GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted
Musings of an Energy Nerd

Calculating Cooling Loads

Part 3 in a series of articles on sizing heating and air conditioning equipment

The sun and high outdoor temperatures lead to external heat gains. Internal heat gains come from people, pets, appliances, and lighting.
Image Credit: IBACOS

A few decades ago, residential air conditioning was very rare in colder areas of the U.S., and cooling load calculations were usually unnecessary. These days, however, new U.S. homes routinely include air conditioning equipment, even in Minnesota, so most U.S. builders are faced with the need to calculate cooling loads.

In my last two blogs (“How to Perform a Heat-Loss Calculation,” Part 1 and Part 2), I discussed the principles behind heat-loss calculations used to size heating equipment. In this blog, I’ll discuss the principles behind cooling-load calculations used to size air-conditioning equipment.

Rule-of-thumb sizing

Although most building codes require load calculations for heating and cooling equipment installed in new homes, the requirement is widely ignored and rarely enforced. Most HVAC contractors never perform cooling load calculations; instead, they size air conditioners by rules of thumb.

The age-old rule of thumb used by most contractors was one ton of cooling equipment for every 400 square feet of conditioned space. In a concession to recent improvements in insulation levels and window specifications, some HVAC contractors have adjusted their rule of thumb, and now size air conditions at one ton per 600 square feet.

Because these rules of thumb almost always result in gross oversizing of cooling equipment, most energy experts have been battling rule-of-thumb sizing for years. However, rules of thumb have their place. Using a rule of thumb is not really the problem; the problem is that HVAC contractors are using a bad rule of thumb.

At least two well-known energy consultants, Michael Blasnik and Allison Bailes, have proposed a new rule of thumb for sizing air conditioners in homes with insulation that meets minimum code requirements: namely, one ton of cooling per 1,000 square feet. According to Blasnik, “Sizing an air conditioner using tons per square foot actually works…

GBA Prime

This article is only available to GBA Prime Members

Sign up for a free trial and get instant access to this article as well as GBA’s complete library of premium articles and construction details.

Start Free Trial


  1. GBA Editor
    Allison A. Bailes III, PhD | | #1

    Great article, Martin! Thanks for all the mentions. I guess I've written about this topic quite a bit.

    I'd like to add a little bit to what you said at the beginning about the rule of thumb I proposed. I don't think it should be used for sizing air conditioners but rather as a way for energy raters, code officials, and builders to do a quick and easy check to see if an AC might be oversized.

    It's also a good way to set a threshold. For example, we could say that in Atlanta, every new home has to be built so that it can be served by an AC that's sized at no more than one ton for each 1000 square feet. If the home is super efficient, it might need only one ton for each 2000 square feet. That's certainly possible because I built one.

    I'm looking forward to next week's installment!

  2. GBA Editor
    Martin Holladay | | #2

    Response to Allison Bailes
    Thanks for the further clarification.

    I agree with you -- which is why I wrote, "A good rule of thumb has many uses; for example, it can be used by builders to get a general idea of whether their HVAC contractor’s sizing method is in the right ballpark or is totally nuts. Of course, using a rule of thumb to size an air conditioner is no substitute for performing a room-by-room cooling load calculation."

  3. Expert Member
    ARMANDO COBO | | #3

    Great summary!!!
    This is your best work that I’ve read in the GBA, Thank you. I will add a few “issues” I commonly see in MJ calculations and HVAC design:
    • Wrong area calculations
    • Assemblies for walls and roof are incorrect. Most HVAC contractors use “standard” assemblies which are usually inferior from my specifications. Too lazy or uneducated to develop a new wall and roof assemblies into the program.
    • Not designing the duct system correctly. Round vs. square ducts, dead corners, lack of turning vanes, wrong duct size, wrong plenum take-off, duct does not match the appliance, wrong fitting tables, too close distance of 1st fitting and better air flow to name a few.
    • Using flex duct. It’s usually installed incorrectly and it reduces the air flow up to 40% form metal duct, not to mention the ducts that are hung to the point it chokes the air flow completely.
    • HVAC equipment and/or ducts in ventilated attic or crawl spaces.

  4. 5C8rvfuWev | | #4

    To size a ductless (e.g., Mr Slim, Fujitsu, et al) is Manual J still used and able to adjust to the absence of ducting?

  5. GBA Editor
    Martin Holladay | | #5

    Response to Joe W.
    Joe W.,
    As far as I know, ductless minisplit AC units are sized according to their output rating (12,000 Btu/h = 1 ton).

    Duct losses only occur when ducts are located outside of a home's thermal envelope. Since ductless minisplit units have no ducts, they have no duct losses (just like ducted systems with all of the ductwork located inside the home's thermal envelope).

    So, yes -- you can use Manual J to calculate the heat gain or cooling load of a home that will be equipped with ductless minisplit units.

  6. user941025 | | #6

    These days, however, new U.S.

    These days, however, new U.S. homes routinely include air conditioning equipment, even in Minnesota, so most U.S. builders are faced with the need to calculate cooling loads.

    Snort. Having lived in a number of cities around the US where people don't seem to understand this fact, I feel I must ask: You do realize it actually does get hot and humid here, Martin, right?

  7. GBA Editor
    Martin Holladay | | #7

    Response to Minneapolis
    I understand two facts:
    1. It gets hot and humid in Minneapolis during the summer.
    2. A few decades back, very few homes in Minneapolis had central air conditioning.

    Times have changed. Lemonade and a ceiling fan aren't enough any more.

  8. gman571 | | #8

    What does right size mean?
    Can someone explain that?
    I understand that someone somewhere said the unit should be running 100% during the hottest day of the year (even reference in this article). I'm skeptical. We had a 100 degree day here in the Boston area last year which isn't supposed to statistically happen (according to Manual J and others). So who decides what the hottest day of the year is?
    Second: What's the financial down side of an oversized unit running on a fractional duty cycle compared to a "right sized unit running at 100%? Presumably the larger unit has a larger compressor (possibly fans, pumps, valves, etc) and will consume more amps while it's running. But shouldn't it be running less often? Or is the delivered cooling not a function of "tonnage"? What's the real delta in kilowatt hours (and hence dollars)? Unless it's ludicrous, I'd rather have an oversized unit to hedge against the statistical temperature anomalies.

  9. GBA Editor
    Martin Holladay | | #9

    Response to Gerard Celentano
    Q. "Who decides what the hottest day of the year is?"

    A. This number is determined from historical weather data. If your designer uses a 97.5% design temperature, that means that the outdoor temperature only exceeds the design temperature for 2.5% of the hours during the hottest month of the year. Believe it or not, this method works.

    Q. "What's the financial down side of an oversized unit?"

    A. It costs more to purchase.

    Q. "What's the real delta in kilowatt hours?"

    A. In most cases, oversized units don't require more energy to operate. But they are expensive to purchase, and in some cases they don't provide efficient dehumidification.

    Q. "I'd rather have an oversized unit to hedge against the statistical temperature anomalies."

    A. Most Americans (but few energy nerds) seem to agree with you.

  10. John_Hansen | | #10

    How do we get HVAC guys to do these calcs?
    Martin, You already know the issue that builders have had for years. Just how does a builder enforce these rules of good engineering without eliminating a very large number of HVAC installers from his pool of bidders? City building departments ask for the manual J calcs, but never check them. Builders require them in their specs, but never really get their favored low bidder to do the calcs. Builders of the new school will align with HVAC contractors who comply, but the changeover to a complete industry did not happen in the last 30-50 years when all this knowledge was available, what must be done now to turn this ship around?
    I did Manual J calcs in the '70s and we knew then that a leaky envelope caused more heat loss than low insulation R-value. We are now 40 years later and I do not see a lot of changes in the average sheet metal man. And builders won't add the cost of the proper calcs to the cost of the house and pay an engineer to do them.
    What do you suggest?

    1. brstina | | #37

      I am building in NW Colorado and my solution was to make my own Manual J calculation. Software is easy and I was able to enter assemblies R values correctly, correct leakage rate and correct window specifications. There were few parameters I was unsure of and I found that they have minimal impact to the required loads when I change them within range I expected. My calculation was 25% less than what my hating contractor came with and less than half from what equipment distributor came up with. It was still 30% higher than what energy rater came with. My research shows that Manual J tables are not ideal for homes that have high levels of insulation as there is too much rounding down when specified insulation level is not found in the tables. This results in lower insulation levels being used for actual calculation by the software and higher loads are calculated.

  11. GBA Editor
    Martin Holladay | | #11

    Response to John Hansen
    I have discussed the problem you talk about -- the fact that most HVAC contractors can't supply a Manual J calculation, in spite of the long-standing code requirements -- in an earlier blog, Saving Energy With Manual J and Manual D.

    I've been mulling the problem for years, and I now think that the solution is fairly simple: Code officials have to insist on a Manual J report, and have to review to be sure it was done. [Later edit: see the next two comments. Armando Cobo makes a good point; code officials will probably need help from third-party energy raters to implement this step effectively.]

    At the recent NAHB Green conference in Nashville, I met a code official from Alpharetta, Georgia, named Paul Ivey. I asked Ivey whether HVAC contractors in Alpharetta produce Manual Js. "They all do," he said. "They have for years."

    "How did that happen?" I asked. "In so many jurisdictions, achieving what you have achieved is seen as impossible."

    Ivey is a calm man. He said, "We just insisted on it." At first, of course, many HVAC contractors grumbled and had to hire a consultant to perform the Manual J. But soon they all realized that this situation was not going to change, and a Manual J became an ordinary step -- and an ordinary cost of doing business -- in Alpharetta.

  12. Expert Member
    ARMANDO COBO | | #12

    That maybe so in Alpharetta, but...
    There are so many ways to do wrong or to cheat a MJ (as you have ponited out) and very few building officials know what to look for. Many building departments are under staffed and under budget, so they'll honestly say that its up to the builder to make sure it gets done right... and there it becomes another issue.
    I hope this series you are writing about becomes a good "short form" guide to all.

  13. GBA Editor
    Martin Holladay | | #13

    Response to Armando Cobo
    While existing regulations allocate responsibility for code enforcement to local code officials, I hope that eventually the responsibility for verifying compliance with a subset of code provisions -- namely, the energy code -- is transferred to third-party experts (BPI-certified or RESNET-certified energy raters).

    This can only happen if we change the building code or change local laws governing code enforcement. Until that happens, it's going to be very difficult for local code officials to require energy code compliance.

  14. user-723121 | | #14

    Big AC's cover inefficient homes
    After air sealing and adding insulation to our home in Minneapolis, we replaced the AC unit with a 2 ton 16 SEER. The house is a rambler with 1664 sf plus full basement, 3328 sf total. Even during the hottest stretches the AC just coasts. We have some good afternoon shading by large trees and our annual electrical usage for AC averages about 250 kWh. For comparison purposes, the heating usage over the last 5 years is about 2.3 Btu/sf/hdd.

  15. wjrobinson | | #15

    Government solutions, more
    Government solutions, more mandates. More more more. Nonsense.

    Markets (and tax manipulated markets) work. Simple works. If using energy unwisely is bad, then tax energy use. Then the customer will make sure the right equipment is used at the right size.

    Inspectors, lots of initials on a card... Cheat too. It just adds cost and inefficiency.

    More for Resnet, which is a closed expensive non simple not needed solution too close to big government.

    In this blog, you all state that the right rule of thumb for your area works just fine. So then be done. Experience is the best and adds no cost.

    Where I am many use AC suddenly and are happy to have some oversizing.

    Volluntary calcs. I am all for.

    I miss the natural building aspect of Robert Riversong. Simple low cost low energy low impact on the planet and small government. That is green building. Not mandatory this that and the other. What next? The governments are broke. Homeowners are under water on mortgages. And the solution is complexity and cost and Resnet? I doubt it.

  16. GBA Editor
    Martin Holladay | | #16

    Response to AJ
    Here's why the market fails to deliver energy-efficient new homes: the builder chooses the specs, but the builder doesn't have to pay the energy bills. The homeowner does.

    An improvement in specifications may greatly reduce energy bills -- but the builder has no incentive to pay 1 cent more for an energy-efficiency measure (especially if it is invisible, like air sealing work), because builders don't pay energy bills.

  17. 5C8rvfuWev | | #17

    RE: greed and cheating -- in this area, HVAC contractors make considerable from installing leaky ducts and over-sizing heat pump and AC. Is that "rule of thumb" somehow suspended in New York? Not likely.

    It's one thing to make a chicken coop pretty. but unless someone is willing to stand watch, the wolf is still going to get what he wants. I damn well know that's true in New York.

  18. wjrobinson | | #18

    Martin, you agree with me.
    Martin, you agree with me. Homeowners are buying cheap energy. The simple solution is to move property taxes of some percentage to energy taxes. Then my way works and is simple. Piling on mandates and Resnet is not simple. Resnet is adding costs to my bottom line and that's the bottom line.

    Joe, sorry to here again you have it out for contractors as I am a contractor. It is not easy to find contractors that have low prices and do the best work. When you do find us, we sure appreciate that glass of lemonade on a hot day. And just to let you in on a little secret, one glass of water and most of us will do everything in our power to really give a top notch service beyond any contracted specifications.

  19. PDaMuGz8Tb | | #19

    Resnet & Energy Star
    I think RESNET and Energy Star have helped raise the bar.

  20. wjrobinson | | #20

    Yes Resnet and Energy Star
    Yes Resnet and Energy Star and all the rest have raised the bar. And the cost. If a customer desires the added cost I am perfectly ok with any added item that is asked for.

    Still my basic point is, we could make all much simpler by increasing the cost of energy.

  21. GBA Editor
    Martin Holladay | | #21

    Reply to AJ Builder
    Like you, I think we need higher carbon taxes so that the price of fossil fuels reflects the environmental damage caused by burning them.

    However, I don't believe that higher fuel prices will be enough to ensure that builders integrate cost-effective energy-efficiency measures in new homes. For example, right now, at current fossil fuel prices, it would be cost-effective for new home builders to pay more attention to air sealing. But they don't do it -- because there are few regulations requiring it, and the regulations that are already in place are never enforced.

    So homeowners get lousy houses -- not because fuel is priced incorrectly, but because there is no incentive for the builder to implement cost-effective solutions to design errors.

  22. wjrobinson | | #22

    We have some PGH builders in Upstate NY
    Maritin, lousy homes, put it this way, every condo in ski country, has had to be rebuilt structurally let alone air leaky energy wasting, So yes the condos of the boom years were built lousy. But around here and Saratoga we have several new insulation companies and they are busy working for the best larger builders. The Showcase of Homes that is held annually may via competitive forces be aiding our area. I see the Owens Corning seal system is in use here now, we have spray foam companies, the one I use has a blower door and rater associated now. So, we are doing I think better than lousy. I build. I stick up for our good builders. In my area those are the builders I know and recommend besides what I can do.

  23. dickrussell | | #23

    Educate the homeowner
    As I went through the process of designing our new house and selecting/working with a builder to execute the plan, I found myself wondering how ordinary folks ever get a really energy efficient, comfortable, and durable home. Most of the time a really good result won't happen if the homeowner's role is little more than to stand aside for 6-12 months and write checks.

    While there are pockets of proficiency among builders of the PGH and better homes, I have to think that making the PGH the norm rather than the exception is heavy upfront education of the people who engage builders for their projects. If the demand and oversight doesn't come from the homeowner, the only construction in the PGH and better class will be done by those select few who want to set themselves apart from and above the others.

  24. GBA Editor
    Martin Holladay | | #24

    Response to Dick Russell
    You're right that an educated person who is building a new custom home can sometimes assure that the building is energy-efficient.

    But the vast majority of new homes in the U.S. are not custom homes -- they are spec homes built by developers. And over the long life of the home -- perhaps 100 years -- all of the subsequent buyers of the home were not involved in the original specifications in any way. That's one reason that we need building codes.

  25. mNyt3iooKp | | #25

    I agree resnet and energystar
    I agree resnet and energystar have raised the bar, we should be designing today to much higher standard!!

  26. user-984364 | | #26

    Energy Star rule of thumb oversizes?
    The 1 ton per 1,000 square feet rule of thumb is interesting...

    The Energy Star site says that for room AC (i.e. window or wall units) 12,000 BTU (1 ton, right?) is good for 450 up to 550 square feet. So that's double the capacity of the "Blasnik and Bailes" ruie of thumb...

  27. [email protected] | | #27

    low BTU mini-splits?
    Seems like the smallest mini-splits are still 9000 BTUs with an 18000 BTU HP. What if you you only need 1/2 that or less? The application is a couple of rooms in a very tight well-insulated house- with CASEMENT windows (also want to get the compressor sound away from the rooms). I get that with EMC motors they can run at 3000 BTUs- but it still seems like a lot more capacity than needed. Any ideas?

  28. GBA Editor
    Martin Holladay | | #28

    Response to Christopher Vlcek
    Obviously, it irks you to have to buy more capacity than you need. However, you can console yourself with the information that oversized ductless minisplits operate at a higher efficiency at part load than at full load, so you'll end up with very low energy bills. (The reason for the high efficiency at part load is the fact that they use inverters and variable-speed DC motors.)

    For more information on ductless minisplit efficiency (especially when a unit is oversized), see Dana Dorsett's comment (comment #7, posted on 5/25/12 at 13:52) on the topic on this blog page.

  29. srivenkat | | #29

    Rule of thumb for basement

    Many thanks for the much valued advice on my other questions.

    As for the revised rule of thumb of 1-ton per 1000 SFT of conditioned space, should I treat the basement area on the same level as the main level? I have 2150 SFT on the main level with another 2150 in a semi-walkout basement underneath the main level. Should I figure 4.3 tons or ignore the basement and figure 2.15 tons?

    I have received a quote from an HVAC contractor specifying a 3-ton GeoComfort GSHP for my home. He took some measurements and mentioned to me the specification was from a Manual J calculation. He normally would see a 4-ton specification for a home my size, but because of the air-tight concrete walls (ICF/EIFS) it came out to be 3-ton. So I was trying to see if this makes sense from a rule-of-tumb perspective.



  30. GBA Editor
    Martin Holladay | | #30

    Response to Venkat
    If your basement is finished, conditioned space -- and you want to include ductwork and supply registers so it is heated during the winter and air conditioned during the summer -- then you should include the area when performing a cooling load calculation.

    If your basement is unfinished and won't be conditioned, then you wouldn't include it.

  31. t_t | | #31

    The 1000 sqft per ton rule of
    The 1000 sqft per ton rule of thumb, is for cooling only?

  32. coolmybldg | | #32

    Cooling in cool climate
    One should take note that each location/climate would have their own rule of thumb since the difference in the set room temperature and the ambient would somewhat have a major factor to the heat gain of the room. I'm in the tropics and our rule of thumb is nowhere near this 1 ton:1000sqft (1.5hp:100 sqm) Our rule of thumb here is around 1hp:10 sqm. That's a whopping 10hp compared to the 1.5 for the same 100 sqm. 6x more! The temperature outside is in the 30s ˚C (86˚ F) and about 50-60 humidity. We're also using inverter DC compressor motors which means the 100% running at the hottest is not a good gauge. If the compressor runs at 100%, the longevity of the motor and the efficiency drops. This article should be revised or at least take note that it's meant for a specific latitude in the US. I would guess that Florida would differ from this "rule".

  33. GBA Editor
    Martin Holladay | | #33

    Response to Victor Ong
    Performing a Manual J cooling load calculation is always preferable to using a rule of thumb.

    That said, the conditions you describe -- an outdoor temperature of 86°F and an outdoor relative humidity of 50% to 60% -- aren't particularly challenging. In Florida and Houston, these conditions are normal -- and in fact, outdoor temperatures and humidity levels are often much higher. So a rule of thumb that works in Florida would certainly work in the climate you describe.

    You didn't mention the name of the country where you live. It's worth pointing out that the rule of thumb mentioned in this article is applicable to a well-built new home in the U.S. -- one built with attention to airtightness, with adequate levels of insulation, and with double-pane low-SHGC windows. In Cambodia, a tropical country I just visited, new buildings have no insulation, lots of air leaks, and single-pane windows. These buildings can't use rules of thumb developed for Energy Star homes in Florida.

  34. GBA Editor
    Martin Holladay | | #34

    Response to Tony Tibbar (Comment #31)
    Q. "The 1000-square-foot-per-ton rule of thumb is for cooling only?"

    A. Yes.

  35. nefrusy | | #35

    Manual J for peak loads, realistic loads, or both?
    I recently got my hands on a residential cooling-load program based on Manual J (guessing an older edition), and I'm wondering what the best use is of the tool; is it true that the calculated load is simply the peak load at the hottest part of the day with highest solar gain?

    I'd like to determine the realistic cooling load on the house, like a daily average for the warmer part of the season. For example, if during a typical hot summer day the load on the house averaged 18K BTU/H, but peaked at 22K for short periods (ex. evening sun, cooking dinner, lots of company, etc)... seems like the right sized AC would be 1 1/2 tons, not 2 tons.

    If Manual J is meant for determining peak loads, does it make sense to size AC equipment to that peak load? If a house is only experiencing those conditions for a brief 2 hour window, then wouldn't the thermal mass of the home help a smaller AC system get through those peak hours without much problem? (ex. maybe a 2 degree rise in temp) Or does it make sense to size the equipment to peak 1% load, because on the really hot 0.1% days, the AC is going to be undersized--and relying on that thermal mass?

    Also, I've seen conflicting advice on how to treat windows; what's funny is you mention that HVAC contractors may "assume that none of the windows have curtains or blinds" to get higher #s, but I read that should be the assumption because homeowners tend to leave window coverings open on a sunny day (for natural light, view, etc). I've also read that tree cover should not count towards external window shade, since "trees can be cut down." But that seems crazy, unless the trees really are likely to not be around in the next 10-15 years.

  36. GBA Editor
    Martin Holladay | | #36

    Response to Nathan Efrusy
    Q. "Is it true that the calculated load is simply the peak load at the hottest part of the day with highest solar gain?"

    A. Not really. The outside design temperature usually represents a temperature that is exceeded only 2% or 2.5% of the time from June through September. These values are referred to as the 2% design temperature or the 2.5% design temperature.

    This approach accounts for your concern that choosing equipment designed for only the hottest hour of the year amounts to overkill.

    When it comes to equipment sizing, you will ideally follow the guidance provided by ACCA Manual S. It's not quite as simple as "now that I know my cooling load, I'll choose equipment than satisfies the load" -- in part because oversizing is undesirable.

    For more information on Manual S, see Manual S brochure.

    If you are the homeowner, and you are in control of your window blinds, you can make any assumption you want concerning whether the blinds are likely to be up or down. Similarly with trees.

Log in or become a member to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |