GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Manual J and equipment sizing

Chad_Attermann | Posted in Mechanicals on

Approaching the final stages of a 60’s ranch major renovation in climate zone 5.  We added an approximately 1000 sf great room that will not be served by the existing ducted system, so I would like to use a ductless system to condition it.  I have performed a room-by-room Manual J using CoolCalc which I feel should be fairly accurate given my inputs and I’ve come up with 16,725 heating BTU and 13,964 cooling BTU (13,955 sensible and 9 latent).  Heating design temp is 14F and cooling is 95F.  For infiltration I specified “semi-tight” in CoolCalc (whatever that means), but I am shooting for under 1.5 ACH50.

My first question is, this being my first Manual J, do these loads seem reasonable for a 1000 sf room with cathedral ceilings and a fair amount of glazing (about 100 sf on the South wall and 180 sf on the East wall)?  The South glazing is solar heat gain and East is low-e.  As for insulation, CoolCalc doesn’t give a lot of options for R-value so what I entered might be a little under, but ultimately we should have approximately R-50 unvented cathedral ceiling and R-30 walls.

My second question relates to equipment selection.  I have read the discussions about single vs multi-zone units and would prefer a single wall unit for it’s increased efficiency and range of output.  Is it reasonable though to expect a single wall unit to be able to evenly heat and cool the entire space, especially at the extremes of the design temps?  The Manual J also calls for 704 cfm which seems high for a single head.  All of the estimates that I received to date call for 2 to 3 heads, but they are all grossly oversized too, so I don’t know if the multiple heads are for coverage or just plain oversizing.  If it does make more sense to go with two heads, is it still advisable to go with single zone units, or would that fact that both heads are in the same space allow a two zone unit to operate at maximum efficiency?

Any advice is greatly appreciated.

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. Chad_Attermann | | #1

    As a follow-up to my own question, I forgot to ask about sizing of equipment. I know there is a science to it and I plan to look into a Manual S, but as a rough guess to start selecting equipment to look closer at, should I be comparing the loads from the Manual J with the "max" heating and cooling outputs of the equipment, or with the "nominal" outputs? One model in particular that I'm looking at (Fujitsu 15RLS3Y) has nominal heating/cooling outputs of 14,500/18,000, but max heating/cooling outputs of 18,400/23,900. With that substantial a difference I want to be sure I'm using the correct values.

  2. hughw | | #2

    I have a different question regarding Manual J and CoolCalc which I apologize if its inappropriate to tag on to this thread....When doing Construction Details, there's no option (at least in the free version) for some typical details that many of us use. For example, there's no roof option for CCF and R49, for frame walls the highest cavity insulation value is R21 when the code calls for R30.

    1. Chad_Attermann | | #4

      That is in my opinion one of the biggest drawbacks of using CoolCalc. I found the following explanation in their FAQ:

      "Construction feature options in CoolCalc are derived directly from ACCA tables. Not all possible constructions are covered in ACCA tables so if you don't see the exact option you are looking for, selecting a substantially similar option is acceptable.

      Newer homes may have better R-values than the available selections from ACCA tables. As R-values increase, the energy efficiency benefit between R-values becomes relatively smaller so using the nearest available R-value is acceptable in most scenarios. CoolCalc does not yet implement custom R-values (MJ8 appendixes 12 and 13), if you believe the construction features of the home are substantially different than the available options in CoolCalc, please follow ACCA procedures in MJ8 appendixes 12 and 13."

  3. Trevor_Lambert | | #3

    That sounds like a lot of BTUs. I'd try a different calculator. You shouldn't need multiple heads in a single room.

  4. STEPHEN SHEEHY | | #5

    We have a single 12RLS3H in our great room which is about 800 square feet or so. Mostly 13' ceilings. Climate zone 6, 99% design temperature is 0°F. R42 walls, R 70 roof, .59ach50. No problem keeping the space as warm as we want. 16,750 btu/ hour seems a little high.

    1. Chad_Attermann | | #6

      Your situation seems fairly similar to mine, with the exception of your lower ceilings and higher R-values. Do you happen to know what the calculated loads were for your great room?

      1. STEPHEN SHEEHY | | #7

        I don't think I have info on room by room loads. The whole house is about 1650 square feet of conditioned space and the total load was only around 12,000 btu/ hour. The unit in the great room also heats the entry, mechanical room and guest room/ office. Probably a total of about 1250-1300 square feet.

        We have a 9 RLS3H in the bedroom, but we probably could have managed without it. It's nice on the few hot days when we use A/C.

  5. Expert Member
    Dana Dorsett | | #8

    You'd only be looking at multiple heads if its an extremely long rectangle footprint, say '15 x 67'.

    The Fujitsu 12RLS3 will probably cover your loads if you're anywhere near 1.5ACH/50.!/product/25324

    I'm a bit skeptical of CoolCalc's & LoadCalc's accuracy with taller rooms of non-standard construction- you're probably getting a significant overestimate of reality.

    Ignore the cfm numbers spit out by the tool. Those are more relevant to ducted systems with 1-2 speeds with the air supplied and returned from duct registers with a fixed direction. Mini-splits can be set up to continuously sweep to provide better mixing over wide areas. If mixing is an issue at low load in a big-ish room like that you can always set the mini-split blower speed manually to something a bit higher. At maximum load the blower would run at maximum speed/maximum mixing

    1. Chad_Attermann | | #9

      Thanks for the reply Dana! My great room is roughly 36' x 24' so definitely not extremely long. I feel like the loads from CoolCalc are exaggerated as well. I have considered modeling in Beopt too. Do you think that would give better results?

      As for mixing, any suggestions on mounting location for optimal mixing? I was thinking of the Southern exposure wall which is 24' long and 12' high, and mounting as close to the top as possible. That way it would be directing air along the long axis of the room.

      Thanks again for your insight.

      1. Expert Member
        Dana Dorsett | | #10

        It's worth running a BeOpt sim on it- I'd trust that more than the online freebie/cheapie Manual-J tools.

        A room layout with window locations and wall heights labeled might be useful for optimizing placement. With a side to side sweep putting it in the middle of the longer wall might work just fine.

        "....mounting as close to the top as possible..." isn't the best strategy. Unless you hail from the Land of the Giants you don't much care what the air temperatures are above 6' from the floor, and in heating dominated zone 5 injecting the heat 10'-12' from the floor it would be taking in warmer, stratified air which cuts into efficiency, and heating the middle stratum of the room, which isn't the most comfortable. Putting it high enough to not be banging your head on it might be necessary, but higher is definitely NOT better, especially in a heating dominated climate.

        Case in point: A project I consulted on a few years ago (also in climate zone 5) there was a 1-ton Fujitsu head mounted about 9 feet from the floor above a bay window in room with a cathedralized ceiling with a ~13' ridge, with the head centered below the ridge. It worked fine during the cooling season, but the sensed temperature offset between the head and where people would sit in front of the bay window was sometimes ~10F or more in winter, and constantly changed with the weather.

        I recommended using wall-remote/thermostat rather than the hand held. Instead they had the tech who had installed it come out and tweak the calibration inside the head to where the setpoint stayed within ~5F of the room temp. The offset still changes with the weather, but they seem OK with it even though it's not great (fussier people would hate it). Had the head been mounted on one of the other walls at about 7 feet from the floor blowing perpendicular to how it had been installed the offset would have been smaller and the heat would have been inserted nearer the humans, and it would have been less of an issue. In their case they usually light up a small wood stove in that space when it's hitting the teens or cooler overnight at which point the ductless head turns itself off, even though the head has the capacity to fully heat the space.

        1. Chad_Attermann | | #11

          I have attached a layout of the room with some wall dimensions. The longer wall on the East has a fireplace in the center which will make mounting a unit there difficult. Opposite wall (West) is shared with the garage and might be an option although it's mostly in the kitchen with upper cabinets at 10'. The North wall will have TV and book shelves.

          I see your point about heating efficiency with the unit mounted too high. I do have unfinished basement below most of the great room if you think that a ducted unit with super short runs up into the floor might make more sense.

  6. KeithH | | #12

    DIY Opinions:

    Can you post the report or a screenshot of the pie charts so we can see the breakdown?

    Regarding missing insulation values in CoolCalc:
    - I'm not convinced that given you are DIYing the sizing (and doing better than most HVAC companies would bother with?) that the difference in R38 and R50 is significant in the ceiling or R21 to R30. I'd say whether you have continuous insulation and entered that is much more relevant.
    - If you really want to pursue the insulation question and you don't have exterior continuous insulation (WHY), calculate the U-value of your assemblies and then find a chart with default wall assemblies U-values. Perhaps R0+R30 matches R6+R21 better than R0+R30. If you have continuous insulation less than R-6, perhaps you can cheat that value to accommodate it. For example, your wall is R-3 (1/2" foam) + R30 fill and you use R-5 + R21 to simulate the U-value. My guess is that you don't see much difference.
    - Absolute numbers: I think the numbers you got are reasonable if you don't have exterior insulation (or didn't enter it) and chose semi-tight.
    - I think choosing semi-tight, unless you know some part of your system is still leaky, is a mistake. It seems like air-sealed and modernized means tight and passive house etc means you can't use the stock infiltration settings. What part of your load is infiltration?
    - How much of your cooling load is AED excusion? I know cooling calc specifies AED excursion. You might have a high peak cooling load due to your glazing. I believe (and I really don't fully understand the science) that appears as an AED excursion penalty on the sizing. If you'd cover the glazing on a hot summer afternoon, you could probably downsize this amount. If you expect interior design temperature at 3 pm on peak design day you probably can't.
    - I'd also check and make sure there isn't a big internal load if the room won't be used a lot in design conditions (coldest night, hottest afternoon). I think the default BTU cooling loads for appliances and lighting are pretty inaccurate these days. Though it may not matter if heating is the larger load for you.
    - You could play around with your model on coolcalc to see what downsizing looks like: I believe you can alter the interior design temperature on coolcalc. Perhaps going down a unit size results in 78 on a design day. Maybe you like that better.
    -But ... all the gurus on here seem to say (on other threads) that the mini splits work more efficiently when they aren't at max load. Perhaps a small amount of oversizing (18k vs 15k; not 36k vs 15k) isn't a disaster. The larger blower included might address your distribution concerns as well.
    - Regarding distribution: rather than substantially lowering efficiency options and increasing install costs by choosing a multiple head, why not install a large reversible ceiling fan that has good low speed operation? In the summer, it will help with distribution and comfort. In the winter, well, ceiling fans can be used to help with distribution but wind chill is an issue too so might not be useful.
    - Bear in mind that design interior temperature is probably 75F. Not everyone work with that. I know I can't work in my home office if its 75F. I need 72F or lower to work. To relax, 75F is typically fine. To sleep, I want lower but then that doesn't happen at outdoor design temperature so every system is oversized for that.
    - Did you modify the default windows and walls and such to make your conditions or create your own? If you created your own and left the defaults alone, I'd double check you got the defaults in correctly. I also found on cool calc that its easy to end up with the default windows per wall layout on a wall with no windows. I'd go back through and check your windows tabs on each exterior wall and make sure there are no stowaways.

    Final: To me your numbers don't seem crazy. You have quite a bit of glazing, you assumed some (too much?) air leakage. I'd check your continuous insulation inputs on CoolCalc and window layout/values and switch to tight construction and then see if you are close to 12k. You can get a more efficient unit if you can get to 12k. It's not like the SEER for the 15k Fujitsu is bad!

  7. Chad_Attermann | | #13

    I suppose I should have posted the report from the start! Here it is. Thanks a lot for all of the great advice. I absolutely did find a lot of window and door stowaways early in the process and meticulously checked every room and confirmed that only my custom windows and doors are present. I'll play some more with the design and see how values for infiltration, continuous insulation and design temps change things. My feeling, like you said, is that the 15K unit is still efficient and not grossly oversized anyway, and is 1/2 the size of the next smallest system that I was quoted by contractors.

  8. KeithH | | #14

    Again, I'm not a pro but your infiltration and internal load members look high. I can't imagine what internal load would result in 3500 BTUs. What's that, 7 people doing aerobics? I'd check to make there aren't some old school lighting and tv loads. Tvs don't make much heating load anymore, same for lights.

    If you airsealed everything, I'd try the tight infiltration setting.

    But as we've both said, there are nice SEER 15k units available so perhaps more downsizing is not very fruitful.

    Good luck. Do come back and report on a design day in August!

    1. Expert Member
      Dana Dorsett | | #15

      It's pretty easy to add 3500 BTU/hr or more of cooling load while cooking dinner. That's one gas burner on a range at mid-fire, or about 1100 watts of electricity use (video games, anyone? :-) )

      Figure 300-500 BTU/hr per conscious human depending on level of activity (250 BTU/hr even for the couch-potato who passed out snoring with the remote in hand.)

      3500 BTU/hr of internal gains is not an insane load under any number of scenarios. The refrigerator is still running, the water heater standby (assuming a conventional tank) is still piling on, it all adds up.

  9. KeithH | | #16

    I thought this sizing and equipment wasn't covering the kitchen. I guess I might have misinterpreted great room. I assumed that didn't include a kitchen.

    If it's just a living room, then TV and possible AV equipment (video game console and stereo receiver being the largest items) do add some load. Some quick research yielded the following:
    75" Samsung LED LCD 140w typical
    Console running a graphics intensive video game 120-140W
    Mid priced 7.1 Denon receiver 500W
    10" Subwoofer 100W
    ~900W. Certainly some additional devices or high power receiver could yield 1100 W.
    Yet another reason to stick the TV room in the basement!

    Well, look at that. Dana's right. 1100 W = 3753 BTU/h and we haven't added any people. So I guess those internal loads are reasonable. My mistake, thanks for the correction Dana.

    What was your thought on infiltration Dana?

    1. Chad_Attermann | | #17

      I questioned the high infiltration heating load too, especially relative to infiltration cooling load. Following is what support at CoolCalc had to say:

      "Heating Load Infiltration = 1.1 * Altitude correction factor * net infiltration CFM x heating temperature difference
      Cooling Load Infiltration - 1.1 * Altitude correction factor * net infiltration CFM x cooling temperature difference

      Your heating temperature difference = the indoor winter temp - outdoor winter temp = 70-14 = 56
      Your cooling temperature difference = the outdoor summer temp - indoor summer temp = 95 - 75 = 20

      Just based on that, the heating infiltration values are expected to be 2.8 times higher than the cooling infiltration values.

      In addition the air changes per hour which are used to calculate the net infiltration CFM values are double for heating than they are for cooling and those values are part of a formula that squares them.

      Once you square ACH, the net infiltration CFM for heating is about a 4x higher than cooling. If you factor this with the fact that the temperature difference is also 2.8x higher, you would expect the heating infiltration values to be somewhere around 11x higher than the cooling infiltration values."

  10. Jon_R | | #18

    > those values are part of a formula that squares them.

    Do you have a link to this formula?

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |