I had a Mitsubishi Minisplit system installed back in July, for both the purpose of adding AC to my house (crank out windows means no AC’s) and to supplement/replace my forced hot water system that was woefully undersized for the house.
Mitsubishi Heat Pumps, model numbers:
First Floor:
MXZ-3C30NAHZ2-U1 (30K BTU Outdoor unit)
MSZ-FH18NA2 (18K BTU Indoor head)
MSZ-FH12NA (12K BTU Indoor head)
Second Floor:
MXZMXZ-4C36NAHZ-U1 (36K BTU Outdoor unit)
MSZ-FH09NA (9K BTU Indoor head) x2
MSZ-FH06NA (6K BTU Indoor head) x2
PAC-MKA51BC (5 port branch box for distribution)
Initially all was fine they worked great for AC, and then fall/winter hit and they started heating. They work pretty well, though I think placement of the units could have been a little better to distribute the heat a little better. The issue is that they are using an absolute monster amount of electricity. I knew that they would use more than I had been before but I feel like they are using way too much.
Part of this may have been my fault, the installer gave me the passcode for the “Installer” settings on the Kumo Cloud app and I have wandered through the settings there. Changing what temperature to switch over to the old heating system, etc to try to improve overall performance and reduce power consumption.
Side note I feel like there should be a way to automatically run both the old and new heating system at the same time at a certain temperature band then only the old below said temperature. I feel that would be more efficient than running the new system and if nothing changes after X time run the old system.
Does anyone have any suggestions as to how I can get the most effective and most efficient heating out my system?
The main issues that I seem to be having are as follows:
First floor: the 12K unit seems to keep the room it’s in about 4-7 degrees (Fahrenheit) above the set point. While the second unit seems to struggle to keep the room it’s in at temperature. The room it is in has a vaulted ceiling with an opening to the second floor so I know heat is lost to the second floor most of the time. I do keep the second floor a couple degrees above the first floor to try to help with this. I also have a ceiling fan to try to keep the warm air lower.
Power consumption: this change was an attempt to save money on heating the house, in combination with Solar panels, but so far compared to last year using only oil to heat to basically the same temperature in the house, I have spent as much if not more… For comparison, in all of 2019 I spent $2,100 on electricity and $1,900 on oil, so far this year in Jan and Feb I have already spent $1000 on electricity and $900 on oil… This is with the addition of Solar panels, and the “more efficient” heat pumps…
I understand last year was a very mild winter for me and this year is a lot colder, but it still seems like something is wrong.
I have asked the installer company about it and they suggested playing with the “Only heat pump” settings, the “Delay” and the “Only backup” settings, as well as the ceiling fan speeds.
Recently I also discovered that the 36K unit even when off draws 2A (measured on amp clamp meter) and the 30k draws between 0.5-1A on that same meter. I know power factor is probably in play so it’s not a true 2A but that’s still a large amount of power.
Replies
I’m giving your question a bump. It’s a good one, and has come up before. It is possible your system is oversized. Read through this Q&A thread thread while you wait for the experts to weigh in. There are some pertinent comments and links to multiple resources you might want to check out.
Sounds like another oversized multi split system. There have been numerous posts here with people with Mitsubishi multi splits having basically the same issue and getting huge electric bills.
Large multi split units have nowhere near the turn down modulation and are much less efficient than single 1 to 1 units. Many single zone units can turn down to like 1k or less btus. Multi splits usually have a 7 ,8 or 9k btu minimum output.
Even some of the state efficiency programs aren't giving out rebates on these multi splits anymore. I know someone in Maine that said their installer said they dont install multiplits anymore and even Efficiency Maine is now only giving out incentives to single zone 1 to 1 units with a hspf of 12.5 or higher.
What size and where is your house? 66k btu is a lot. I'm using 2 12k units in New England for my year round heat/ac in my 1800 sq ft house.
I have heard from another forum (though they were rather hostile about it and suggested tearing it out and just beefing up the hot water) that it may be oversized.
I'm in MA, and oddly they won't provide rebates for larger multiple single splits, they prefer a large unit with a splitter box, though it's confusing since with the splitter box I'm not sure if multiple units can run at once meaning it should be paired with a smaller outside unit right?
My house is rather large (2,800-2,900 sq ft) and does have a lot of things going against it, multiple large windows (21 windows all on the larger size) a few sky lights, high ceilings in about half the square footage.
I ran through an online Manual "J" calculator and came up with 5 tons, so not far off from the 5.5 installed.
Side note the Mitsubishi Kumo app is absolutely terrible, crashes, can't connect, etc.
The splitter (branch boxes) allow multiple units to run simultaneously in the same mode (heat/cool) but most residential equipment does not allow both simultaneous heating and cooling to occur. (so in the winter put everything into heating mode, and vice versa in the summer).
Thanks for the clarification, they're always all in the same mode, and usually set to similar if not the same temperature range.
IMHO a HSPF of 12 should be the baseline in 2020.
I'm glad they're tightening the screws on this. Low efficiency multisplit (due to sizing, or equipment factors) installs are giving a fundamentally good technology a bad reputation and creating disappointed customers.
I agree. Didn't you post with a similar issue in Massachusetts? Seems like there have been a ton of bad multi split installs in Massachusetts lately.
As I posted earlier it seems even some of the state efficiency programs are realizing these multi splits are not as great as they seem. Efficiency Maine requires single zone units to get the $1500 state rebate. Multi splits only get tier 1 $500
Tier 2 heat pumps:
AHRI-rated HSPF 12.5 or greater
Each system is single-zone
Wall-mounted indoor unit
Installed on or after 1/1/2020
Home not served by natural gas
Somewhat of a followup:
I did some data collection and ran the numbers and... for all of 2019 before getting Solar, and before the heat pumps were installed at all I spent a grand total of $2,100 in electricity and $1,900 in oil.
So far this calendar year I've spent $1,000 in electricity and $900 in oil, I have played with the changeover temperatures and such to rely on oil more at lower temps and such. Things do not bode well for this to save me anything at all, in fact it's going to cost me twice as much, that's with a Solar offset lessening the electric bill a little bit.
More data collection, I do not believe that it is an oversized heat pump system. Well it might be overall, but I don't think that's what is causing the issues.
I got a power meter and looking at the larger unit with the branch box it is in an almost constant ~20 minute cycle.
The weird bit is that none of the heads seem to be running in that timeframe, none of the heads have double LEDs on, and none have the fans on, and through the app the temperature readings are satisfied. Sometimes one or two of the heads feel like they are warm (Can feel heat near the vents) but that's about it.
I attached a couple hours of cycling and then a more detailed ~45 minute cycle graphed out.
Thoughts?
How is the thermometer setup? Some people have found that getting the thermometer setup correctly paid dividends:
https://www.greenbuildingadvisor.com/question/mitsubishi-mini-split-behaving-very-differently-with-external-thermostat-vs-without
I seem to remember that an external one was better than the one in the head. Can't seem to find the thread, though.
Currently I only use the internal thermostat for the upstairs.
I will be looking into external ones, though I do have fans in all of the upstairs rooms pushing air upwards, and they are all enclosed rooms so it should "pool" and maintain temp I would think.
Your first graph definitely looks like short cycling. If none of the heads are running that are connected to it then there is something major wrong with it. How much power is the peak on the first chart?
Remember when the outdoor unit is running it is sending hot refrigerant to every head regardless if your using it or not.
For reference here is what one of my mini splits look like when its modulating. It never turns on or off just ramps up or down to the load.
I suspected there was something majorly wrong since the I would sometimes watch diligently to see if any of the heads were running and what their temp readouts in the app were and for the majority of the time they all read at or above the setpoint. For at least two of these cycles they were all reading at least one degree over the setpoint.
The peaks are all about 2.6-2.8kW, there were a couple where it was higher than that, but only when there was actual draw on multiple heads.
I just took a look at the NEEP cold climate site for your unit. It lists the minimum btu output as 22k at 47 and 10k at 17. Perhaps someone else can chime in but it would seem the MXZMXZ-4C36NAHZ-U1 is ridiculously oversized connected to 2 9k's and 2 6k's.
https://ashp.neep.org/#!/product/31973
See information like this makes me question why there aren't settings to run heads in parallel, that 10k is not far off from the 9k heads and there have been multiple ~17 degree days where I have drawn virtually the same as the warmer or cooler days.
Sounds very frustrating. These units do a bad job of providing data to diagnose these sorts of issues.
One thing to look at is whether you have the indoor units set to "thermal off" (ie stop the fan when set point reached in a zone). Different units require different approaches to this (and the KumoCloud app sometimes shows the option even for units that don't have that as a software settable option, requiring a hardware "clip").
I have found the KumoCloud wireless sensors to help a lot, they have run without trouble. Source them online to avoid the silly markups and shipping delays in going via HVAC supply houses.
For your year to year comparisons, can you look up actual heating degree days and weight accordingly (ie divide costs by hdd), that might give some peace of mind.
They are all currently set to "Selectable in Cool and Heat" which doesn't really tell me much as to its operation.
I do have a wireless thermostat for the downstairs since they both work together to heat/cool the first floor as whole.
I thought since I had fans the temperature would be more stable since there is constant air movement throughout the rooms so the hot air wont just stagnate on the thermostat.
I will look into them, cheapest I could find with a quick search is $55 each plus shipping, I may push back on my installer and see if they can provide me some since this system is only 9 months old.
Quick googling hasn't resulted in much information, any suggestions as to what they should be set to?
I kind of have a sense of what they do but any advice would be appreciated.
My interpretation:
Extra Low in Heat: Leaves the fan on always when it is in heat mode extra low when satisfied but to the setting when calling for cooling
Stop in Heat: turn off the fan when it hits temp in heat
Stop in Cool and Heat: Same as above but for cooling also
Stop in Cool: Same as above but only cooling
Selectable: determined by the remote setting (more on that later)
Ext. Lo in Heat. Stop in Cool: same as first and stop in cool combined.
Relevant to Selectable from what I have observed they are always set to Stop in Both and I see no clear way to change that in either the app (other than deep in the installer settings) or through the remote.
Intuitively it seems like I would want the last option (seems oddly specific) or the first option to keep using some of the residual heat in the coils to continue to heat the room etc.
Your unit is definitely cycling.
Since you mention that some of the units are feeling warm while off, it also means that the heads that are on are too small for the minimum capacity of the outdoor unit so that it needs to bypass refrigerant through the other heads. This makes sense as the min capacity on the unit is 22kBTU ( http://enter.mehvac.com.s3.amazonaws.com/DAMRoot/Original/10006/M_Submittal_MXZ-4C36NAHZ.pdf ) and you have a bunch of small heads on it.
There is no easy answer here without changing equipment or relocating some of the zones to add load to the outdoor unit. If I had to take a stab at it, a reasonable solution is to ditch the main floor outdoor unit, move the main floor heads onto the 4C36 and replace all four wallmounts upstairs with an 18k ducted unit. Not simple or cheap.
Each individual head is too small, yes, so it will run some refrigerant through at least one of the unused heads while another is running, same thing happens with the first floor, the larger unit will run and the smaller will have some heat coming from it even though it is off.
The issue is that there will be no heads calling for heat when the system runs pushing heat to these heads.
The first floor works fine, and uses relatively very little power (~17.5% of my total usage on Mar 7th, vs ~64% consumed via the second floor)
And unfortunately there really is no way to put in ducted units to one of the rooms on the second floor, there is no attic space connected, three walls are exterior and it has a pitched ceiling following the roof.
Also whatever the fix is I will only be paying the difference in system price since this was installed by a company in July last year and if they messed it up it's on them.
I don't think your understanding the 22k btu minimium capacity that both me and Akos pointed out. When your outdoor turns on it starts at that as its minimum output. For comparison a single 12k Mitsubishi can usually start out as low as 1-2k btus. Those are the 2-2.5 kw spikes your are seeing. If those 6 and 9k heads are in bedrooms on the 2nd floor the heat loss could be as low as 1000 btu/hr in each bedroom on a mild day. Those four rooms would only need 4k btu but your unit turns on and provides almost 6 times that at its minimum capacity.
To give you a comparison I use 2 12k cold climate splits just south of you in CT for my full time heating and cooling. On milder days like today my 2 12k units are only putting out like 1-2k btus each drawing around 150-200 watts to heat my entire 1800 sq/ft house.
The only possible easy solution might be to see if they could install a condenser that has a lower minimum capacity
I do understand the issue it's just that I'm not so much concerned about the overall power usage when it's running the concerning bit is that it is running as frequently as it is when the setpoints are satisfied.
For example, the second floor temps (per the app) were all set to 68 and were all reading either 69 or 70 as their current temperatures, the system started ramping up and "ran" for about 15 minutes before shutting off again, none of the head units kicked on to deliver heat, one ran on low for a few minutes. (I adjusted thermal fan off to " Extra Low in Heat")
Then it shut off again, sat idle for 10-15 minutes then ramped up again, this time all the heads were again at 69-70 but one that was at 68 (its setpoint)
If the second floor unit operates anywhere close to the first floor when it works properly I should be saving money every month on heating vs the oil system, I just need to get it functioning correctly.
These units don't work like a traditional thermostat connected to a furnace. As Akos pointed out there can be a 2 degree up or down range that will kick them on or off. Your 68 setpoint might have a 2 degree range up or down the units will maintain. My Midea mini split actually has a switch to adjust the offset in 1, 2 or 3 degrees increments.
One of your heads is definitely calling for heat. Have you tried setting them all really low to like 4 degrees below the room temp?
Yesterday evening every time it started ramping up I would go and look at all the head units, and the one that I suspected was the culprit (had both LEDs on a couple times that it was ramping up) I shut off and it did immediately stop drawing power. BUT shortly after that it would ramp up again, this time all turned on heads were showing 1-2 degrees above the set point and none of them had both LEDs on.
As to the temp offset, there is a setting called "Display offset" and they are all set to 0 degrees but that is the only setting like that that reads to me like that, though it may be hidden internally in the system.
That being said it ramps up at the set point and stops running while still at the set point, so that argument doesn't sound like it would be the issue, since when it ramps down all fans are off on the heads, so it's not like it satisfied and it shutting off in a short cycling situation, its more like the system doesn't know what to do.
I have an oversized multi split that I use only for cooling.
What I have seen is that no matter what setpoint I have the head set to it draws about the same power (~500W for a 9k head, 600W for a 12k head). It also seems that no matter how much cooing the head is doing it is still sucking that power until the indoor temperature gets about 2F bellow the setpoint, at which point the compressor shuts off.
This means that if the head is doing 9k of cooling or 4k of cooling, it is consuming the same amount of power, thus the efficiency sucks. I can definitely confirm this, a ultra budget one-to-one mini split providing similar amount of cooling uses about 1/2 the energy.
The only way to get efficiency out of the multi split is to match it to your load. 4 bedrooms would be around 12000BTU, so you need a much smaller unit. The problem there are no small 4 head multi splits out there, so you are pretty much stuck.
You might be able to get some better efficiency out of the whole setup if you move some of the smaller heads to the 3C30 and move the 18k head onto the 4C36. This also could result in both units short cycling, so it it is hard to tell if it will be better overall.
There is not enough technical information out there from any of the manufactures. I would not blame your installer for the setup, most manufacturer's brochure tout the one wall mount per bedroom setup. This somewhat works, it just doesn't work well.
Another update.
Over the last 45 minutes the second floor unit has ramped up and down twice, though ALL heads are at least one degree above their setpoint and as far as I can tell none have even pushed any heat out.
Crazy thought, Is it possible that some how the upstairs system is listening to the downstairs thermostat, even though the upstairs heads aren’t calling for heat. Sounds like you’re running two heads off of one stat down there and the assigning of heads gets confusing. (I think Mitsu controls are terrible... sorry) It would be weird, but running through head assignments might be worth while.
I would believe it, but the sensor was disconnected until Tuesday last week when a tech came out to check things out (though dispatch said I "hadn't used the app at all and needed help" rather than the actual issues)
The only place I see anything from the sensor is from one room on the first floor.
The bottom line is this setup is never going to run efficiently. Even if you figure out the random turning on and off the 22k minimium output is still absolutely insanely oversized for a bunch of bedrooms. Your unit is just going to cycle on and off drawing 2.6-2.8 kw. You are getting zero modulation. Its running like a regular off on heat source and its going to cost you a fortune to run as you already found out.
To give you a perspective its 51 degrees right now in CT nd both my 12k splits are running at their minimiium speed. The current heat loss of my entire house is probably around 2k btu/hr right now. If your unit was doing my entire 1800 sqft house when it turns on its minimum it would be oversized by 20 times my current heat load.
The heating load for 4 2nd floor bedrooms on a day in the 40s or 50s like today is tiny. Your 4 bedrooms might only need like 1-2k btu/hr. Your unit is oversized by a factor of 20 on a day like today.
I agree that it seems oversized. But if it runs in a more normal cycle, similar to the first floor, then it will save me money over heating with oil. The goal right now is to get it running properly then compare with oil heating costs and go from there. If there is still a large cost increase over the oil then I will petition the HVAC company to remove the outside units and replace them with individual units or at least something more efficient.
Today neither second or first floor needed to run at after 7am or so, theoretically my solar panels will actually generate enough to offset the heat pumps today.
Hi GreenGiant117! This was a super helpful thread for me to follow - I'm seeing similar oversizing issues with my setup as well (https://www.greenbuildingadvisor.com/question/is-my-mitsubishi-heat-pump-short-cycling).
I'm curious whether you were able to solve this puzzle, and if so, what your solution looks like.