How to Estimate Furnace Gas Usage From Heating Load and Degree Days

Divide the Manual J heat loss by the design temperature difference it was run at, multiply by 24 and by the location's annual heating degree days, then divide by AFUE. For an Atlanta house with a 40,000 Btu/h heat loss at 70°F inside and 25°F outside, that's about 568 therms a year at 95% AFUE on standard 65°F degree days, and 296 therms if the house balances at 57°F.

That's ASHRAE's degree-day method, the same one its Handbook uses for a Denver house. The heat loss can come from any Manual J, including the free furnace size calculator, which prints the design temperature it used with the result. The base temperature nearly halves the answer in Atlanta, and the sections below show why.

How does the degree-day method turn heat loss into fuel?

It multiplies the house's heat loss coefficient by the degree-hours in a year, then divides by the furnace efficiency. ASHRAE writes the heating energy as degree-days times 24 times the overall heat loss coefficient in Btu/h per °F, covering conduction through every surface plus infiltration. A Manual J heat loss divided by its design temperature difference gives that coefficient, with infiltration at its design-day rate and any duct loss the load includes.

ASHRAE Handbook - Fundamentals 2017 § Chapter 19, 2.1 Degree-Day Method

"In the degree-day method, the building heat load (i.e., the amount of heating energy input or cooling energy extraction) is estimated as the number of degree-days times 24 (to convert to degree-hours), times the overall building heat loss coefficient Btu/h · °F." - ASHRAE, handbook.ashrae.org

The same section adds that "the heating or cooling loads must be divided by the efficiencies of the HVAC system." Put together, with heat loss Q at design and design difference dT:

1. Heat loss coefficient. UA = Q / dT, in Btu/h per °F.
2. Heat delivered in a year. UA x 24 x HDD, in Btu.
3. Fuel burned. Heat delivered / AFUE, in Btu.
4. Therms. Fuel / 100,000.

ASHRAE's Example 1 runs a 1,800 sq ft Denver house with 5,940 degree-days, a surface UA of 400 Btu/h per °F, 0.7 air changes per hour in 14,400 cu ft, a 78% AFUE furnace and a 10% duct loss. The infiltration adds 0.7 x 14,400 x 0.018 = 181.4 Btu/h per °F, and the Handbook's answer is 118.1 million Btu of fuel a year. The same steps reproduce it: 5,940 x 24 x 581.4 / (0.78 x 0.9) = 118.1 million Btu, or about 1,181 therms.

Which design temperature difference do you divide by?

The one the heat loss was calculated at. When every component of a Manual J heat loss is figured at the same design difference, as on the free calculator, dividing by that difference gives back UA no matter which design temperature was used. Mix them, say a load run at an approximate zone value divided by a station's difference, and the coefficient comes out wrong by the ratio of the two.

On the free calculator, the heating load is transmission plus infiltration plus attic duct loss, all figured at 70°F indoors minus the 99% heating design dry-bulb for the ZIP code. The calculator page states that every component uses that same difference, and the result line prints the design temperature it used. That design value is an approximate estimate for the ZIP's IECC climate zone, not a station value, but it doesn't bias UA as long as you divide by the difference the load used.

Here's the example house we'll carry through. Its Manual J heat loss is 40,000 Btu/h at 70°F indoors and 25°F outdoors, the 99% heating limit ENERGY STAR lists for Fulton County, Georgia. That's a 45°F difference, so UA = 40,000 / 45 = 888.9 Btu/h per °F.

If you divided the same 40,000 by a station's milder 50°F difference by mistake, UA would drop to 800 and every therm figure below would come out 10% low.

How many degree days does your area get?

Use NOAA's 1991-2020 climate normals for the nearest airport station. At the standard 65°F base, Atlanta Hartsfield averages 2,531 heating degree days a year, Dallas-Fort Worth 2,220, Sacramento Executive 2,436, Houston Intercontinental 1,266 and Phoenix Sky Harbor 874. The same files give lower bases too, and at 57°F those stations drop to 1,318, 1,144, 1,023, 510 and 211.

NOAA publishes annual totals for each station as long-term averages, at a 65°F base and at 60, 57, 55, 50, 45 and 40°F.

Here's what the example house's UA of 888.9 burns at 95% AFUE at each of those five stations.

All five stations carry NOAA's "S" completeness flag, defined as meeting "WMO standards for data availability for 24 or more years." These are averages over 1991 to 2020. A colder winter than normal burns more, so compare an estimate with an average year of bills, not last January alone.

Why does the base temperature move the estimate so much?

Because the 65°F base assumes only 5°F of free heat from the sun and internal gains at a 70°F setpoint. ASHRAE says better-insulated, tighter houses can balance lower and puts the average for modern houses at 55 to 57°F. At the five stations in the table above, moving the base from 65 to 57°F removes 48% of the degree days in Atlanta and Dallas and 76% in Phoenix.

ASHRAE Handbook - Fundamentals 2017 § Chapter 19, Variable-Base Degree-Day Method

"The average balance point temperature for modern residential buildings is now 55 to 57°F." - ASHRAE, handbook.ashrae.org

The balance point is the outdoor temperature where the house's own gains match its losses. Above it the furnace doesn't run. Below it, ASHRAE says, the house needs heat "in proportion to the difference from the base temperature," which is why UA x 24 x HDD applies at a lower base the same way, as long as the base is the house's balance point.

ASHRAE's 55 to 57°F is an average across modern houses, and a given house can sit above or below it. A house with a small UA and the same gains balances lower than a leaky one, and ASHRAE notes the balance point also swings between day and night. So read the 65°F and 57°F results as two reference points. For the example house in Atlanta at 95% AFUE, that's 568 and 296 therms a year, and at 55°F, the low end of ASHRAE's modern-house average, it's 243.

The method has other blind spots, and ASHRAE lists them. It ignores solar heat gain and thermal mass, can't follow changes in infiltration, and doesn't handle a night setback. The Handbook calls it a quick answer that works "as a starting point or check for more detailed calculations," and gives "light-construction residential buildings with low solar gain in cold climates" as an example of where it may be adequate. The five stations above are mild.

What does a higher AFUE save?

The fuel scales with 1 / AFUE, so going from 80% to 95% cuts gas for heating by 1 - 0.80 / 0.95, about 15.8%, whatever the climate. For the example house in Atlanta that's 675 down to 568 therms at the 65°F base, or 351 down to 296 at 57°F. Multiply the therms saved by the rate on the customer's gas bill for the dollar figure.

Two things to check before quoting the savings:

1. Ducts. ASHRAE's example divides by AFUE and then by 0.9 for a 10% duct loss, so the two are separate. If your Manual J heat loss already includes duct loss, as the free calculator's does for attic ducts, don't divide by a duct factor again. If it doesn't and the ducts run outside the conditioned space, add the loss one way or the other.
2. Other gas loads. A water heater, range or dryer on the same meter burns gas all year. Compare the heating estimate with the bill only after taking those out.

The free calculator doesn't ask for AFUE. It sizes the input bin with 80%, 90% or 95% depending on year built, so for a fuel estimate use the AFUE of the furnace actually installed or quoted.

Running it for a real house

1. Get the heat loss. Run the furnace size calculator with the ZIP code and the house details, and note the heating load and the heating design temperature on the result line.
2. Find UA. Divide the load by 70 minus that design temperature.
3. Pick the degree days. Take the nearest NOAA station's 1991-2020 normals at 65°F and at 57°F.
4. Compute both results. UA x 24 x HDD / AFUE / 100,000 gives therms at each base.
5. Check it against the bills. Take the water heater, range and dryer out of an average year's gas use, then see which base the heating gas lands closer to.

The free calculator gives the whole-house heating load without an account. For the cooling side of an equipment change-out, the SEER2 energy cost calculator runs the electric bill comparison. Design temperatures by ZIP code are on the design temperature page.

Sources

• ASHRAE, 2017 ASHRAE Handbook - Fundamentals (I-P), Chapter 19, Energy Estimating and Modeling Methods, section 2.1 Degree-Day Method, Example 1 and Variable-Base Degree-Day Method: handbook.ashrae.org
• NOAA National Centers for Environmental Information, U.S. Climate Normals 1991-2020, annual and seasonal station files for USW00013874, USW00003927, USW00023232, USW00012960 and USW00023183: ncei.noaa.gov
• NOAA NCEI, Normals Annual/Seasonal Documentation 1991-2020 and Normals Calculation Methodology 2020 (completeness flags): ncei.noaa.gov
• U.S. Energy Information Administration, Glossary, Therm: eia.gov
• U.S. EPA, ENERGY STAR Certified Homes Design Temperature Limit Reference Guide, revised September 1, 2018, Exhibit 1 (Fulton County, Georgia): energystar.gov

For estimation purposes only. Not a substitute for a licensed engineer, and not ACCA-approved for permit submission.