Minisplit heating system for high altitude
I would like to relate my recent experience and outline the present configuration and performance of my minisplit heating system to solicit comments on whether it is reasonably sized and balanced, and how I might improve it. We bought this house 5 years ago and are trying to improve it as we can. It was terribly leaky and we have air sealed everywhere we can. It is fairly tight now, certainly not great, but enough to get the interior humidity up into the 40+ percent range during the winter.
I have a 3 year old retrofit Mitsubishi H2I (MXZ-8C48NA outdoor unit, MSZ indoor heads) system in my 1250 sq ft ranch style house, built in 2000, located at 8100 ft in the Colorado front range. The climate zone seems to be generally identified as 5 by county, but the location is nearby high mountains identified as zone 7. The house is 2X6 construction with FG batts, and the windows are new Pella Impervia. The bedroom half of the house has a conventional attic with 16 in of cellulose, and the front great room has a vented cathedral ceiling with R-24 FG bats. The house sits on a full basement with about 1/2 of the perimeter fully buried, and the other half gradually exposed to a walkout at one end. The walkout end of the basement is finished with 2X6 walls and FG bats. The climate over the last 5 years has been similar to the front range plains, but generally 10-15 degrees C cooler. We have a lot of wind from the late fall through spring, often 30-40 mph, and gusting to 50+ for days.
The outdoor unit is 48K feeding 3 indoor units upstairs: 9K and 12K in the bedrooms, and 18K in the front great room. The basement has a 12K unit in the main room at the walkout end. The system was sized by eyeball by the HVAC installer company before I learned much about such things from this site. I used to think it was significantly oversize based on what I have read, but I recently learned several things that suggest it may actually be about right or even a bit small. Experience over the last three years has been very good, and we stay at 70 degrees F overnight when it goes down below 0 degrees F, and even managed 68 overnight on a -17 degree F night. When it is predicted to go below 0 degrees F at night, I turn off the basement unit to maximize heat for the main floor. I recently wanted to put in an additional 9K unit in the basement so I could warm the back end to 70 during the winter (my workshop). Typical temperature in the basement without heat is low 60’s, down to upper 50’s in very cold weather. The Mitsubishi Diamond dealer I contacted thought this would work since the Mitsubishi manual for the outdoor unit said the outdoor unit would accept up to 62K total indoor unit capacity (130% of the outdoor unit rating). But when the HVAC contractor checked with Mitsubishi to determine the appropriate total refrigerant charge, the answer was very different. They said that the system was already maxed out because the outdoor and indoor unit ratings were derated 24% due to altitude, and further due to the total lineset length (about 140 ft total for all the units). I did not find any of this in the unit specifications in the install manuals. Considering this, it seems I actually have a 36K system or even less (the derating for the lineset length is unknown). If I add up the heating capacities for the upstairs indoor units (derated for altitude), they are close to the derated outdoor unit capacity as that capacity drops below 10 degrees F (down to about 80% at -10 degrees F), which happens 5-10 times per winter. Does this add up to a reasonable system balance or design point? It seems to work well in practice. I realize we are designing by doing rather than computing the heating requirement, but this is where we’ve ended up and it seems to be working.
Apparently, to add additional heating capacity for the basement I will need to add an additional outdoor unit, probably a 24K unit derated for altitude to 18K, which will also handle the current basement indoor unit since I have found that it would be better to split the heating up between the upstairs and the basement. In the shoulder seasons we often wish we could have the basement on heat while we have cooling upstairs to counter high solar heat gain in the late afternoons. There is only one location around the house where I can shield the outdoor unit(s) from severe winds. Can I locate the additional outdoor unit close to the present one?
We have not found a HVAC or heating in the Colorado front range that actually seems to know the engineering of minisplit heating systems. I have low faith in whether the system I have is properly charged, and am having trouble actually getting any contractor to measure what is in the system now and determine what it really should be. I can never get any actual numbers, just arm-waving declarations, and Mitsubishi technical won’t talk to me directly. Can anyone recommend a good HVAC contractor/engineer in the front range that actually understands these systems? One further question I have no idea about is how all the derating affects efficiency, and how to optimize? All I presently know is that it works better and is cheaper to run than the previous propane heat. Any comments or guidance would be appreciated.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Are you sure your outdoor unit is a MXZ-8C48NA or do you actually have a MXZ-8C48NAHZ? There is a huge difference in the cold weather performance between those two models.
Also, what is your current branch box setup like?
>"There is a huge difference in the cold weather performance between those two models."
No kidding- it's not even subtle:
https://ashp.neep.org/#!/product/29012
https://ashp.neep.org/#!/product/28986
At 8000' most air-source heat pumps need to be derated to about 75% of the nameplate capacity at sea level. See Table 1:
https://s3.amazonaws.com/greenbuildingadvisor.s3.tauntoncloud.com/app/uploads/2016/09/08060512/Mitsubishi%20M%20and%20P%20series%20Altitude%20correction.pdf
That would make the -8C48NA good for about 24,000 BTU/hr @ +5F, compared to about 40,000 BTU/hr @ +5F for the -8C48NAHZ.
At typical tightened up 2x4 framed 1250 rancher with code min windows would a heat load of about 20,000 BTU/hr @ +5F if the foundation is insulated, about 25,000 BTU/hr @ +5F if the foundation insulated. For a walk-out basement it's more like 30K (insulated foundation) , 37K (no foundation insulation.
So if it's the NA you might start running out of capacity even at +10F, but the NAHZ should be good into negative single-digits.
Run a formal heat load calculation on the basement before picking a compressor.
It's fine to install the compressors side by side or one above the other, as long as they beat the minimum clearances in the specs. Don't install them perpendicular to one another at close proximity to avoid capacity defeating interactions when running at high speed.
BTW my name is Bill. Thanks for the review and comments. I am sorry I accidentally mislead - the outdoor unit is an MXZ-8C48NAHZ. The branch box is a 5-port unit in the basement. It is about a 40 ft run from the outdoor unit to the branch box, almost entirely in the basement. The indoor unit linesets run through the basement until they get to the plan view location, then go out the rimjoist, up the side of the house, and back in to the indoor unit. The foundation is not insulated, and only the foundation walls that are exposed outside have interior insulation. Since I have 3 years of operational history, I thought it could guide the evaluation and expansion of the system. Apparently we have one of only two such systems up here, and we're 500 ft higher yet than the other.
>"The foundation is not insulated, and only the foundation walls that are exposed outside have interior insulation. "
At your cool altitude & location it's well worth insulating the foundation walls all the way up to the foundation sills to a code-min R15 continuous insulation even where it's not above-grade, and even including spaces not normally actively heated. A poured concrete wall is about R1, quite thermally conductive compared to even an inch of rigid foam.
If that's impossible to do without a major gutting of the place it may be easier and less disruptive to install 3-4" of rigid EPS foam on the exterior of the foundation down to at least 2' (4' is better), which would make a significant difference in the heat load of the lower level.
Current code min would call for 4" of EPS all the way down to the footing, which is easy enough to do on new construction using insulated concrete forms for the foundation pour, but a major PITA as a retrofit.
A 12K head isn't sufficient to cover the heating load of a basement with no foundation insulation, but even derated for altitude it should pretty much cover it once the foundation is insulated. Again, running a Manual-J on the before and after insulation condition of the lower level should be able to determine that.
You may have been the 2nd install in your area at the time of your install, but I doubt you are anymore...and definitely not for long.
Checking the amount of refrigerant in your system isn’t any different from “traditional” HVAC system. Any competent technician can vacuum out your refrigerant charge and weigh it. Weighing out the charge is time consuming so maybe that’s why there was some reluctance.
Calculating the required charge is also not rocket science. If you follow the instructions (below, section 5)
https://meus.mylinkdrive.com/files/MXZ-8C48NA_MXZ-4C36-5C42-8C48NAHZ_Install_RG79D595H05_03-15.pdf
this handy video hits on just about all the variables.
https://youtu.be/RiBjwQbpFe8
Although you can install up to 130% indoor capacity(in most scenarios) that doesn’t mean you get a “free” 30% more heat. So if your indoor units were rated at 130% and ALL turned to max the outdoor unit would only supply 100% and all the indoor units would have to share and not operate at full capacity.
You have one more set of ports available on the branch box but it seems you are maxed out, based on your info.
As you said “seat of the pants” feel is that your system is a pretty good fit...so that’s good.
BTW: Mitsubishi makes a wind guard and I have seen some even more effective looking wind screens online.
Dana, the basement insulation is on the to-do list. It will probably be interior. The 2nd basement head was partly for heat distribution since there is a separation of front living spaces with rear workshop/storage.
FluxCapacitor, it would be great if there were more systems like this up here. It actually works great. My information on only two systems is based on the nearest Diamond Dealers who proposed on the additional head. Neither caught the problem. I was clear on the 100% capacity, hence turning the basement heads off in really cold weather. The surprise has been what that 100% actually was. I've computed the neccessary charge based on the install manuals, but wanted a check through Mitsubishi because I think the original charge was mis-computed and may be several pounds low at 16.3 lbs. Thre is much reluctance to bother to check. I am just told the pressures look 'good'. This is frustrating because I just want to confirm it is right.
I still don't have an explanation for additional derating due to the length of the linesets. My outdoor unit is tucked into a corner between the garage and the house and screened by a very leaky, slatted cedar enclosure exceeding Mitsubishi's specs, but buffering the gusty winds. However, there is not enough room for an additional outdoor unit except on top of the present double fan unit - maybe that is what I need to do.
From my understanding the high altitude de-rating is the similar to temperature effects. In hot weather, you get less BTU out of the unit, the same as higher altitude.
The 130% value is for the refrigeration side, it has nothing to do with high altitude de-rating. It is there to make sure the unit can flow enough refrigerant for the heads to work properly.
Adding in the extra head, won't change anything but since your compressor is maxed out, you won't get any extra heat into the house. The setup would still work but now the extra head will be stealing some of the heat that could be going elsewhere.
The extra line length adds restriction to the refrigerant flow, thus slightly reduce capacity on the head with the long runs.
It’s impossible to know the refrigerant charge by line pressure. Refrigerant must be pulled and weighed to know. It’s a PIA.
Mitsubishi does not recommend cold weather units be installed above/below each other.
https://meus.mylinkdrive.com/files/Cold_Climate_Installation_Guidelines.pdf
If you mount new unit above you better have a really good strategy for not allowing the defrost condensate to foul or freeze up the lower unit.
> interior humidity up into the 40+ percent range during the winter.
This isn't good for your house. How much so depends on the design details.