published, and 1.34 times 0.91 is a 22 percent increase against 21 percent published, a difference of rounding in the percentages. Propane passes ($1,245 against $1,246). Electricity passes. Natural gas comes out at $645 against a published $640, a 0.8 percent gap, which is within the rounding of a 48 Mcf figure that is itself rounded to a whole number. The check is not about catching EIA errors. It is about learning the units. The natural gas consumption column on the page is labeled “million cubic feet” in the header, but the prose calls the unit Mcf, and the arithmetic only works if the 48 is thousand cubic feet. If you ever read 48 million cubic feet per household, multiply it out and notice the bill would be $645,000. The gas units guide covers the conversions in full.

The same identity explains why spending and price can move in opposite directions. In this outlook the propane price falls 2 percent and consumption falls 1 percent, so spending falls about 3 percent. In the West, natural gas prices fall 8 percent but a 25 percent colder winter (heating degree days) lifts consumption by 17 percent and spending rises 7 percent. Quantity is the weather signal and price is the market signal, and the page reports them separately so you can tell which one moved your fuel.

Weather: the unit behind every quantity

Quantity in the outlook is driven by heating degree days (HDD). The EIA defines a heating degree day against a 65 degree Fahrenheit base, computed from the day’s average temperature, and weights regional values by population. Its own example: an average of 50 degrees gives 15 degree days, and thirty such days give 450 for the month. Space heating consumption is estimated as proportional to degree days, using the relationship between weather and use in the survey data. Because the forecast uses a trend of the previous 30 years, a base-case winter is a typical winter, not a prediction of this one.

Table 3. Heating degree days, base case 2026-27, by region

RegionNovemberDecemberJanuaryFebruaryMarchSeasonVersus last winter
Northeast6339171,0719147954,331-12%
Midwest7261,0621,2109978064,802-2%
South2954775424102972,021+1%
West4626636445384742,781+25%
United States4797147916445263,1550%

Source: EIA, Winter Fuels Outlook 2026-27, weather assumptions (checked October 23, 2026).

The regional spread is the point. Midwest degree days of 4,802 are more than double the South’s 2,021, so a Midwest propane home needs more than twice the heat of a Southern one before any difference in fuel prices. The Northeast forecast is 12 percent milder than last winter, which is why Northeast consumption falls across fuels. The West forecast is 25 percent colder, which is why West electricity (up 9 percent) and gas (up 7 percent) buck the national direction. Always read the weather line before a price line.

The EIA also includes two side cases that move all regions’ degree days 10 percent colder or 10 percent warmer. A caution on that table appears below, because on the page as checked it did not match the 2026-27 base case.

Who is in each table: the household counts

The household table is the denominator that tells you how large each fuel group is. For 2026 the EIA counts the following homes by primary heating fuel, in thousands, with the change from 2025.

Table 4. Households by main heating fuel, 2026 (thousands)

FuelNortheastMidwestSouthWestUnited StatesChange from 2025Share of the total
Natural gas12,39318,07715,02415,46860,9620%45.4%
Electricity4,5387,24534,88811,20857,880+1%43.1%
Propane1,2952,1911,9019746,3610%4.7%
Heating oil3,3881744111614,134-5%3.1%
Wood3453873415491,622-4%1.2%
Other or none6394159011,4383,392+6%2.5%

Source: EIA, Winter Fuels Outlook 2026-27, household characteristics (checked October 23, 2026). The share column divides each row by the six-row total of 134.4 million households, this page's arithmetic.

Natural gas (60.96 million homes, 45 percent) and electricity (57.88 million, 43 percent) dominate; propane serves 6.36 million, oil 4.13 million and wood 1.62 million. Two readings follow. First, the national headline for oil, up 21 percent, applies to 3 percent of homes, and 82 percent of those homes (3.388 million of 4.134 million) are in the Northeast. Second, electricity serves 34.9 million homes in the South alone, 60 percent of the national electric-heating total, in the region with the fewest degree days. That is why the electric spending average is so different between the South ($1,079) and the Northeast ($1,637). The EIA notes that in all-electric homes space heating is a smaller share of consumption, especially in the mild South.

The regional tables: where the average hides the gap

The same block gives regional detail. A short selection shows the gaps worth knowing before using a national number.

Table 5. Regional spending and price, 2026-27 base case

Fuel and regionSpendingChangePriceConsumption
Natural gas, Northeast$879-14%$16.12 per Mcf55 Mcf
Natural gas, Midwest$613-10%$10.38 per Mcf59 Mcf
Natural gas, South$511-15%$14.13 per Mcf36 Mcf
Natural gas, West$590+7%$15.10 per Mcf39 Mcf
Electricity, Northeast$1,637-3%26.19 cents per kWh6,252 kWh
Electricity, South$1,079+3%15.75 cents per kWh6,849 kWh
Propane, Northeast$1,687-15%$3.28 per gallon515 gallons
Propane, Midwest$1,305-4%$2.04 per gallon641 gallons

Source: EIA, Winter Fuels Outlook 2026-27, regional columns (checked October 23, 2026).

Look at the pairs. Natural gas costs $16.12 per Mcf in the Northeast and $10.38 in the Midwest, a 55 percent premium, but the Midwest home uses 59 Mcf against the Northeast’s 55, because the Midwest is colder. Propane in the Midwest averages just over $2 a gallon against $3.28 in the Northeast, while the Midwest home burns 641 gallons against 515. Spending is the product, so the regions land closer together than prices do. Treat a regional price as a feature of local supply, and a regional quantity as a feature of local weather and housing.

How far a forecast moves: the 2025-26 tracker

The top of the page keeps a record of earlier releases for the previous winter. It is an uncommon piece of evidence because it shows the EIA’s own forecast error in public. Reading the seven release rows for the winter that just ended (November 2025 to March 2026):

Table 6. Winter 2025-26 outlook by release: average household spending by main fuel

ReleaseHeating oilNatural gasPropaneElectricityU.S. degree days
October 15, 2025$1,390$642$1,210$1,1333,153
November 12, 2025$1,429$654$1,293$1,1343,153
December 9, 2025$1,453$671$1,291$1,1443,194
January 13, 2026$1,467$670$1,297$1,1353,166
February 10, 2026$1,650$688$1,344$1,1593,327
March 10, 2026$1,749$696$1,334$1,1573,273
April 7, 2026 (final estimate)$1,749$686$1,286$1,1463,140
October to final+25.8%+6.9%+6.3%+1.1%-0.4%

Source: EIA, Winter Fuels Outlook 2026-27 page, forecast updates tracker for the 2025-26 winter (checked October 23, 2026). The release label follows each block on the page. The last row is this page's percent change from the first to the final column.

Two things stand out. Heating oil moved most: the final estimate is about 26 percent above the October forecast, while natural gas moved 6.9 percent, propane 6.3 percent and electricity 1.1 percent. And season-total weather barely moved at all, 3,153 heating degree days in the October release against 3,140 in the final estimate. Since spending equals price times quantity and the quantity side is built from degree days, a change that large with little change in season weather came from the price side or from where the weather fell within the winter. The February and March releases show season totals of 3,327 and 3,273 degree days before the final estimate of 3,140, but the page publishes the totals without a reason for each revision.

What to take from this: a heating oil forecast in October is a snapshot of prices at that date. The December and January releases moved little, then the February release lifted the forecast by $183 in one month. For a household, the useful behavior is to treat the October number as a starting estimate and compare it with the tracker for any later release before making a decision that depends on it. Such a decision is not one this site can make for anyone, and none of the tracker implies what this winter will do.

A caution: stale text on the page

When this page was checked on October 22 and 23, 2026, the tables for 2026-27 and the narrative text under them did not agree. The heating oil paragraph said the EIA expected prices to average about $3.50 per gallon, 4 percent below the previous winter, and household spending of $1,390. The table above it says $5.26 (up 34 percent) and $2,115 (up 21 percent). The natural gas paragraph described expenditures close to last winter’s level, while the table has a 9 percent drop. The three-case table, which shows base, 10 percent colder and 10 percent warmer, listed a natural gas base of $642 (down 1 percent).

The older figures match the first row of the tracker, the October 15, 2025 release for the winter that ended in March: $642 for natural gas, $1,390 for heating oil, down 1 and 8 percent. The most likely reading is that the narrative and the three-case table had not been refreshed for the new year, and the EIA may have corrected them since. Whatever the cause, the lesson is the same. When a page carries more than one vintage, trust the numbers you can tie out. Here the tables satisfy the price times quantity identity, and the EIA’s Today in Energy article of October 7, 2026 independently reports heating oil expenditures up 21 percent, natural gas down about 9 percent, propane down about 3 percent and electricity up about 4 percent, which agrees with the tables. This site’s own analysis uses the tables and the Today in Energy summary, and nothing from the stale paragraphs or the three-case table.

A four-point freshness check for any forecast page

1. Find the release date and the forecast-completed date. 2. Confirm that the narrative numbers match the table numbers for the same fuel. 3. Check that price times quantity equals spending. 4. Look for a second EIA publication on the same release, such as Today in Energy, and compare the percentages.

Using the outlook without misusing it

Compare your own year with last year

The EIA frames the product as a way to compare this winter with the last. The cleanest personal use is therefore to take the percentage changes for your fuel and region, apply them to your own last-winter bill, and treat the result as a direction. A household that spent $2,400 on oil last winter has no reason to expect $2,115 from this outlook; the outlook would imply about 21 percent more, or $2,904, at the EIA’s average weather and price path. That is a scaling exercise, and it carries all of the cautions above.

Pair it with data that updates weekly

The outlook is monthly at best. The EIA’s Heating Oil and Propane Update, which publishes weekly retail prices from October through March, supplies the current price; for the week of October 19, 2026 it shows U.S. residential heating oil at $6.190 a gallon and propane at $2.689. The weekly natural gas storage report and the weekly petroleum status report supply inventory context. This site explains how to read the gas storage report and the weekly petroleum status report, and the release schedule shows when each one lands.

Know the pass-through speed of your fuel

The page explains that natural gas and electricity retail prices lag wholesale prices because of utility regulation, while heating oil and propane track wholesale prices within weeks. That difference decides how much weight a wholesale move deserves. The purchased gas adjustment guide follows a gas cost true-up through a regulator’s order so the lag can be seen in dollars.

Six checks before quoting a figure from the page

  • Ranking fuels by the headline spending. The EIA warns against it. The homes behind each average are different.
  • Reading the national heating oil number as national. It describes 4.1 million homes, mostly in the Northeast, not the 134 million total.
  • Reading an average spending figure as a bill. The average includes water heating, cooking and drying for the same fuel, excludes other fuels, and excludes fuel bought before the winter or under a fixed contract.
  • Skipping the unit labels. The natural gas consumption header says million cubic feet, the prose says Mcf, and the identity says thousand cubic feet.
  • Using an old release. The tracker is there because forecasts change. The heating oil forecast for the previous winter ended 26 percent above its October estimate.
  • Taking a forecast as a prediction. The outlook uses a 30-year weather trend and notes that a strong El Niño pattern could cause large temperature variation. It is a scenario built from a trend, not a call on prices or weather.

Reader questions about the outlook

When is the Winter Fuels Outlook published, and how often does it update?

The 2026-27 outlook was released on October 6, 2026, with the forecast completed on October 1, as a supplement to the October Short-Term Energy Outlook. EIA updates the forecast values with each STEO released during the winter and compares the new values with earlier ones in a tracker on the page. The next STEO release is on November 10, 2026.

Does the outlook tell me what my own bill will be?

No. EIA calls the averages a broad guide for comparing the upcoming winter with the previous one, and says readers should not assume that switching to a fuel with lower average expenditures would lower their own. The averages describe homes that already use each fuel, which differ in size, insulation, equipment and climate. Use the percentage changes and your own bills to estimate direction, not the dollar levels.

What does the outlook count as a household’s fuel use?

It counts consumption of the home’s main space-heating fuel, including other uses of that fuel such as water heating, cooking and clothes drying. It leaves out a second fuel used in secondary equipment unless the secondary equipment burns the same fuel. Electricity totals are the all-electric home’s winter electricity, with an assumed April to October cooling season removed.

Why are the percentages in the table not simply the price change?

Because spending is price times quantity. Heating oil prices are forecast 34 percent higher and consumption 9 percent lower, so spending is about 22 percent higher by multiplication (1.34 times 0.91), against the 21 percent EIA publishes; a one-point difference is within the rounding of the published percentages. The weather forecast moves quantity and the fuel market moves price, and the table reports both so the two can be separated.

Are the dollar figures adjusted for inflation?

No. EIA states that all values are nominal and that it assumes consumer prices rise about 2 percent over the winter, so an increase must exceed roughly 2 percent to be larger than general inflation.

Why did EIA’s heating oil forecast for last winter rise so much after October?

The tracker shows the 2025-26 heating oil forecast moving from $1,390 in the October 2025 release to a final estimate of $1,749, a rise of 25.8 percent, while the winter's total degree days moved from 3,153 to 3,140, a change of -0.4 percent. Season-total weather therefore does not explain the revision, which came from price assumptions and from the monthly pattern. EIA's page does not give a reason release by release.

Data and sources: U.S. Energy Information Administration (EIA), Winter Fuels Outlook 2026-27, released October 6, 2026 (forecast completed October 1, 2026; next release November 10, 2026), checked October 23, 2026: expenditure, price and consumption tables by fuel and region; heating degree day table; households by main heating fuel; methodology and inflation notes; forecast-update tracker for the 2025-26 winter (releases of October 15, 2025 to April 7, 2026); narrative text and the three-case table as discussed above. EIA, Today in Energy, “Mixed outlook for energy expenditures this winter,” October 7, 2026, checked October 23, 2026. EIA, Heating Oil and Propane Update, week of October 19, 2026, checked October 23, 2026. EIA, Glossary: heating degree days, checked October 23, 2026. The identity checks, share-of-total column, percent-change rows and the scaling example are this page’s arithmetic.