What Size Heat Pump Do You Need?

Two contractors look at the same house and quote different tonnages. One did a load calculation, the other multiplied floor area by a number they have used for twenty years. Both quotes look equally confident on paper, and the difference between them is usually money you either spend on capacity you cannot use or save by asking one question.

What size heat pump do you need?

Size means capacity: how much heat the equipment can move per hour, in BTU/h. Quotes often express it in tons instead. One ton equals 12,000 BTU/h, a convention inherited from refrigeration, so a 3-ton unit is a 36,000 BTU/h unit. Tons say nothing about efficiency — that is SEER2 and HSPF2, a separate question.

Two numbers have to meet. The building has a load: heat lost on a cold design day, heat gained on a hot one. The equipment has a capacity that changes with outdoor temperature. Sizing matches one to the other in both heating and cooling, and the two rarely land on the same tonnage. No universal answer exists, because the load belongs to your house, not to its floor area.

Heat pump size by square footage: what it can and cannot tell you

We do not publish a lookup table you could size a system from, and the reason shows up fast. Take one 1,800 sq ft house and run it through three per-square-foot assumptions.

The spread is 32,400 BTU/h wide, nearly three tons. The hypothetical Manual J for this house, worked through below, returns a 30,000 BTU/h heating load — one number inside that band. The middle assumption would have specified 39,600 BTU/h, about 32% high. A rule that can miss by a ton either way flags an implausible quote; it cannot tell you what to install.

Why square footage alone is not enough

Floor area is one input among many, and not the strongest. These factors move the answer, and each is worth asking about.

Manual J, Manual S and Manual D

Three ACCA standards do three different jobs, and quotes frequently name the first while skipping the second.

• Manual J — residential load calculation, to the national ANSI-approved standard. It produces the design heating load and the design cooling load.
• Manual S — residential equipment selection. It covers eleven equipment types and explains how to use the manufacturer's performance data. This is the step that turns a load into a specific model, and the one most often missing.
• Manual D — residential duct design. ANSI-approved standards for duct sizing and airflow. It matters whenever ducts are added, replaced, or asked to carry more air than before.

Building America is explicit that equipment should not be selected on nameplate capacity: the capacity at design conditions has to be determined instead. A load calculation with no equipment selection behind it leaves that open.

Heating load vs heat pump capacity

A heat pump's headline capacity is rated at 47°F outdoors. Your design heating load happens at a much lower temperature, and capacity falls as the outdoor temperature falls, because there is less heat outside to move. The rated number and the number that matters are not the same number, and the gap widens the colder your design day.

For cooling-led selection, Building America documents a window: the manufacturer's listed capacity at design conditions should sit between 95% and 115% of the design cooling load, widening to 95% to 125% for heat pumps in IECC climate zones 4 through 8. That window is written against capacity at design conditions, not against the nameplate.

Cold-climate sizing

Design temperature, not the annual average, is what sizing runs on: an average conceals the cold snap the system has to survive. The companion idea is the balance point, the lowest outdoor temperature at which the heat pump still supplies the whole heating load alone. Below it, backup heat covers the difference. Building America sets out four approaches, and knowing which one a contractor used explains most disagreements between quotes.

1. Size to the design cooling load. A cold-climate model is not required, but backup heat carries a large share of winter.
2. Choose a variable-capacity unit whose mid or low cooling range meets the design cooling load, raising heating output without oversizing cooling.
3. Size to a stated fraction of the design heating load — 80% is the example given — or to a target balance point, with backup for the rest.
4. Size to the full design heating load at the location's design heating temperature, minimizing or eliminating backup heat.

The first three all lean on auxiliary heat in deep winter. Electric resistance backup runs at a COP of 1, so it costs about three times as much per BTU as a heat pump running at a COP of 3 — how often it runs is a running-cost question, not only a comfort one. Building America adds that upsizing or rounding up beyond the calculated load is not needed, and that a unit whose modulating range sits above the actual load will cycle rather than modulate.

One published benchmark is worth carrying into a showroom. To earn the ENERGY STAR cold-climate designation, a model's heating capacity at 5°F must reach at least 70% of its rated capacity at 47°F, with a COP of at least 1.75 at 5°F. Those are certification thresholds, not a promise about any particular unit and no substitute for its performance table. Many models clear them comfortably; the model's own table is the only way to know.

A worked example

What happens if a heat pump is too large?

You pay up front for capacity you never use. A single-stage unit also reaches the setpoint before taking much moisture out of the air — Building America describes it shutting off before it has had time to remove humidity, causing high humidity problems in humid climates. Short cycling adds starts and noise, and an oversized unit can demand more airflow than the existing ducts deliver.

A variable-speed unit softens this by modulating down instead of cycling, and it does not remove the problem. Every inverter has a minimum output, and where that minimum sits above your typical load it cycles anyway. Modest oversizing is not automatic failure. It is a cost with no offsetting benefit.

What happens if a heat pump is too small?

It runs long, leans on backup heat, and on the coldest days may not hold the setpoint at all. Where the backup is electric resistance, the bill shows it at roughly three times the cost per BTU.

How to check a contractor's heat pump sizing

For ductless systems, add one more. Building America directs that the load calculation be run for individual zones, one per indoor unit, as well as for the whole house: head capacities cannot be added up and matched to an outdoor unit, and an oversized or badly located head short-cycles just as a central system does.

Red flags in a heat pump quote

• The size rests on floor area, or on a per-square-foot figure with no calculation behind it.
• A safety margin was added on top of the calculated load, unexplained.
• A larger unit is recommended for cold weather, with no capacity figure at the design temperature to justify it.
• Tonnage and efficiency are used interchangeably, as though a bigger unit were a more efficient one.

How sizing changes cost and payback

Sizing reaches your money three times: the purchase price, any duct or electrical work a larger unit forces, and the running cost through how often backup heat gets called. Right-sizing has no universal payback — the saving depends on what the alternative quote was.

The running-cost side reduces to one expression: annual operating cost = annual electricity use × electricity rate. Your own rate is what matters. EIA's U.S. residential average was 18.44¢/kWh in May 2026, while state averages that month ran from 12.35¢ in Idaho to 52.00¢ in Hawaii — a four-fold spread that swamps most sizing decisions. Take your effective rate off the bill: total dollars divided by total kWh.

What to do next

One question does most of the work: what does this model deliver at our design temperature, and what fraction of our design heating load is that? A contractor who has done the sizing answers it from the manufacturer's table.