Are minisplits a good solution for heating (and cooling) a smallish addition to my 85-year-old house?
I’m in zone 4 (NYC suburb) and the original house is a 2 1/2 story wood framed house heated by gas fired one pipe steam. Cooling is from window units. Our addition (currently rough framed and dryed in) is two stories with an 18′ x 15′ exterior footprint. 3 foot block stem walls make the lower level slightly smaller than the upper. The upper level, which connects to our main floor, will be our new kitchen, and the lower level, which connects to a walkout basement, will be a play/TV room. The main floor of the house will be pretty open with large openings between the kitchen, dining room and living room. The resulting main floor will be just under 1100 sq. ft. The lower level of the addition will have a direct connection with a room that is about 300 sq. ft. The addition is 2 x 6 construction with one inch of XPS on the exterior and planned cellulose in the walls and above the kitchen ceiling. There is two inches of XPS under the addition slab The original house had blown in cellulose retrofitted to the 2 x 4 walls and some other air sealing and insulating work at the rim and in the attic.
Mini-splits seem like a good option because of their efficiency and the fact that they heat and cool. I’d love to have main floor cooling and dehumidifying provided by the ductless rather than installing window units every summer and taking them back out every fall. They also seem like a more streamlined option compared to trying to add on to the steam system, or adding a new small hydronic zone, or both. Neither of which would offer improved cooling.
Does my situation sound like two mini-splits, one up and one down, would be a viable option?
I know that the rooms themselves are likely too small for any size unit. Does the open floor plan change that significantly?
I would welcome any advice from someone who has experience with this kind of setup. Does the thermostat for the steam get set back a few degrees? Would I need to relocate the thermostat out of range of the mini-split?
Thanks in advance
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Mike,
Yes, you should be able to use a ductless minisplit to heat and cool the main floor. You can adjust the thermostat on your steam heating system anyway you want -- to a higher point than the minsiplit if you want to heat with gas, or to a lower point than the minisplit if you want to heat with your minisplit.
You may not need a minisplit for your basement TV room -- it depends on how often the room gets used, and how hot the room gets during the summer.
Run a heat load calculation (either Manual-J or I=B=R) for both the upper & lower floors of the addition (and include the adjacent rooms if separated only by an open archway that convects freely.)
If a zone doesn't have at least 4000BTU/hr of heat load at the 99% outside design temp it's not a great candidate for a ductless solution. To get the efficiency out of mini-splits/multi-splits the ductless head should not be oversized by more than 50% (25% oversizing is fine, and may even improve efficiency in some cases.) The smallest ductless heads out there deliver the 6000-7000BTU/hr range at NYC-burb outside design temps (+10F to +15F).
Try this freebie Manual-J room load calculator:
http://www.borstengineeringconstruction.com/Heat_Loss_Analysis_Calculator.html
Be sure to use credible inside & outside design temperatures (if you're OK with letting those rooms drop to 65F on the coldest night of the year, use 65F even though you keep it 73F when you're watching TV with grandma.) Estimate your 99% temperature bin from nearby cities on the ACCA Manual-J list:
http://www.energystar.gov/ia/partners/bldrs_lenders_raters/downloads/Outdoor_Design_Conditions_508.pdf
When you come up with some load numbers it'll be easier to tell if this is in the sweet spot for ductless (or mini-ducted) solutions or not. Don't rule-out multi-splits- 2- zone multi-split with the heads & compressor right-sized for the loads can beat the as-used performance of two oversized mini-splits on both efficiency and comfort.
Martin,
Thanks so much for your answer. I have many follow up questions, but I'll try to boil it down to a couple. I'm wondering if I should be sizing the kitchen unit to the whole main floor? Certainly for cooling that makes sense. If yes, should I turn off the minisplit when I want the steam to provide more heat throughout the first floor and also second floor bedrooms?
The lower level TV room does connect to a basement, but it's a walkout so the room will have three fully above grade walls. The basement it connects to has been heated by the uninsulated steam pipes. It stays pretty comfortable; not too hot and only a little cool on the coldest days. I imagine the more I use the main floor minisplit, the colder the basement will be and the more I would want another unit on the lower level. Does that sound right?
Hi Dana,
Thanks for that information. Particularly the two singles vs. a dual zone. The efficiency numbers of MSZ-FH units are so high compared to 2 zone units. It sounds like that doesn't necessarily hold true in use.
I did a load calculation for the main floor on another free website and came up with 16.6K btu/h for heating and 16.8K btu/h for cooling. I used 6 degrees as the design temp because that's what the architect specified on the plans. I'm about 20 miles north of White Plains if that helps any.
Is that too conservative?
If so I'll run it at a higher design temp. The indoor temps were the default at 70 heating and 75 cooling. I set our thermostat to 68 in the winter, so that can be adjusted. If it's dry inside, 78 feels luxurious in the summer, so I can bump that up too.
Dana,
FYI, when I ran the calculation with a 12 degree design temp and 68 and 78 for indoor temps, I came up with 15400 btu/h cooling and 14300 btu/h for heating.
Not all free load calculators are the same. The Borst Construction freebie at least uses ACCA Manual-J methodology, and will be more accurate than most online tools. A big WAG in all of these calculations is the natural ventilation rate, which is usually estimated at higher than reality. Even most Manual-J calculations overshoot reality by ~15%, sometimes more. Which online calculator were you using?
A +6F outside design temp seems low for any location (other than a mountaintop) within 20 minutes of White Plains. Plug your ZIP code into weatherspark.com dashboard graphing temperature, then compare how much colder (if any) your location seems to be than White Plains ( http://weatherspark.com/#!dashboard;a=USA/NY/White_Plains )on the coldest 5 days of the year.
If it turns out that the heat load of your upstairs zone calculated 14K @ +12F using Manual-J, the -FH12NA is going to be better matched than the -FH15NA. A calculated 14K load is more likely to actually e closer to 12-13K unless you were VERY aggressive in subtracting out every last possible plug load and warm-blooded body, and have blower-door test numbers to better estimate the infiltration rate.
But re-run the load numbers with a couple of tools, and be very aggressive about the load-mitigating assumptions. Being undersized for the cooling load by 2-3000 BTU/hr is not a disaster- the thing will run nearly continuous cycles and do a better job on the (fairly substantial) latent load on days where it might not be able to hold the line at 78F. (At 35-40%RH even 80F is still pretty comfortable.)
Run the load numbers on the downstairs zone too.
Dana,
The other website I used was loadcalc.net. I think the weather data weatherspark.com uses for my zip is actually White Plains, which happens to have +12F design temp. I will run the numbers again on the Borst site. My thinking is to go with the new FH model units as it does dip below zero on occasion. Several times last winter. Would that capability be a reason to eliminate dual zone units that can't heat below zero?
The H2i multi-split units all have a rated output down to -13F, they are just not quite as efficient as the single-head units. The -13F output of any H2i unit is typically 60-75% of it's full output at +5F. (It varies by model)
Basement heating and cooling loads are often extremely low, but if your's is a walk out it may warrant it's own ductless head. If the heat load in the downstairs isn't too much less than the -FH09NA there is no advantage (other than up-front cost) for going with a separate unit.
Does anyone have a source for "Internal Appliance Sensible Heat Gain" numbers? I'm trying to find realistic numbers for new efficient appliances so I can calculate the cooling load. Or, instructions on calculating this myself would be much appreciated.
Mike,
It's very difficult to perform cooling load calculations by hand. For lots more information on this topic, see Calculating Cooling Loads.
The article notes, "The sensible heat gain (body heat) emitted by one occupant is usually assumed to be 230 Btuh (67 watts). The default assumption for a home’s appliances is usually 1,600 Btu/h (469 watts); lighting is assumed to add another 1,600 Btu/h (469 watts). These default assumptions can be adjusted up or down if necessary. If a house has an unusual number of appliances or equipment — for example, a home business that includes a computer server room — internal loads will be higher than these default assumptions."
Thanks for that information Martin. I'm trying to avoid the pitfalls discussed in the link you provided. My goal had been to find, or calculate, actual heat gain for the appliances and lighting that I'm going to install so as to not use assumptions that oversize my cooling equipment. The free calculator I'm using calls for the simple number of occupants but needs a number in Btu/h for appliances, appliance latent gain and lighting. Using the assumptions you provided has me at about 20K Btu/h for cooling my main floor which, in minisplit models I've seen, would be oversized for my heating load (14K Btu/h). I do have some poor performing, western facing windows that also contribute to heat gain so that may just be my situation. Ideally I'd have exact numbers for every entry, but it seems that even with pure intentions, there is a bit of art to choosing the numbers that go into the calculations.
Is there a service, independent of HVAC contractors, that would make sense for consulting and calculating on such a small project? I'm wary of going with an installer's numbers since the people that have given me estimates on installation so far haven't even pulled out a tape measure.
Mike,
A HERS rater or an energy consultant should be able to perform a Manual J calculation for you. (A Manual J includes room-by-room heating load and cooling load calculations.)