How Old Is America’s HVAC Equipment? 20% of Heating Systems Are Past 20 Years

Last reviewed: September 2026 · By the HVAC Brief Editorial Team · Our sourcing and editorial standards

Every page that answers “how long does an HVAC system last” gives a range, usually 15 to 20 years, and none of them cites a source. There is no source to cite: lifespan means age at failure, and no federal survey asks households when their equipment died.

A different question is answerable, and it is the more useful one. EIA asks 17,745 households how old their main heating equipment is, and the same for cooling and water heating. For your own unit, the serial number on the data plate gives the build date; our serial number decoder by brand shows how to read it. That gives the age profile of the entire installed base, which has not been published as a distribution.

The age of what is installed

The single most quoted number in this topic is the 15 to 20 year lifespan. Against the distribution, 32.4% of installed heating equipment is already 15 years or older and 20.0% is past 20. That is 23,537,032 homes running heating equipment older than the upper end of the rule of thumb.

Cooling is a younger fleet. 61.6% of air conditioning equipment is under 10 years old against 48.6% of heating equipment, and the 20-year-plus share is roughly half the heating figure.

It depends almost entirely on what the equipment is

Steam and hot water systems are the outlier and it is not close. 38.1% are 20 years or older, with an interpolated median of about 16 years, against 18.2% and about 10 years for the 74,420,786 homes on a central furnace.

Central heat pumps sit at the other end, 9.6% past 20 years with a median around 8. That is a statement about when heat pumps were installed, not about how long they survive. A technology that grew recently has a young installed base by construction, and this data cannot separate the two effects.

Window and portable units turn over fastest, with medians around 6 and 4 years. They are cheap enough to replace rather than repair, which is the clearest case in this data of price shaping the age of the stock.

Geography, and the Northeast boiler problem

The spread runs from 31.3% in the Middle Atlantic division to 11.2% in the West South Central. The two oldest divisions are the two with the most boilers and the oldest housing, and the pattern follows the equipment mix rather than any difference in maintenance.

Housing vintage matters less than might be expected, and not monotonically: homes built in the 1990s show a higher share of 20-year-old-plus heating equipment than homes built in the 1980s. Equipment age is not housing age, because equipment gets replaced and houses do not.

Homes built from 2010 onward are excluded from that table. Almost none has heating equipment over 20 years old, which leaves the estimate below EIA’s publication threshold rather than at zero.

Frequently asked questions

Methodology and limitations

• Source: EIA Residential Energy Consumption Survey 2020 public use microdata file v7. Heating base is 17,745 sampled households using space heating equipment, weighted to 117,738,794 homes; cooling base is 16,171 households, 109,511,484 homes.
• Variables: EQUIPAGE, ACEQUIPAGE and WHEATAGE, each defined by EIA as the age of the main equipment of that type. The code −2, Not Applicable, is excluded.
• Ages are bracketed, not exact. RECS offers six categories: under 2 years, 2 to 4, 5 to 9, 10 to 14, 15 to 19, and 20 or more. Shares are reported directly from those brackets. Medians are interpolated across bracket boundaries and are approximations; the top bracket is open-ended, so any median falling inside it would not be estimable and none here does.
• Weighting with the final analysis weight NWEIGHT. Standard errors computed by EIA’s documented Jackknife method using the 60 replicate weights, and every figure published here clears EIA’s standards of a relative standard error at or below 50 with at least 10 sampled households. Cells failing either test are omitted and said to be omitted.
• This measures installed age, not age at failure. A distribution of how old working equipment is cannot be converted into a lifespan or a survival curve without knowing the age of the equipment that was removed, which RECS does not record. No lifespan figure is derived here, and any page quoting one is not using federal data.
• Equipment age is self-reported by the householder, who may not have installed it.
• Differences between equipment types partly reflect when each technology was installed rather than how long it lasts. Heat pumps and mini splits have young installed bases in part because they were deployed recently.
• Vintage figures describe the age of equipment in homes of a given construction era, not the original equipment.
• Collected in 2020. The 2024 RECS cycle has released preliminary tables, but air conditioning and space heating detail is scheduled for spring 2027.

Sources

1. US Energy Information Administration, Residential Energy Consumption Survey 2020 public use microdata, variables EQUIPAGE, ACEQUIPAGE, WHEATAGE, EQUIPM and ACEQUIPM_PUB.
2. US Energy Information Administration, Using the Microdata File to Compute Estimates and Standard Errors, June 2023.
3. US Energy Information Administration, RECS 2024 preliminary housing characteristics, for the release schedule of 2024 equipment detail.

The HVAC Brief is an independent trade publication. We do not sell equipment, take manufacturer advertising, or accept payment for placement in our research. Corrections welcome.