CertifiedSpecs

How to work out heat pump running cost from HSPF2

The arithmetic is two lines long and you can do it on a phone. The hard part is knowing what the answer means, which is where most running-cost estimates — including the ones in sales literature — quietly go wrong.

The formula

HSPF2 is BTU of heat delivered per watt-hour of electricity consumed. A million BTU is 1,000,000 BTU, and a kilowatt-hour is 1,000 watt-hours, so:

kWh per million BTU = 1,000 ÷ HSPF2

cost per million BTU = kWh per million BTU × rate in $/kWh

annual cost = annual heat demand in MMBtu × 1,000 ÷ HSPF2 × rate

The 1,000,000 and the 1,000 cancel down to a single 1,000, which is why the formula looks almost too simple.

Worked through with numbers

Take a system certified at HSPF2 9.0 and an electricity rate of 17.0¢/kWh, which is close to the U.S. residential average.

Electricity per million BTU
1,000 ÷ 9.0 = 111.1 kWh
Cost per million BTU
111.1 × $0.17 = $18.89
Over 40 million BTU of delivered heat
4,444 kWh, costing $756

A quick sanity check you can use on any of these: at HSPF2 10 exactly, a million BTU takes 100 kWh, so at 20¢/kWh it costs $20. If a calculator disagrees with that, something is wrong with the calculator.

What the same demand costs at a different rating

Cost moves inversely with efficiency, so the saving from a higher rating is never proportional to the rating increase. Going from HSPF2 9.0 to 10.5 — a 17% better rating — takes the same 40 million BTU from $756 to $648, a saving of $108, or 14%.

Three places this goes wrong

1. Confusing energy consumed with heat delivered

This formula wants heat delivered into the house, in million BTU. Published household averages usually report energy consumed for space heating, which is a different quantity: a gas furnace consuming 50 million BTU delivers rather less than 50 million BTU of heat, and how much less depends on its efficiency.

You cannot convert a national average of consumption into delivered heat by applying one efficiency, because those averages mix homes with gas, electric resistance and heat pump systems together. That is why this site leaves the demand field empty and labels its example as a hypothetical rather than offering a "typical" figure it cannot defend.

The figure actually worth using is a Manual J load calculation from a contractor, or your own metered consumption converted with the efficiency of the system that produced it.

2. Treating HSPF2 as a prediction

HSPF2 comes from a standardised laboratory procedure at reference conditions. Since 2023 that procedure is AHRI 210/240-2023 Appendix M1, which raised the external static pressure used in testing from 0.10 to 0.50 inches of water column to sit closer to real duct systems. That change made the numbers more realistic, and it is also why SEER2 and HSPF2 figures read lower than the older SEER and HSPF for the same equipment — typically around 4–5% lower for SEER.

More realistic is not the same as predictive. Your climate, how the system is sized, duct leakage, thermostat behaviour and the quality of the commissioning all move real consumption. Use HSPF2 to compare two systems fairly under identical assumptions, which is exactly what it is good for.

3. Using the wrong HSPF2 for the system being quoted

The largest error is usually not in the arithmetic at all. It is using the outdoor unit's best certified rating when the quote pairs it with a different coil. 44.3% of certified outdoor models carry more than one HSPF2, so the rating you plug in has to be the one on the AHRI certificate for your combination. Why that happens, and how to get the right number.

What this does not cover

Run this with your own rate and model

Related

Sources

By Adrián Criado · About the author and this dataset · Updated 2026-09-26