Mitsubishi Ducted Heat Pump Short Cycling
I have a Mitusbishi SVZ-KP30NA & SUZ-KA30NAHZ (hyper heat) installed at the beginning of 2023. The unit seemed to by cycling on/off a lot, so I bought an Emporia Vue 2 energy monitor to gather the attached data. The energy monitor confirmed my suspicions; the unit is generally kicking on to a fairly constant ~2500W load then kicking off with very little fluctuation in energy use. Shouldn’t the energy use be continuously changing as the unit tries to modulate to maintain our tstat set temp? I understand the unit may not be able to modulate low enough when outdoor temps are >45F, but I was expecting this thing to be able to run constantly when it is <35F outside. Based on the energy use, this things is behaving more like a single stage furnace. From the attached Mitsu Partial Load Capacity Chart, it appears the energy use should fluctuate from ~500W – 3800W when outdoor temps are 20F – 40F. Is the unit not acting properly, or am i misunderstanding how the modulation aspect works as it relates to energy use?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Are you using the Mitsubishi wired thermostat or their PAC thermostat interface module and a standard thermostat?
Using the Mitsubishi wireless interface with Mitsubishi tstat. Part numbers are below.
MITSUBISHI MHK2 WIRELESS REMOTE KIT
MITSUBISHI PAC-USWHS002-WFI
KUMO CLOUD
I haven't used the MHK2 but for their wired units you have to configure which temperature sensor the unit uses for sensing room temperature. This should be set to the sensor in the remote not the one in the intake (which is the default). It is somewhere in the installer options.
Strange, that should work well enough. I know with Midea units setting the fan to anything other than Auto causes the unit to cycle a fair bit. You can try adjusting the fan settings and see how it runs.
The fan is an auto on the thermostat. We actually had a service call a couple weeks ago because the system was blowing cool air more often than it was blowing warm air. The tech said there was a factory default setting that was basically causing the fan to run 24/7 even though the thermostat fan was set to auto. The air felt cool because it was basically circulating 67° air with no added heat for 85% of the time. He changed something in the programming. Now the fan does not run constantly, but that is when I noticed the unit short cycling frequently and running for short, five or 10 minute periods. I purchased the energy monitor so I could see specifically when it was running to help better understand what was going on.
You can try setting the fan to low and see how it runs. I can't see anything obviously off on your setup, it should run reasonably well. The dual zone Mitsubishi unit with a ducted air handler I checked power use on did ramp down to near spec min (~700W).
Good to hear that it at least should work the way I thought. My contractor is a Mitsubishi Diamond dealer, but I think it was the first time the tech who did thr startup had ever set up one of their wireless tstats so perhaps there is some programming component that isn’t set up right. They’ve always been good to work with, though. I sent them the info today and will give them a call on Monday.
Just to make sure, you don't have some home automation that is messing around with the temperature setpoint on the unit, the target temp is simply set on the MHK and the system allowed to run.
My understanding is the same as yours - these are supposed to modulate, and thus you are not supposed to have the spikes you are seeing (and based on kw usage, that I expect mine has).
As far as t-stat settings, download the install manual so you can adjust these settings yourself. Mitsubishi makes unbelievably complicated products that apparently work as advertised if you get everything installed and set right, but they give "Diamond Elite" status to companies who might have one person who went through the training, and then someone else installs the unit. It's a mess and noone will care as much about this thing working properly as you do.
I presume this thing qualified for the tax credits? And since this is the higher static pressure unit, you were able to use your existing ductwork? In general I thought these higher static pressure unit with a traditional type air handler and outdoor unit heat pumps. Are these units technically mini splits?
Yes, Mitsubishi makes really high quality stuff but I’m continually amazed at how complicated installations can be. One-to-ones are usually simple. But multiheads and ducted systems can be a bear.
And what’s just as troublesome is that the product line is extremely complicated and difficult to navigate, with wide variations in performance, sometimes where you don’t expect. So knowing what equipment to recommend for a given situation is sometimes more difficult than installing I think.
For ex a one to one muz-fs**/suzfs**nah will maintain 100% output down to -13 and has a great turn down ratio—maybe 5:1. Put the same fs indoor unit on a multihead, and suddenly the turndown is 1:2 at best and might even disappear completely. A ducted hyper heat svz-kp**-ka**naz system will hold its own down to 0F or so but if you read the fine print you’ll notice COPs are low (esp 24 and 30) and some are not energy star, and the performance drop below 0 F is precipitous. So that’s probably fine in a climate like PA/NJ or if you lean on a strip heat kit but in climate zone 6 and North a client might actually be better off with the p-series commercial pva-a**aa7/puz-ha**nka.
And the published data is mostly optimal best case—output and efficiency derates for lineset lengths and bends can be quite surprising.
I’m all for electrification but I think it’s going to be a bumpy road.
Yes, it is basically a mini-split. The outdoor unit is the exact same unit Mitsubishi sells with their ductless mini-splits. I just have it paired with their indoor air handler instead.
Great to see this kind of data. I agree, something is not right here. You’d expect to see some sign of turn down and what you see instead is pretty much bang on/bang off and no modulation.
First try cycling power off and on again if you haven’t already. Sometimes that resolves strange electrical and control issues. Leave it off for a good ten minutes or so.
I’d look at the thermostat location—best not on an exterior wall and certainly out of direct throw from a register.
Try what Akos says—put the fan speed on low on the mhk2 and see if you get lower energy consumption. In theory that should limit the compressor and you should see lower wattage.
The installer programming on these ducted units is surprisingly complex. The only (far fetched) possibility I can think of there is that maybe the hysteresis is somehow tricked out so that the on condition in heating is substantially below set point and then the unit responds by running full out to come up to temp as rapidly as possible? Not sure that’s even possible. Mode 24 in the install settings is the only possibility I see, oddly translated as “heating height offset” if I interpret right. Not sure what the default is there either.
More likely I’d look at the possibility of refrigerant under/over charge. Not sure how a kink in the lineset would manifest on an energy monitor either. You might hear that—squealing and other weird harmonics in my experience.
I’d start with the basic tests—whether you have temp across the coil and whether static pressure is within range. If those check out, then it’s unlikely to be the installation as far as I can see.
Good luck, let us know if/how this gets resolved. Im curious.
Couple things.
The unit is either severely oversized or the thermostat placement is bad. Your energy consumption pattern looks about right for the above- mentioned context where the thermostat gets satisfied very quickly. The compressor ramp up looks right, but is cut too short. If you set the temp to 80 for example (but you need to manually set fan to low) so the unit runs longer you will see that the unit will back off a bit after 15 mins or so and then you might see some wavy pattern/ modulation. But it doesn’t play out as settemp is reached quickly in your use case and heating cycle is cut short. It takes anywhere between 15-30 minutes for the unit to get into cruising mode and start stepping slowly up/down the compressor depending on indoor/ outdoor conditions (ie modulating)
Changing the fan from auto to low will Not make a difference as it is running on low anyway in your day to day operation most of the time.
My suggestions are:
1. Rebalance registers open/ close so the ones close to the thermostat are closed a bit or completely.
2 move thermostat away from registers.
3. Install a remote temp sensor away from registers.
You need to increase your run times.
One last thing- at 45f outside (especially with some heat gain) most well insulated houses have very low heat load and you are guaranteed to short cycle in most of the scenarios.
Good luck
Deleted
Seems like we read this complaint from different posters from time-to-time without a resolution. The one thing that seems to be common is the Kumo Cloud consider disconnecting it.
Walta
Walta, Kumocloud is usually associated with connectivity latency issues ie overshoot settemps. I believe in his case this is not the case...
But of course it never hurts to disconnect it (if present) to rule out connectivity issues.
How do I disconnect the kumo cloud component? I read thru the MHK2 install manual but don’t recall seeing anything about how to disconnect kumo cloud.
Your unit looks way oversized especially in milder weather due to its high minimum capacity. In your chart you can see in mild temperatures from 45-65 the minimum on that unit is 11-17k btu/h. Looking at your heat loss data your are going to be vastly oversized and cycling in mild weather. A 2400sqft place with a max heat loss of 33k btu/h in a cold place like Lincoln NE is not going to have a anywhere near a heat loss of 11-17k btu/h on a mild day in the 40's and 50s. Notice the only day you were running at a steady modulation was on the 18th which was the coldest day in the teens on your charts.
Also, the NEEP data is showing this unit can really only turn down to 50% of its max capacity in cooling and 47 heating. Your unit is also rated at 2.4 kw at 47 which lines up with your units 2.5 kw peaks on your chart. So your units appear to be just cycling on off at its max rated capacity even when its in the 30's as its way oversized. Since your monitoring the electricity at around what temperature does it start modulating steady?
https://ashp.neep.org/#!/product/34582/7/25000///0
I understand the system will short cycle during the swing seasons and there is nothing we can do about that. When daytime highs get around 50F, we typically have enough internal heat gains that by late morning we can switch the unit off altogether.
I've only had the energy monitor installed since March 15, so the data I have posted is everything I've got to date. Unfortunately, our 10-day forecast shows highs in the 50s - 60s, so I may not get another cold snap where outdoor temps consistently stay below 30F.
I had the entire system disconnected from power for a couple hours yesterday afternoon while I swapped out the cable between the indoor and outdoor units. I installed the electrical for the initial installation myself (I worked for a commercial electrician for 12+ years, so yes, I am qualified) and used a 30A/3P motor-rated switch as the service disconnect for the fan coil unit. The tech who commissioned the heat pump recommended that I eliminate the switch and run a single uninterrupted cable between the indoor/outdoor units, as he said Mitsubishi does NOT recommend breaking these wires with a switch because it has the potential to cause communication issues. To rule this out, I just went ahead and replaced the cable.
@greenright,
Our tstat is located in the hallway on our upper level. Nearest register is 10' away inside a bathroom, so there shouldn't be any air blowing directly on the tstat.
I went through the installer settings in Kumo Cloud and have posted an attachment with our current settings. Light blue is how the system was set up. The yellow are settings which I adjusted this morning. I changed the humidifier (we don't have one) and Indoor Temp Detection (changed from '1 - average data' to '3 - remote controller' so it should now only be using the readings from our tstat). Anyone see any settings that appears incorrect which would cause problems?
The last page of the attachments are some screenshots from Kumo Cloud, as there were a handful of Kumo Cloud settings where it wasn't blatantly obvious which programming mode/function each was associated with. If anyone has experience with these Kumo settings and can offer any insight it is appreciated.
Finally, there is an ISU #145 deadband setting which allows you to change the deadband from 3-8F when heat/cool auto changeover is enabled. Is there a similar setting which would allow me to set a deadband on our heating setpoint (i.e. heating tstat setpoint is 68F, but the system is set to range +/- 2F, or 66-70F)? I thought this might help the short cycling by allowing the system a 4F range so it has more time to find its sweet spot before the set temp is satisfied. I didn't see any setting that appeared would accomplish this.
I'm really hoping the system is not oversized. Our contractor recommended a 3T, but I asked for the 2.5T because it appeared it was better-suited for our manual J (especially in the summer when even the 2.5T is oversized). The contractor does offer a 2-year 'right size guarantee' in which they claim to replace the unit if it is not sized correctly and does not heat/cool the home to the owner's satisfaction. I imagine they wouldn't do anything for me if the unit was undersized since I asked for the 2.5T, but hopefully they will honor the guarantee if it turns out the 2.5T is still too big.
"tstat is located in the hallway on our upper level"
Right sizing and issues aside, this generally doesn't work well in heating climate due to hot air buoyancy. As soon as the air handler turns on, fair bit of the the hot air rises to the 2nd floor which causes the thermostat to shut off.
I would move the tstat to the main floor and see how it runs.
I moved the tstat to our main level earlier this evening. If the new location seems to help, I can move it permanently.
dtkrie,
A non pertinent side track if you don't mind: what is the labeling on the top horizontal axis represent? The 1,2,3,4...?
Does Emporia allow you to export as .csv and these graphs were made in excel?
The top horizontal label is the secondary axis which I had to use to plot the outdoor temperatures. It represents the data entry # from the weather station. Depending on the day, the weather station I used had varying recordings (for instance, on 3/15 the weather station had 28 different data points, but on 3/16 the station recorded 42 data points). I meant to remove the labels when I saved the graphs as PDF's to avoid any confusion, but I forgot.
Yes, Emporia can export the data to .csv files which I then used excel to create the graphs I have posted. I could have just sent screenshots from the Emporia app, but the graphs allowed me to show entire days in a single graph and allowed me to plot the outdoor temps alongside the energy use to paint a more accurate picture.
My contractor has requested the local Mitsubishi factory rep to stop by and test the system, but it will be April 24 before he is back in this area. If anyone has any ideas or recommendations to try prior to the 24th let me know.
Try closing all registers on the second floor...
As Akos and I mentioned - we do not believe your t-stat is at a good location.
Do me also a favor - turn temps to 80 heat, set fan to "low" and show me a screenshot of your Emporia in "Minutes" energy consumption of your outdoor. The screenshot should cover roughly 45 minutes at which point you can shut it off/ return to 68f or whatever your indoor you maintain. Also take a note of the outdoor temp during this exercise.
Thanks
Kicked the system up from the 67F setpoint to 80F at 7:02 AM today. When I checked the system at 7:47 AM, my Kuhmo Cloud app showed an indoor temp of 70F. Outdoor temp was 38F. Attached is a screenshot of the energy use.
Regarding the tstat, it is completely wireless, so I can move it wherever it is needed. The contractor installed it over the old tstat opening in the hallway, probably to cover the existing hole in the wall more than anything else.
Have you done any simple performance checks like measuring the return and supply air temperatures?
One easy test to do is repeat the same test and set the fan on high. Confirm your running near your max of 4kw for a few mins and go outside and measure the temperature of the larger copper pipe on your outdoor condenser. That pipe should be around 125-150+ degrees. Report back what your find with those measurements. A basic meat thermeter will work.
I assume the graph includes also a defrost cycle in which case it looks perfect load- response- wise. So yes- the system operates as expected, but it short cycles day-to-day as thermostat location is not optimal. Also as suspected you are a bit oversized as you raised the temp from 67 to 70 in 45 minutes. That is very aggressive for an ashp...
Lastly, completely anecdotical, but I suspect you are actually a wee bit undercharged as 4kwh for a 30k indoor is a bit too much at 37 outdoor and also the load is going up instead of down over time. Have tech check high pressure - it should be between 375 and 410 psi in heating test mode after 30 minutes for your unit depending on outdoor temp
Thanks, Greenright. I'll check the pressure readings from the initial install startup report when I get home today, and I'll ask my contractor to check again when they come back out.
With the heat pump raising the temp 3F in 45 minutes, does the expected timeframe change at all since we have a hyper heat outdoor unit instead of a traditional? To be clear, I agree it is still oversized; I'm just wondering if the hyper heat would change the expected performance in this regard.
Regarding the original system sizing, when I looked at our Manual J again, I realized it has ~5500 btuH ventilation heating load at 0F outdoor temp. Our home does not have any continuous ventilation system; the only exhaust air is the bathroom fans (no kitchen hood). The air infiltration was set using the results from a blower door test. If anything, the infiltration may be a bit less than shown on the manual J due to a bit of air sealing that occurred post-blower door test. Had I noticed the ventilation load prior to ordering the heat pump, I probably would have requested they step down to the 2T unit.
So turns out the tech didn’t record the refrigerant line pressures when the system was commissioned, at least not on the paperwork I have. I’ll have to request they recheck the pressures when they come back.
I did notice that when you asked me to run the test at 80F set temp I was supposed to set the fan to low, which I failed to do. It was set to auto during that test. If it’s beneficial to reset the fan to low and run the test again, just let me know.
Finally, I took the attached photo of the refrigerant line set right where it enters the indoor FCU. When the system was installed they sent a different tech to commission the system than the one who installed it. I asked the guy who commissioned it if this kink would cause issues, but he said he wasn’t concerned as long as the system tested within specs.
The kink in the lineset is not a problem. it is probably a 90 degree elbow there. As long as there arent many of those it should not be a problem.
It also appears that your unit is precharged for up to 70 feet of lineset (ie distance from outdoor to indoor unit). I assume you are within spec there so charge is alright.
Given the above I would move the t-stat away in an effort to increase run times. If that is not enough I would ask installer to turn-down the blower in the indoor unit - dont know this particular one, but most of those have a 3-speed ranges to match better to duct work. Your system is probably on the "mid" blower setting/ harness- ask to move to "low" if available.
Other than that- downsizing the equipment is the only other option.
Thanks for the reply. Blower fan has always been set to the lowest speed setting. I'll see what it does over the next couple days with the tstat moved to the main level.
Hmmm, that looks like a pretty tight bend radius on the upper pipe. Could be fine but I’d slit the insulation and take a look. The copper should be circular in cross section and not oval or constricted.
There is an ever-so-slight kink in the lineset that you can see in the attached photo. Not sure it is enough to justify replacing the lineset, but I also don't install heat pumps for a living so take my opinion with a grain of salt.
FYI, there are (7) total 90's in the lineset between the indoor and outdoor units, but this is by far the tightest. The rest of the bends are pretty wide and gradual.
Photo above looks fine, no worries. They used a bender which is good practice esp on larger dia tubing.
I have the same setup and some similar concerns. I replaced a gas unit and my energy bill has increased. Here is some of the info from my setup:
1) I understand that there are two kinds of blower settings. There is the "static pressure" setting which defaults to the "middle" setting. And then there is the fan blower setting on your thermostat itself. So when we are talking about the blower setting I hope that you have been able to go into the "installer" settings (for me it is in the thermostat) the unit and set the static pressure to the lowest setting. Also if you do this I would be interested if the temperature of the air in your ductwork increases, as it would on a gas fired unit.
I have yet to test my unit with a clamp meter to see what is really going on with power usage, but I may get a chance this weekend to do so.
2) When our system was first set up in October we had a different default problem. The unit was set to use the intake air temperature (in our attic!) rather than the room temperature where we have our thermostat installed. Once that issue was fixed by the tech the unit performed much better.
3) I have the PAR40 thermostat but I think we would be better of with MHK2. One problem is that the minimum temp on the Par40 is only 63 degrees. Based on where our thermostat is located we would do better with a temp of 62 or 61 to get our best sleeping temperature. I might have to swap the thermostat to the other side of the wall because right now there is air rising from a stairwell that is not yet well insulated. That causes the system to keep running even when it has worked in heating up the other 90% of the house.
4) Our thermostat has an energy saving mode that appears to keep the blower off a significant part of the time. However the documentation on this (which can be set to 80% or 50%, whatever that means) is fairly poor. Maybe it is better documented in the Japanese manual!
5) Our system should work better in spring temperatures then yours because we can turn it off and just run an 18K unit that we have in place, downstairs.
Finally due to the relatively poor HFPS rating of these units (I think mine is a 9.0 where other combinations perform around 12.0) there may be some inherent defect in the units themselves that causes them to underperform. I hope this is not the case but if this issue presents itself, then perhaps we can document that they don't even get the rated HFPS. And if this is the case we can get the outdoor unit swapped out for one that does better in heat mode. If you were to install an inexpensive Mr. Cool 12K unit for your low energy seasons then you could run your main Mitsubishi unit a lot less, and it would likely last longer as well.
Depending on gas and electric rates you rarely get more btu per $ with ashp than with high efficiency condensing gas unit.
So switching from gas to ashp for heating is almost never a winning proposition money- wise.
ASHP wins over resistive heat, propane and oil, but not against natural gas. For now.
Deleted
Deleted
Where I live, as long as the COP of the heat pump is about 3.0-ish, the heat pump is cheaper to operate than an NG furnace.
In my case the electricity rate was almost constant compared to winter 2022, while the natural gas rate went up a lot due to the war in Ukraine. I am concerned about the usage. I used 1072 kwh from the grid and 366 kwh in solar, for the month ending March 19. That is a total of 1438 kilowatt hours. Last year in the same period, I used 570 kwh from the grid and 420 kwh from solar, for a total usage of 990 kilowatt hours. Also I put ~150kwh into new electric appliances last month (car plus induction range). So the increase is from 990 kilowatt hours to about 1288 not including the new appliances.
Maybe that is not so bad after all?
"Based on where our thermostat is located we would do better with a temp of 62 or 61 to get our best sleeping temperature."
My understanding is that big night-time setbacks on heat pumps are generally a bad idea. I've always heard that heat pumps are more of a "set it and forget it" type of system. Maybe some experts will weigh in.
You are right but my setback is only 2 degrees F right now, which doesn't seem to be a big deal for a heat pump. I guess it can be an issue at 1 or 2 in the morning if the unit has to do a defrost cycle to start heating again. Or at 6am when I turn the unit up a couple of degrees. Having it run in the morning is beneficial for keeping the bathroom humidity under control when people shower.
I did move the thermostat temporarily inside the master bedroom a couple of hours ago so it won't be affected by drafts of return air coming up the stairs when the HVAC is on. I have a hunch that this is going be a big improvement. Ideally I would be able to heat my whole house with just one of the heat pumps operating, so Dec-Feb I would run my larger 30K system and November/March/April I would be fine with the 18K system.
My understanding is the ~9 HSPF of the central ducted units is due to the inherent energy penalty when you factor in a blower tasked with distributing the air through a house full of ductwork; the energy consumption of the blower effectively reduces the efficiency of the system as a whole. This logic makes sense to me.
It actually is not the whole story because there is a similar product for mid atlantic states that comes with a blower that does very well. 12.5 HSPF and 18 SEER. The outdoor unit is Mitsubishi-SV-KP30NAH2 or Mitsubishi-SV-KP30NA.TH . So I think the issue is that the hyper heat technology is not particularly efficient at temperatures outdoors of 25 to 35 degrees F, compared to the non hyper heat version. See the specs for these paired with the same blower here:
https://hvacdirect.com/mitsubishi-svz-ka30na2-30-000-btu-18-seer-ductless-mini-split-heat-pump.html
https://www.ecomfort.com/Mitsubishi-SV-KP30NA.TH/p96639.html
Its a pretty huge energy penalty and it is the main reason single zone units are just so much more efficient. It takes a lot of energy to move air through ducts. Let's compare some fan data below.
The fan motor full load amperage (FLA) of your SVZ-KP30NA appears to be 3.3 amps/792 watts.
My single zone gree mini split with a 38 seer/15 hspf rating has the fan FLA listed at .3 amps or 60 watts. According to my Emporia electricity monitor it actually consumes way less than that around 30 watts.
What I would do is install a 12k single zone to use in mild weather and use the Mitsubishi in severe cold/heat when its more appropriate sized.
People should be installing smaller units if they want low modulation. They are just so much more efficient as well. Just off the SEER and HSPF numbers these single zones use like 50-60% less electricity than ducted and most multisplits.
I
For what it's worth, I have a 5-year old single-stage Lennox furnace that I monitor with an Emporia Vue 2. It draws 55 watts in "fan only" mode pushing about 500 cfm. When the furnace kicks on, the fan is pushing about 1300 cfm, and it's drawing about 550 watts. This is a 50,000 btu furnace. So I question whether the SVZ-KP30NA would ever really be drawing 792 watts. I bet it's actual usage is much, much lower than that.
Edit: Another data point. At a different house, I had a 2-ton Trane XV18 variable speed heat pump. That unit drew 20 watts pushing about 300 cfm in "fan only" mode. Unfortunately I no longer remember what amount of power it used when pushing the full 800 cfm.
I think you are onto something though my blower is probably running at only 250 watts. It only runs full blast when there is a huge temperature difference between the thermostat and the set point.
The bigger issue may be that my ductwork was oversized with the old, powerful natural gas blower in mind. One duct run (insulated) goes through my unheated basement, and I'm sure it carries a lot of cold air, especially after the blower has been off for a half hour.
I should be able to remove 50% of basement ductwork, and reduce the diameter in about 8 feet of the runs, resulting in less heat loss to my basement.
Let me offer a suggestion to the OP on your problem. It seems to me that the efficiency problem could be addressed with the features of certain thermostats. I have reviewed the PAR40-MAAU thermostat manual (link: https://www.mitsubishielectric.com.au/product/par-40maa-wall-mounted-controller/ ) and there are "energy saving" settings that could help. It won't be perfect but it would be better than swapping out the unit.
Basically the thermostat can tell your unit to only run at 50, 60, 70, 80, or 90% of max capacity. So you still might have an oversized unit but at least if you set it to run at less than 100% it would be as if you had an 18 or 24K system and the cycling problem will be reduced.
Also the thermostat can be set to change temperature up to 8 times in a day, so you could set up the unit to do a longer run in the morning when people are waking up, and at that time run it at 80 or 90% of capacity instead of the 50% you might be using just to maintain temperature.