Start from the house, not the climate

Every estimate models the same reference house: 2000 square feet, with a heat-loss coefficient of 0.3 Btu/hr per square foot per °F of indoor-outdoor temperature difference — typical for an existing US home, not a new build. That fixes how much heat the house loses at any given outdoor temperature, before climate ever enters the calculation.

Degree days, corrected

Annual heating demand comes from heating degree days (base 65°F) at Minneapolis, MN, where 6022 HDD a year drives a real heating load, versus Phoenix, AZ at just 1212. Raw degree-day totals overstate real heating, because a house gains free heat from people, appliances, and sunlight, so the model applies a 0.6 correction factor on the heating side and 0.9 on cooling, where the same internal gains work against you instead of for you.

Sizing off the cold snap, not the average day

Equipment gets sized against the 99% winter design temperature — the cold snap it must still keep up with, not the seasonal mean. That figure sets both the system size and, through the COP curve, how efficiently it runs; see the cold-climate units guide for how that curve changes below 25°F. Los Angeles, CA and San Diego, CA both sit in the mild range where a standard unit holds up fine.

What is actually being replaced

The existing furnace is modelled at 80% AFUE (older, non-condensing) and the existing central AC at SEER 10; the new heat pump is priced at the federal SEER2 15.2 minimum. The decision-relevant cost is INCREMENTAL: a whole-home heat pump installs for about $13,000 (plus a $2,500 premium where a cold-climate unit is called for), against a like-for-like replacement furnace at about $4,500 and a new AC at about $5,500 — because almost nobody rips out a working furnace, they replace one that already died.

The tax credit and the payback window

IRA §25C applies a 30% credit, capped at $2,000 a year, to the gross equipment cost before the incremental comparison. See the tax credit guide for how much that moves the answer. Every city page also reports a 15-year net position, because a payback that technically clears past the equipment's own working life — covered in the guide to where a heat pump never pays back — is not a real recommendation.

Where the raw numbers come from

Climate data is NOAA's published normals for the nearest weather station to each city. Electricity and gas rates are EIA's state-level residential tables. Neither is modelled or estimated — both are looked up per city and cited on that city's page, alongside the electric-resistance comparison for homes with no gas service at all.