Heat Pump Design Temperature: Verify Output and Size Backup Heat

Three numbers settle whether a proposed heat pump covers your house in winter. Your 99% winter design temperature. Your Manual J design heating load in BTU/h. The delivered heating capacity of the exact indoor plus outdoor combination at that temperature. Divide the third by the second and you have your answer.

Almost every quote gives you none of the three. It gives you a brand, a tonnage, and a marketing claim about how low the unit will run. This page publishes the design temperatures, shows where the capacity number lives, and walks the arithmetic with a real published capacity table so you can do the check yourself before you sign.

We have not installed or measured any of these systems. Every figure below is transcribed from the published documents listed under Sources, and each one carries the date we checked it.

The five steps, in order

1. Find your 99% heating design temperature. Use the table below, or look up the nearest station in ACCA's Outdoor Design Conditions Guide. This is the outdoor temperature your area stays above for 99% of annual hours, roughly 88 hours a year colder.
2. Get the Manual J design heating load. A room-by-room load calculation at that design temperature, holding 70°F indoors. Not a square-footage rule. See our Manual J and sizing guide.
3. Get both model numbers and the AHRI certificate number. The outdoor unit alone does not have a capacity. The certified rating belongs to the pair, and in ducted systems sometimes to the pair plus a furnace.
4. Pull the expanded performance table for that pair. Manufacturer submittal sheet or the NEEP ccASHP product list. Read delivered BTU/h at your design temperature, at the airflow the installer plans to set.
5. Divide capacity by load. NEEP's sizing tool calls 90 to 120 percent the band for whole-home heating without backup. Below that, the gap is what backup heat has to cover.

Step 1: your 99% heating design temperature

The values below are the 99% heating dry-bulb temperatures published in ACCA Manual J 8th Edition v2.0, Table 1A, which is the table Manual J requires a designer to use. Every row names the specific weather station the number belongs to, because the station matters more than the city name. Transcribed from ACCA's Outdoor Design Conditions Guide on August 16, 2026.

ACCA states the rule plainly in the same document. Under Manual J don'ts, listed as a mandatory requirement: do not add a safety factor to the Table 1A design conditions. A contractor who designs to a colder number than the table gives is not being careful, they are inflating the load, and an inflated load is how a system ends up two sizes too big.

The station matters more than the city

Table 1A carries several stations for most metro areas, and the spread inside one metro is often larger than the difference between two states. These are all published Table 1A values, same source and same check date.

Fifteen degrees separate Logan from Pittsfield, eleven separate JFK from Poughkeepsie, and four separate two Denver airports twenty miles apart.

What that costs in equipment, using the 3-ton Daikin capacity curve published further down this page and a 42,000 BTU/h design load in both houses. At the Logan design temperature of 12°F the system delivers 39,660 BTU/h, which is 94 percent of the load and needs no backup at all. At the Pittsfield design temperature of -3°F the same system delivers 28,440 BTU/h, which is 68 percent, a shortfall of 13,560 BTU/h, or 4 kW of electric strips and the breaker to feed them. Same house, same equipment, different station.

Two caveats. Manual J Note 6 says Table 1B micro-climate data supersedes Table 1A for Arizona, California, Hawaii and Nevada, because coastal and inland values inside one county diverge too far for a single station to describe. That is why no metro from those four states appears in the table above; ask your designer for the Table 1B entry instead. And ENERGY STAR's county-level reference guide, which certification programs use as a limit, assigns each county the lowest heating design temperature among all stations inside it, so a county figure is deliberately colder than most addresses in that county.

Step 2 and 3: the load, and the two model numbers

The load has to come from a room-by-room Manual J at the design temperature you just looked up, holding 70°F indoors. NEEP's own guidance for its sizing tool is blunt about what happens next: resist adding in extraneous safety factors to avoid oversizing.

Then get both model numbers. An outdoor unit on its own does not have a certified capacity. NEEP describes its database this way: each individual listing is its own AHRI certificate number representing a unique combination of outdoor unit, indoor unit, and in some cases paired indoor gas furnace.

How much does the indoor match move the number? Bosch's own IDS Ultra specification, document BTC 762008303 B dated January 2026, publishes AHRI 210/240 ratings for the same 3-ton outdoor unit against different indoor units:

Same outdoor unit, a 1,800 BTU/h spread and a full point of HSPF2. That is why the quote needs both models and the AHRI certificate number, not a brand and a tonnage. Look the number up yourself in the AHRI Directory of Certified Product Performance.

Step 4: three claims that get confused

Manufacturers publish three different cold-weather numbers, and quotes routinely quote the weakest one as though it answered the question.

The gap between rows two and three is where whole-home proposals fail. A system can hold 100 percent of rated capacity at 5°F and still be 78 percent at -5°F and 55 percent at -10°F, which is exactly what the Daikin table shows.

Bosch shows the same tension inside a single document. The IDS Ultra feature list says the unit is capable of providing 100% heating capacity at 5°F outdoor temperature. The extended heating table in that same specification, for the 3-ton BOVB19-36 plus BIVB19-36 pair at 1,000 CFM and 70°F indoor, lists 34,800 BTU/h at 47°F and 32,500 BTU/h at 7°F. Both statements can be true at different rating conditions. Bosch does not state which condition the feature bullet refers to, so ask, and get the answer against your airflow.

A published capacity curve, read the way you should read yours

This is the Daikin FIT AURORA DH9VS expanded heating data, normal heating mode, from submittal SS-DH9VS-R32, checked August 16, 2026. Every cell is BTU/h delivered at that outdoor temperature by that specific matched pair.

Daikin prints these model numbers with a trailing wildcard character covering finish and revision variants, for example DH9VSA361CA*. The capacity figures are the same across those variants. Your proposal should still carry the full model number as the installer will order it.

Two things in that table are worth more than any brochure line. Output at 17°F is higher than at 47°F on every one of the four sizes, because the published maximum shifts to a much higher compressor speed below about 40°F: input power on the 3-ton pair goes from 2.27 kW at 40°F to 5.22 kW at 35°F in the same table. And the fall past 5°F is steep. The 3-ton pair loses 45 percent of its 5°F output by -10°F. Neither pattern is visible in a SEER2 or HSPF2 number.

Published COP for the same 3-ton pair falls in step: 4.34 at 47°F, 2.77 at 17°F, 2.40 at 5°F, 2.15 at -5°F, 1.98 at -10°F. Even the worst of those beats electric resistance backup, which sits at 1.0. For the general shape of that curve across equipment classes, see heat pump efficiency by outdoor temperature.

Heat Pump Guide

Want a second read on the quote?

Paste the proposal or upload the PDF. No lead auction.

Check my quote

Free. No sales pitch. Personal reply in ~2 business days.

Still researching? Send me the guide

Step 5: the arithmetic, worked all the way through

House: Portland, Maine. Table 1A design temperature 4°F at Portland International Jetport. Manual J design heating load 42,000 BTU/h at 4°F, holding 70°F indoors.

Proposed system: Daikin DH9VSA361CA outdoor unit with the CAHEA3630 coil and MBVK16CH air handler, the 3-ton row above.

Capacity at 4°F. The table publishes 34,200 BTU/h at 5°F and 31,200 BTU/h at 0°F. That is 600 BTU/h per degree, so at 4°F the pair delivers 33,600 BTU/h.

Percent of design load served. 33,600 divided by 42,000 is 80 percent. NEEP's tool calls 90 to 120 percent the band for whole-home heating. Eighty percent is below it, so this system needs backup at design conditions.

Shortfall. 42,000 minus 33,600 is 8,400 BTU/h.

Backup in kW. Electric resistance heat delivers 3,412 BTU/h per kW. 8,400 divided by 3,412 is 2.5 kW.

Balance point. Draw the load line the way NEEP's tool does, linear from 42,000 BTU/h at 4°F down to zero load at 60°F. That is 750 BTU/h per degree. Where the load line crosses the published capacity line is the capacity balance point.

So the compressor carries this house alone down to about 9.5°F, and below that backup fills a gap that reaches 8,400 BTU/h at the design temperature and 22,150 BTU/h if the weather goes nine degrees past it.

The alternative worth pricing. Move to the 4-ton DH9VSA4810A, and capacity at 4°F rises to 42,300 BTU/h, which is 101 percent of the load and inside NEEP's band with no strips at all. The tradeoffs are real and also in the published data: COP at 5°F drops from 2.40 to 1.90, and rated maximum cooling rises from 34,200 to 46,000 BTU/h, which is a lot of cooling capacity for a Maine summer and a familiar route to short cycling and poor humidity control. That is the actual decision, and it is a numbers decision, not a brand decision.

Sizing the backup, if there is a gap

Convert once and the rest is arithmetic. Electric resistance heat delivers 3,412 BTU/h per kW, and draws kW times 1,000 divided by 240 amps on a 240V circuit.

One trap worth naming: heat kits are rated at 240V, and resistance output scales with the square of voltage. On a 208V supply a 240V-rated 5 kW element delivers about 75 percent of its nameplate, roughly 12,800 BTU/h. If your service is 208V, say so and get the kit sized against that.

Then pick the strategy against the gap you calculated.

Whichever route the proposal takes, the electrical scope has to be priced, not implied. Ten kW of strips is 41.7 A of new load, which is where a panel upgrade stops being hypothetical. Compare strips against keeping a furnace in the backup heat guide.

Send this to every bidder

Copy it as written. It asks for exactly the evidence this page uses and nothing more.

Please send the AHRI certificate number for the exact outdoor and indoor combination you are proposing, plus the page of the manufacturer's expanded performance data showing delivered heating capacity in BTU/h at [your design temperature]°F for that combination, at the airflow you intend to set. Also state the design heating load your Manual J produced and the design temperature and weather station you used. If the system does not cover the load at that temperature, state the shortfall in BTU/h, the backup heat in kW, the breaker and conductor size, and the aux lockout temperature.

A bidder who can answer that in a day is running a real design. A bidder who sends back a brochure page with a minimum operating temperature on it has told you what you needed to know.

What we could not verify

Publishing the gaps is part of the method.

• Which ASHRAE edition your designer's software uses. ACCA's own Note 4 says Table 1A uses the 1997 ASHRAE 1% and 99% data where it was available for a location. NEEP's sizing tool assigns the 99% heating dry-bulb from the 2021 ASHRAE Handbook of Fundamentals. We did not reconcile the two station by station and we are not publishing a claim that they agree. Expect small differences, ask which source produced the number on your proposal, and treat a several-degree gap as a question rather than an error.
• Hours per winter below the balance point. None of the documents we retrieved publishes an hour count per station. NEEP's tool derives percent of annual load served from station temperature-bin data, so run your own case there rather than trusting a rule of thumb. We are not publishing an hours figure we cannot source.
• The condition behind Bosch's 100 percent at 5°F claim. Not published in the specification we read. Reported above alongside the extended-table numbers, not resolved.
• Whether your county's assigned value fits your address. ENERGY STAR's county reference guide assigns the coldest station in a county to the whole county by design. Useful as a ceiling for certification, deliberately conservative as a description of your site.
• Anything about the systems in the field. We have not installed, tested, commissioned or measured any equipment named on this page, and we publish no performance data of our own. Every number here came from the documents under Sources.

Red flags in the answer you get back

• The proposal names a brand and a tonnage but no indoor model number and no AHRI certificate number.
• The cold-weather evidence is a minimum operating temperature rather than delivered BTU/h at your design temperature.
• A single-zone performance claim is reused for a multi-zone outdoor unit.
• The designer used a colder design temperature than Table 1A publishes for the nearest station, and calls it a safety margin.
• Backup heat appears as a line item with no kW, no breaker size and no lockout temperature.
• The capacity number quoted is the 47°F rating, presented as though it applied in January.
• Nobody will say which weather station the load calculation used.

For the full proposal checklist, work through how to read a heat pump quote. For a shortlist of systems that publish usable low-temperature data, see best cold-climate heat pumps, and for the ducted pair used in the examples above, Daikin FIT versus Bosch IDS.

Frequently Asked Questions

What is my heat pump design temperature?

It is the 99% heating dry-bulb temperature for the weather station nearest you, the value Manual J Table 1A requires a designer to use. Your area stays above it for 99 percent of annual hours. Find your metro in the table above, or look up the station in ACCA's Outdoor Design Conditions Guide. Boston Logan is 12°F, Portland Maine is 4°F, Minneapolis is -8°F, Denver is 7°F, Seattle is 29°F.

How do I know if a heat pump is big enough for my house?

Divide the delivered capacity of the exact matched system at your design temperature by the Manual J design heating load. NEEP's sizing tool treats 90 to 120 percent as the band for whole-home heating without backup. Below 90 percent you have a shortfall to cover, above 120 percent you are buying capacity you will not use and inviting short cycling.

Is the capacity on the spec sheet the capacity I get?

Only if the spec sheet row matches your system. Capacity is certified for a pair, not for an outdoor unit. Bosch's own AHRI table shows the same 3-ton outdoor unit rated at 36,600 BTU/h with one air handler and 34,800 BTU/h with a cased coil. Airflow and indoor temperature move it again. Ask for the expanded performance data at the airflow the installer will actually set.

How much backup heat do I need?

Take the shortfall in BTU/h at your design temperature and divide by 3,412 to get kW. The worked example above has an 8,400 BTU/h shortfall, which is 2.5 kW. Then check that the electrical service carries it: 10 kW of strips is 41.7 A of new load at 240V.

What is the balance point and where should it be?

It is the outdoor temperature at which the heat pump's maximum capacity equals the house load, so below it the system needs help. There is no universally right value. What matters is that the proposal states it, that the aux lockout is set consistently with it, and that the backup sized for the design temperature is actually priced.

Does an ENERGY STAR or NEEP cold-climate listing prove the system covers my house?

No. NEEP's ccASHP specification, version 4.0 effective January 1, 2023, screens equipment for IECC climate zone 4 and higher and adds test points below 17°F. That is an equipment screen. It does not calculate your load, choose your indoor unit, set your airflow or commission the install. Some listed products carry no published data below 5°F at all, which NEEP flags in its own tool.

Where to go next

1. Run your own case in the heat pump sizing calculator, then get a real Manual J before signing anything.
2. Maine buyers should read the design temperature next to the money in Efficiency Maine heat pump rebates; Massachusetts buyers, Massachusetts heat pump cost and rebates.
3. When the proposal is written and you have the capacity table in hand, send it for a free quote review before the deposit.

Sources

All documents retrieved and read on August 16, 2026. Manufacturer submittals describe equipment capability at stated rating conditions; they do not replace the AHRI certificate, NEEP listing or expanded performance page for your specific proposed match.

• ACCA Outdoor Design Conditions Guide, the stand-alone publication of Manual J 8th Edition v2.0 Table 1A
• NEEP, User Guide: Cold Climate Heat Pump Sizing Support Tools, February 2024
• NEEP ccASHP specification and product list, version 4.0
• NEEP cold-climate air-source heat pump product list database
• Daikin FIT AURORA DH9VS submittal, SS-DH9VS-R32
• Bosch IDS Ultra condensing unit product specification, BTC 762008303 B, January 2026
• ENERGY STAR Certified Homes County-Level Design Temperature Reference Guide
• PNNL Building America Solution Center, cold climate heat pump sizing and selection
• AHRI Directory of Certified Product Performance