Mini split heating for Colorado foothills
I’m interested in placing ductless mini splits into a house in the Colorado foothills. The home currently has electric baseboard heat and can easily use 5000KW in a cold month. It seems like many of the new mini split models remain efficient at temperatures well below freezing. In my town the daily average lows during the winter are 10-15F. I’m going to ask several HVAC contractors for proposals but I’ve read that system sizing is often done sloppily. What should I be looking for in terms of:
1) System size? House is ~3000sqft on 3 levels. Can I use my electic heating bill to back into a BTU requirement?
2) Any particular systems to look for or avoid? I’m assuming I want something with a high HSPF (12-14?).
Any advice would be appreciated.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
>"The home currently has electric baseboard heat and can easily use 5000KW in a cold month."
That would be one of the least efficient 3000' houses in Colorado. 5000KW is a energy consumption rate of 17 million BTU/hr.
What I expect you really mean it uses 5000 KW_H_ in a cold month, an ENERGY (not power) use of 17 million BTU (170 therms) over 30 days (not one hour), which would be a pretty efficient 3000' house.
>"1) System size? House is ~3000sqft on 3 levels. Can I use my electic heating bill to back into a BTU requirement?"
Yes. To get a rough handle on it take your lowest electricity use month kwh and subtract it from your wintertime months' usage, then use weather data to correlate the remaining power use to heating degree-days. See:
https://www.greenbuildingadvisor.com/article/out-with-the-old-in-with-the-new
That will help figure out the whole-house load, but not the zone loads. Doing a room by room IBR type load calculation can be useful in apportioning that load, which will help determing how to best zone it, and match equipment to those zone loads.
Derating capacity for altitude is also important for selecting the equipment :
http://meus1.mylinkdrive.com/files/Application_Note_2003_-_High_Altitude_Applications.pdf
>"2) Any particular systems to look for or avoid? I’m assuming I want something with a high HSPF (12-14?)."
The bench-tested HSPF takes a back seat to capacity at your 99% outside design temperature.
Using only vendors that have a reasonable amount of support in your areas is also more important than the nameplate HSPF.
As a general rule, using multiple single zone units will be more efficient, and usually cheaper than a mulit-spit Medusa. Having multiple outdoor units also provides redundancy- if one compressor craps out you're not completely out of heat (though you would presumably have the baseboards as low efficiency back up.)
Yes, sloppy language on my part - 5000+ KWH in a month. In a peak month the electric bill is $450-600 @ 10c/KWH. In a summer month it's closer to $100 @ 13c so that leaves a usage related to heating of roughly 3500-5000 KWH depending on how cold the month was. How would I translate this into a BTU system size?
For the vendors is there any way to know besides just calling a bunch of HVAC companies and seeing who they recommend?
Great to know about the single units vs. multi zone, that was one of the things I was trying to evaluate.
Really appreciate the help!
Take a look at the vendors' " contractor finder" web pages, see how many installers are listed within an hour's drive of your location, the call a few, get some references. eg:
https://www.mitsubishicomfort.com/get-started/contractors
http://contractors.fujitsugeneral.com/
Throwing dollar signs into the electricity billing periods just confuses the math- keep it in kwh, since those are the units of energy.
To figure out how determine the 99% heat load from the net heating kwh, read the article- it's the same math, correlating the power used to heating degree-day data from a nearby weather station, then making a linear approximation of kw per heating-degree based on the ratio of kwh per degree hour. Put in a single sentence it sounds more complicated than it actually is- there is no rocket science involved- just 5th grade arithmetic. :
https://s3.amazonaws.com/greenbuildingadvisor.s3.tauntoncloud.com/app/uploads/2016/09/08060512/001%20Altitude%20Capacity%20Correction.pdf
The article uses therms & BTUs as the units of energy rather than KWH, but if you can come up with the KW load at the 99% outside design temp it's a simple conversion constant: 1 KW = 3412 BTU/hr
Here's another altitude derating document for heat pumps:
https://s3.amazonaws.com/greenbuildingadvisor.s3.tauntoncloud.com/app/uploads/2016/09/08060512/001%20Altitude%20Capacity%20Correction.pdf
Here's what I got using the methodology you linked:
4500 KWH = 15.3m BTUs
Sum of 65 degree measurement from Boulder was 1090 (TBH there isn't a station that well approximates where we are, as we are 2k ft higher than Boulder...i tried a station further in the mountains but that measurement would have been 1700 which seemed less realistic...so using Boulder for now)
15.3m/1090 = 14,087 / 24 = 587 BTU per hour * 65 degree difference (99% load in our area is ~0) = 38k
38k * 1.4 ASHRAE = 53.4k * 1.26 (adjusting for 8k elevation) = 67k BTU system?
Does that seem right?
Assuming ~3F per 1000' adiabatic cooling, there would be an additional 6 HDD per day beyond what Boulder experienced. Aren't you usually ~6F cooler than Boulder? If yes the calculation is a bit of an overshoot.
Assuming that was a 30 day period that adds 180 HDD to that 1090 HDD number for a total of 1270 HDD.
4500KWH/1270HDD= 3.54 KWH per degree day or (/24= ) 0.1476 KWH per degree-hour
With 65 heating degrees that would be 65F x 0.1476 KWH= 9.60 KW which is (9.6 x 3412 =) ~32,750 BTU/hr.
Call it 33K instead of 38K. Either would be a credible number for a 2x6 framed reasonably tight 3000' house.
That 4500KWH number is a suspiciously round number to have been the adjusted power use after subtracting out some background.
The 1.4x oversize should be a fairly firm upper bound (I usually try to keep it under 1.5x, but over 1.2x) when using ductless, so that the system would be running in it's high efficiency modulation range.
The correction factor for altitude depends on the manufacturer, and with a multi-split you need to look at the corrections of both the compressor (sometimes incorrectly referred to as a condenser) and the heads/cassettes independently- it's not a single number.
At 8000' a Mitsubishi MXZ-8C48NAHZ compressor puts out 94% of it's specified capacity at sea level, which is 54,000 BTU/hr @ +5F, so call it 51K @ +5F, and about 49-50K @ 0F (without looking up the extended temperature capacity tables). That would be a 1.3x oversize factor if the load is really 38K, a 1.5x oversize factor if the load is more like 33K, which is fine, but try to work those kwh number as accurately as possible, and over periods longer than just one month.
With more carefully calculated load numbers it may make sense to drop back to the 3.5 ton -5C42NAHZ, which would be good for about 43-44,000 BTU/hr @ 0F at 8000'.
http://meus1.mylinkdrive.com/files/MXZ-8C48NAHZ_Submittal.pdf
http://meus1.mylinkdrive.com/files/MXZ-5C42NAHZ_Submittal.pdf
At 8000' the individual zone HEADS only put out 74% of their rated capacity, so pay attention to the zone-by-zone loads and upsize the heads by ~35%. eg: Half ton Mitsubishi head like the FH06NA is good for 87oo BTU/hr heating at sea level when married to an MXZ series compressor, but would only be capable of 8700 x 0.74= 6438 BTU/hr at 8000'. (That's still a lot of heat, more than 1/6 of your whole-house load.) If the zone load is 7500BTU/hr you'd want to bump up to a 3/4 ton head.
Similarly, the 3/4 ton -KD09NA4 mini-duct cassette is good for 10,900 BTU/hr at sea level when married to an MXZ compressor, but only 0.74 x 10,900 BTU/hr= 8066 BTU/hr at 8000', so if the zone load adds up to 9500-10,000 BTU/hr, bump it up to the 1-ton.
http://meus1.mylinkdrive.com/files/MSZ-FH06NA_For_MXZ_MULTI-ZONE_SYSTEMS_ProductDataSheet.pdf
http://meus1.mylinkdrive.com/files/SEZ-KD09NA4_For_MXZ_MULTI-ZONE_SYSTEMS_Submittal.pdf
It's worth paying a qualified third party to run an aggressive room by room Manual-J on this house that can be sanity checked against the power use numbers. The room by room Manual-J would also allow you to figure out in advance what heads/cassettes might be needed in which zones, using those derating factors, and be able to put together the zoning map and list of equipment you think really works rather than just throwing it out to an HVAC contractor for proposals. If you do the latter I'm sure you'll get proposals with an oversized head in every room, and 2-3x more compressor than needed to keep the place warm.