Using Minisplit for Efficient Heating
THE STORY:
Our home in New England (just north of Boston) had a heat pump system installed this past August to do the following:
1) Heat and cool a new space added to our home and
2) Cool three bedrooms that are already heated by an efficient gas boiler and hot water radiators.
The system installed is comprised of the following:
1 – Outside unit (Mitsubishi MXZ-4C36NAHZ-U1) with Branch box
4 – Indoor units
New space – Kitchen/Dining – Mitsubishi MSZ-FH12NA
Bedroom 1 – Mitsubishi MSZ-GL06NA
Bedroom 2 – Mitsubishi MSZ-GL09NA
Bedroom 3 – Mitsubishi MSZ-GL06NA
THE PROBLEM:
Electric bill went up considerably ($120 additional for same time period as last year). Was not expecting such a jump.
We’ve spoken to a few professionals in the area but are getting mixed information. We are told by some that only running the single 12k head for heat is the issue because of short cycling and offered very expensive solutions. Another said no, the MXZ-4C36NAHZ can modulate down so there is no excess. I tied to check specs myself but I’m seeing conflicting numbers online and I’m way out of my league trying to make sense of this understanding a lot goes into designing an efficient system.
I guess my primary question is does it make any sense to heat a 281 sq/ft space with a MSZ-FH12NA being served by a MXZ-4C36NAHZ-U1?
Any help in understanding what is going on and suggestions on how to resolve it would be greatly appreciated.
Thank you!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
>"I guess my primary question is does it make any sense to heat a 281 sq/ft space with a MSZ-FH12NA being served by a MXZ-4C36NAHZ-U1?"
As long as there's glass in the windows and fluff in the walls & ceiling it's unlikely ANY 281 square foot space in MA (other than a greenhouse) would need a 1 ton cold climate ductless head, summer OR winter. Half-ton heads (or bigger) in bedrooms is usually way oversized too.
As much as I detest rules of thumb, they're useful when more accurate load calculations aren't available. That said...
Typical insulated houses with double-panes (or storms over single panes) in the Boston area would have =a whole-house design load to square foot ratio at 0F outside (colder than your design temp) of 15-25 BTU/hr per square foot if 2x4/R13, 12-20 BTU/hr per square foot if 2x6/R20. There would be varations on that ratio on a room by room basis, but unless it's a lot of glass (or huge air leakage) the design load of the 281 square foot space is going to be no more than about 7000 BTU/hr @ 0F (less when it's warmer, which is more than 99% of the time), and could be as low as 3500 BTU/hr.
>"Another said no, the MXZ-4C36NAHZ can modulate down so there is no excess."
The MINIMUM modulated output of the 4C36NAHZ at +5F is ~6700 BTU/hr (probably more than your design load for that single 281 foot zone), and at +17F it's about 9700 BTU/hr. At 47F it can "only" pull back to a whopping 22,500 BTU/hr, which is many times the heat load of that zone at +47F (and probably more than the whole-house load @ 47F.) See:
https://ashp.neep.org/#!/product/31973
Sure the 12K head modulate down to below it's multi-zone rated 13.6K when the compressor can throttle back that far (or when other heads are using some of the excess compressor output), but it's still going to cycle the head (a lot) due the gross oversize factor, and somewhere in the 30s F outside it will short cycle the compressor, since it can't dump the minimum compressor output on it's own:
https://nonul.mylinkdrive.com/files/MSZ-FH12NA_For_MXZ_MULTI-ZONE_SYSTEMS_ProductDataSheet.pdf
There's really no way to suppress cycling of that compressor even when it's cold out unless you can give it more load (say open up a few windows or something! :-) ), and when it's over 40F outside it'll short cycle the compressor.
In this area unless it's under a flat black asphalt roof or there are huge "sunset view" picture windows the peak whole-house floor area per ton ratio is going to be over 1000' per ton, could easily be 1500' or even 2000' per ton. A half ton head has enough capacity for at least 500 square feet. In cooling mode the minimum output of the compressor @ 82F ( a bit cooler than your % outside design temp) is 9500BTU/hr, a bit more than 3/4 ton, and at 95F (a bit warmer than your 1% outside design temp) it's about 15,500 about 1.25 tons) so as long as all zones are "on" there would probably be enough load to avoid crazy short-cycling in the cooling season, but it's not really optimal.
Design temp source:
https://higherlogicdownload.s3.amazonaws.com/ACCA/c6b38bda-2e04-4f93-bd51-7a80525ad936/UploadedImages/Outdoor-Design-Conditions-1.pdf
The high minimum modulation levels is why I dislike multi-split solutions, but the mis-match of head size to load size is a gross error.
Unless it's the leakiest house in MA you're unlikely to have enough zone load on ANY of your individual zones to not cycle the heads under most conditions, and with zones dropping out it will short cycle the compressor whenever the combination of still active heads can't emit the full minimum compressor output.
The half-ton heads can only spit out 7200 BTU/hr at max speed in heating mode:
https://nonul.mylinkdrive.com/files/MSZ-GL06NA-U1_For_MXZ_MULTI-ZONE_SYSTEMS_ProductDataSheet.pdf
The GL09 maxes out at 10.900 BTU/hr:
https://nonul.mylinkdrive.com/files/MSZ-GL09NA-U1_For_MXZ_MULTI-ZONE_SYSTEMS_Submittal.pdf
Clearly there is no way your bedroom loads are anywhere near that high, so the heads will be cycling no matter what. Multiple heads have to be running simultaneously to not short-cycle the compressor when 40F outside.
To sanity check this, run a room by room load Manual-ish calculations using LoadCalc.net or CoolCalc.com. Note that those tools are likely to overestimate the real loads by 25% or more, even when being fairly aggressive on air leakage & R value assumptions.
Unfortunately, it seems like this system design is really inappropriate.
Ideally, the contractor that put it in would apologize for this big mistake and offer to replace it with a more appropriate system at no charge. That seems unlikely. So what to do?
One option would be to try running all four heads on heating , and using the gas heat less or maybe not at all. That will help the oversizing problem considerably and boost the efficiency. I'm not sure how the total energy bill will compare, but it might be worth an experiment, even just for a week, with keeping track of your electric meter readings.
But it seems that what you really want is a much smaller stand-alone minisplit for the new space, plus a much smaller system for summer cooling in the bedrooms--maybe a mini-duct unit mounted in a closet and feeding those three rooms. The good news is that a smaller system will be cheaper.
I wonder if there is a used market for the oversized equipment you have.
Thanks for replying Charlie. Yes, I plan on getting back to the installer once I have some solid facts that it is indeed the wrong design for our needs. I'm hoping he will do the right thing.
We did try running all four units at once at 69 degrees and turning down the h/w system to 65 and the bedrooms were unbearable to sleep in.
Yes, I already looked on Craigslist to see if there may be a market in the area and it looks like there may be one. It may come to that for my sanity because I don't think I can just let this go.
I'm curious about the experiment with all four running--were the bedrooms unbearably hot to sleep in?
Yes! And dry.
Wow. Thank you Dan for your prompt response. There is a lot of glass (Anderson 200's low E) and a portion has a shallow 9' cathedral ceiling in a section (I'm attaching a photo and floor plan) but the room is well insulated with closed cell foam and very tight. The units in the bedrooms that are off are "exhaling" occasionally. Another sign I guess of wasted energy and money. :-(
Based on what you said I guess we have a bad design and we can with live with it or spend the money to fix it. I've worked hard insulating, air sealing, and rebuilding original windows with custom seals and installing good storms (as per New Life for Old Double-Hung Windows, John Michael Davis - FH Dec 2007) in the original part of the house so this problem is really bothering me. The suggested fixes have been as follows:
1) Replace the MXZ-4C36NAHZ-U1 with two new outside units. One 12K hyper heat to the 12k kitchen/dining space and one 24k to the three bedroom heads for cooling.
2) Just add a 12K hyper heat outside unit to serve the 12k kitchen/dining space and leave the MXZ-4C36NAHZ-U1 for cooling the bedrooms in the summer.
Option one was quoted at 10k. We already had over 16k invested in the system.
New questions:
1) I expect our electric bill will continue to rise as we head deeper into winter. Should we expect the same increases thru summer with cooling or is that a different scenario?
2) What would you do? :-)
The 12K outside unit to match the kitchen/dining space is a reasonable solution. But either of the two solutions for the bedrooms seems oversized. I think that before spending more money, it would be wise to do a careful room-by-room load calculation.
How do I get a room-by-room load calculation? Are they expensive or do I just give it my best shot with LoadCalc.net?
I have a 3000 sqft house, 12 feet ceiling height, no insulation at all, single pane window, heated by several Gree mini splits that are considered to be crazily oversized here. The electrical bill for everything in the whole house for one month in winter is only ~$150 ($0.125/kWh).
How much is the electricity in your area for one kWh? if it is less than $0.2/kWh, your mini split uses >600kWh per month. That's a lot for just 281 sqft space for the setup you have. There might be some issues with installation (low refrigerant, kinked lineset, brazing lineset without nitrogen purge, wrong cable connections) or the unit itself. Mitsubishi MXZ-4C36NAHZ-U1 uses a branch box, which requires some good skills to install properly.
HoverDA - We pay $0.10521/kWh. I had a local HVAC company check the outside unit and they said all the voltage readings checked out (for $180 for about 20 min.). Would there be any other symptoms with the problems you described besides a high electric bill?
Are you on municipal power? Massachusetts has some of the highest electricity prices in the US. According to the EIA the average price is .22 kwh. If that is really your price then it may be cheaper or equal to what gas would cost you.
Sorry, I just provided the supply charge. My total cost is $.2309/kWh.
Could you buy a kWh watts meter to monitor the power consumption of your outdoor unit? And also measure the temperature of the air blown by your 12k indoor unit.
T B,
Who did the design and installation? 16k seems cheap for what you had done around Boston compared to the multiple quotes I've received. I'm outside of Boston and have been planning a new HVAC project. Dana (and other experts) always recommend a 3rd party manual J to be performed prior to installation. Doing it yourself with loadcalc will likely add reassurance but I would rely on someone who had done it multiple times especially given the poor design of your current system. I haven't found anyone in the Boston area who can do this reliably for me. I may reach out to Allison Bailes to do this unless Dana has any other recommendations.
What do you mean unbearable to sleep in? Too hot? Why would you set the gas at 65 and the HPs at 69? That would mean 2 systems running with already oversize HP? Sleeping temps ideally should be right around 65 deg. Have you done blower door numbers? May be helpful since that number would go into your manual J.
Will
Forgot to answer your first question.
The install was done by the plumber that was sub-contracted by our GC on the addition. He did good work on the project and at one point during all the chaos we made the decision to invest in a heat pump instead of standard baseboard heat from the existing boiler to gain the AC. Then we were curious what it would cost to add the bedrooms so we could finally be done with the AC window units. His quote came in at $12,800. We were surprised at how low it was and agreed, fully trusting that he knew what he was doing. I usually research these investments thoroughly but I dropped the ball.
The contract with him was separate from the addition which was finished in April. Due to a series of problems on the plumbers end (broken truck, deaths in the family) and COVID the system wasn't installed until late August AND, he mounted the outside unit on a wall bracket. Big mistake. Glasses were clinking in the cabinets from vibration and the noise was ridiculous. Being so fed up with his delays and mistake we went with a local Mitsubishi Diamond Elite installer to move it off the wall to a stand (see attached pictures), replace the communication line (which was wrong) and recalculate the refrigerant pressure (which was also wrong). We were charged $3,800. A few weeks later, when our first heating bill came in and the outside unit seemed to be running all the time, we had them come back to check it again for another $180 which I mentioned above.
I am now in one of those contractor nightmares. You get what you pay for.
We were advised to not set back the splits because you lose their efficiency so we left them at 69. The gas is set back to 62 at night as per Nest thermostat so it was too hot for us.
Yes, I have had a blower door test courtesy of MassSave but it was done prior to new construction. Results came back 1,636 at 50Pa. I think I would get a better result now since I've done all the window work and added the Harvey storms. I'm interested in Allison Bailes availability/thoughts. Thank you Will!
You might be able to set the minisplits to 62 in the bedrooms--set and forget--and to 69 in the kitchen/dining area. Just as an experiment.
Will try tonight, but even if we can comfortably sleep do you think that will drop the electricity consumption to a reasonable level and offset the gas charge enough justify the current design of the system?
Keep us updated! Sounds like a brutal process
New question:
After a long drawn out series of events we finally have our system running well. In short, we had another installer replace the outside unit with two seperate units and surprisingly the original installer refunded a good portion of the original fee.
The new installer did the following:
Installed an MUZ-FS12NA dedicated to the MSZ-FH12NA for heating and cooling the kitchen/dining area and a MXZ-3C24NAHZ2 to serve the three bedrooms just for cooling (2 - MSZ-GL06NA and 1 - MSZ-GL09NA).
New issue:
After a few months I called the MassSave program to check on my rebates and they said there was some confusion about the larger unit and wanted to know if I had a MXZ-3C24NAHZ2 or MXZ-3C30NAHZ2. I checked the plate when I got home and the MXZ-3C30NAHZ2 was installed mistakenly!
I called the company and they will replace it with the MXZ-3C24NAHZ2 but they said there really was such a small difference in efficiency that I can keep MXZ-3C30NAHZ2 and get the $592 rebate which I would not get with the MXZ-3C24NAHZ2.
Not sure what the correct choice is. Would the MXZ-3C24NAHZ2 be the right choice. My gut says go with the original design and have it swapped.
Any advice would be greatly appreciated. Thanks
A guess - they are about the same unit with different programming and, at the same output, will have the same efficiency. If you go by specs, MXZ-3C30NAHZ2 has a better HSPF, which I expect to be the most important specification.
>" Would the MXZ-3C24NAHZ2 be the right choice. My gut says go with the original design and have it swapped."
Let's take a look:
https://ashp.neep.org/#!/product/26173
https://ashp.neep.org/#!/product/26170
The minimum modulated outputs & efficiency at those minimum outputs is pretty much identical across temperature, so there isn't any comfort or shoulder-season efficiency hit that comes with the 3C30
But you'll note that the efficiency of the 3C30 is HIGHER at mid range @ +17F, so it may in fact the bigger unit will likely prove to be slightly MORE efficient than the 3C24 (but not so much so that you would be laughing all the way to the bank.)
Don't swap it out. If the minimum modulation & efficiency @ 47F were dramatically better it might be worth considering the swap, but they're not. The risk of causing refrigerant leaks in flared fittings goes up every time the equipment gets swapped, and there is no realistic chance of seeing an efficiency gain big enough to notice in the power bill, indeed it's more likely to be a (hard to measure) loss in seasonal efficiency.
Thank you for the quick replies.
Dana - you mentioned efficiency ratings at various temps. Since this unit will only be used for cooling (since the bedrooms have H/W heat) do you still think it makes sense to keep the 3C30?
I'm attaching a shot of the Manual J report showing the bedroom calculations marked in red for more detail.