How to Compare Heating Costs by Fuel — From Your Own Bills, Not Averages
General information only. Fuel prices, delivery charges, and taxes change constantly and differ by address. The worked examples below use round placeholder prices to show the method — replace them with the all-in rates from your own bills before drawing any conclusion about your home.
"Gas is cheaper than electric" is the most repeated sentence in home heating, and it is meaningless as stated — because nobody buys heat. You buy therms, kilowatt-hours, gallons, and cords, and those units contain wildly different amounts of heat. Comparing a price per therm with a price per kilowatt-hour is comparing the price of a crate with the price of a bottle. The U.S. Energy Information Administration says it flatly: comparing prices of fuels in dissimilar units is not meaningful. The comparison only exists once every fuel is converted to the same thing — the cost of a million Btu of heat, delivered into your rooms.
The method, in three steps
- Find your all-in price per unit. Take a winter bill and divide the total by the units consumed, so delivery charges, riders, and taxes are inside the rate. A commodity-only rate understates every fuel, but it understates delivered fuels and electricity unevenly, which corrupts the ranking. Do this for each fuel you are comparing, from your own documents.
- Convert to cost per million Btu purchased. The EIA formula: price per unit, divided by the fuel's heat content per unit, multiplied by one million. The standard heat contents used in extension-service comparison tables are: natural gas, 100,000 Btu per therm; heating oil, 138,500 Btu per gallon; propane, 92,500 Btu per gallon; electricity, 3,412 Btu per kilowatt-hour.
- Divide by your appliance's efficiency. Purchased heat is not delivered heat. A gas furnace rated 90 percent AFUE delivers nine-tenths of its therms into the house. Resistance electric heat delivers effectively all of its energy — an efficiency of 100 percent that still loses, as the arithmetic shows, because its purchased Btu cost so much more. A heat pump is the special case: it moves heat rather than making it, so its effective efficiency is above 100 percent — a coefficient of performance of 3 delivers three units of heat per unit of electricity — which is the only reason it can undercut gas.
What success looks like: a single ranked list, in dollars per million Btu delivered, computed from your own winter bills and your own equipment ratings — a list no national article can hand you, because no national article has your bills.
A worked example with placeholder prices
Say a winter gas bill works out to $1.50 per therm all-in, electricity to $0.20 per kilowatt-hour all-in, and oil to $4.00 per gallon. The arithmetic, step by step:
| Fuel and appliance | Purchased $/MMBtu | Efficiency | Delivered $/MMBtu |
|---|---|---|---|
| Natural gas, 90% AFUE furnace | $1.50 / 100,000 x 1,000,000 = $15.00 | 90% | About $16.67 |
| Electricity, resistance baseboards | $0.20 / 3,412 x 1,000,000 = About $58.62 | 100% | About $58.62 |
| Electricity, heat pump at COP 3 | About $58.62 | 300% effective | About $19.54 |
| Heating oil, 85% furnace | $4.00 / 138,500 x 1,000,000 = About $28.88 | 85% | About $33.98 |
The table demonstrates the two lessons the method exists to teach. First, efficiency is not a footnote: the same electricity is the most expensive heat on the page through baseboards and the second-cheapest through a heat pump. Second, rankings are local. Halve the gas price or double the electricity price — both entirely plausible at different addresses — and the order shuffles. This is why comparison calculators tell you to replace their defaults with the rates on your own bills: the defaults describe someone else's house. Extension tables make the same point in shortcut form, pricing delivered heat directly — natural gas at 90 percent efficiency costs roughly the per-therm price multiplied by 12.5 per million Btu, while resistance electricity costs the per-kWh price multiplied by 293 — a factor gap that no amount of shopping changes, because it is physics, not retail.
What the per-MMBtu ranking does not decide
The EIA itself lists the factors standing beside fuel cost: the efficiency of the appliance, the cost of installing the system (including any credits or rebates), operation and maintenance, the availability of the fuel during the heating season, and environmental considerations. Delivered-heat cost ranks fuels for a home that already has the equipment. It does not price the new furnace, the heat pump's cold-weather capacity, the oil tank's remaining life, or the value of a system that also cools in summer. Use the ranking to answer the question it can answer — what does a unit of heat cost me, fuel by fuel, today — and let the equipment decision weigh the rest openly rather than pretending one number settles it.
Keeping the comparison honest over time
Redo the arithmetic each autumn, from that season's first full winter-style bill, and keep the dated results. Fuel prices move at different speeds — oil with the delivery market, electricity with regulated rate cases, gas with both commodity and distribution charges — so last year's ranking is a hypothesis, not a fact. For homes with two systems, typically a heat pump with a backup furnace, the dated comparison also sets the changeover point: the outdoor temperature or the price movement at which the backup becomes the cheaper heat source is a number you can compute rather than a habit you inherit. Households that run this calculation yearly rarely discover they have been heating with their most expensive fuel out of routine — which is, in the end, the expensive outcome the whole method exists to prevent.
Canadian bills: gigajoules, cubic metres, and the same method
Canadian readers meet the identical problem in different clothing. Natural gas may be billed per gigajoule or per cubic metre rather than per therm, and electricity per kilowatt-hour with delivery and regulatory charges itemised separately. None of that changes the method — it only changes the heat-content line in step two. One gigajoule is about 948,000 Btu, so a per-gigajoule all-in rate converts to a per-million-Btu cost almost directly. A cubic metre of natural gas contains roughly 35,000 to 38,000 Btu depending on the gas supply, and utilities print the conversion factor they actually bill with on the statement or in its notes, which is the figure to use rather than a textbook value. Electricity needs no conversion at all: the kilowatt-hour is 3,412 Btu in every province. The discipline that survives every border and billing format is the same three moves — all-in price from a real winter bill, conversion to a common heat unit, division by the efficiency of the equipment in your own basement. A household that can state its delivered cost per million Btu for each of its fuels knows more about its heating economics than most comparison websites do, because those sites are answering for a house that is not theirs.
Questions homeowners ask about fuel costs
Why does my neighbour swear oil is cheaper than gas? Possibly, at their address, with their equipment, it currently is — or their comparison stops at the sticker price per unit, which the EIA warns is not meaningful. Run both homes through the same three steps before arguing.
Should I use the rate printed as the "price" on my bill? Use the all-in rate: total divided by units. Fixed monthly charges belong to every fuel alike, but per-unit delivery and rider charges change the ranking and are part of what a unit of heat actually costs you.
Does a heat pump always beat a gas furnace? No. It wins when its effective efficiency overcomes electricity's high per-Btu price at your rates — the COP your climate actually delivers across the season, not the brochure figure. The worked example shows the mechanism; your bills supply the verdict.