Minimal Modulation with Ducted Heat Pump
Mitsubishi ducted heatpump; minimum modulation of 40,000 BTU
I read all the great articles and question/comments here on ducted mini and multi-splits and the importance of minimum modulation to reduce cycling and to enable efficient run energy consumption.
Even after reading all the articles, for my situation a ducted unit looked best and so about a year ago I went with a Mitsubishi (MXZ-SM48NAMHZ with a PVFY-P54NAMU) rated at 54,000 BTU and compressor specs has a min of 27,000 btu at 47f (and 11K at 17f). I started measuring output and surprised to found out it does not modulate below about 40,000 btu at 50f outside and the lowest modulation that I have seen when colder is still over 36,000 btu. Based on the daily BTU output the unit seems to be correctly sized for load at design temp (5-10f).
Figured this out by logging supply temperature in supply duct close to indoor unit and tracking fan setting. I haven’t seen the supply air temp modulate under 100f (or about 32 degree delta T) = Using cfm*1.08*delta T = 32*1.08*1040 = 36,000 btu at (lowest fan speed). Minimum supply duct temp at 50f outside was 103f before turning off.
I posted about a month ago regarding high energy usage and this could explain a good bit of that I guess. Incredible how little modulation I see with this ducted heatpump.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
My only question is how do you know the CFM? Do you know the pressure curve?
Thanks for response and great question as it is the point of uncertainty in my number as I don't have a system CFM measurement. I have static pressure readings and the fan characteristics curves from Mitsubishi and also power monitoring which supports lack of modulation. Strangely enough I was previously told by a tech that I had high energy use due to too much airflow and ducting. I have a hot wire anemometer but haven't figured out a clear place to traverse.
My take on these modulation issues is that the heat load calc itself is the problem. When a larger unit is spec'd for an incorrect (as in, too large) heat load vs actual, then it uses more power cycling vs running at lower demand. How was the heat load calculated in the first place?
The local Mitsubishi reps and distributor are more concerned the installed unit is undersized.
I calculated the heat load first using historical oil use data, and did follow these methods generally https://www.greenbuildingadvisor.com/article/planning-a-furnace-to-ducted-heat-pump-retrofit-part-3-sizing-calculations
Post install I have also calculated using output and if anything likely a touch undersized.
I knew the turndown was poor on this multi-split type unit, even through installed 1:1 with a single indoor blower but even so what I see is much worse than expected.
You're not alone with this issue. We have a Daikin Fit 2 ton, VS, ASHP with 2 ton VS air handler installed in Spring '21. When running in standard heating mode (aux. locked out), it will slowly ramp up to full compressor speed and 800 CFM indoors, then run at full power until it overshoots the set point, then it will quickly wind down and shut off. The only way to overcome this is to put the unit in the lowest level of 'Quiet Mode.' This results in app. 40% turndown ratio, or 14k Btu/hr. of the 22.8k Btus/hr. rating (47F). I know this because the OnePlus stat displays CFM, air handler liquid and suction temp., etc. The only information I lack is power and energy monitoring, although Daikin occasionally sends monthly energy reports.
The advantage of this turndown is the HP fan coil now blows at 576 RPM, which is quieter. It should also use less electricity, assuming the relationship between turndown and power/energy isn't fully linear. The unit is also operating at a lower pressures. One disadvantage is that register temps. are lower (83F- 89F), versus 93F- 99F in standard heating mode. I don't mind this, especially at the lower fan speed.
I'm not sure how they can truthfully call these units 'variable speed,' at least in heating mode. Cooling mode does display lower modulation and longer run times.
I'm also disappointed with the NEEP database. For our unit, it clearly shows 65%-75% turndown ratios in the data and graphs. If the unit cannot ever do this in standard heating mode, let alone in 'Quiet Mode,' then where did they get this data?
Daikin's expanded data on our unit only shows maximum outputs, like it's a single speed compressor. The words 'variable-speed' are also used in the document to describe the compressor, and in multiple sale's brochures for the heat pump.
I was actually surprised to see that was a valid combination. Usually most multi splits have a 2 unit min.
I would look through the installation and engineering manuals here for your unit:
https://www.mylinkdrive.com/usa/M_Series/R410A_Systems-3/R410A_Outdoor
The Mitsubishi units usually have an installer menu where you can set options, on some multi splits I know evaporator target temperature at least in cooling is adjustable. There might be something similar for your unit.
I also find it odd that they didn't go with the 4 ton indoor unit, this is probably not helping things.
That ship has sailed, but I think the 3.5 ton P series hyper heat (maybe with backup coil if too small) would have been cheaper and without some of these issues.
Thanks for the suggestions and I have read all of those install and engineering manuals to no avail. Actually tried Mitsubishi phone support which was helpful and they told me that the unit is not performing correctly but I cannot get service on it. Local diamond service group reps have everything is fine so kinda stuck.
The specs overall all almost identical to the P series version, PVA-A42AA7 & PUZ-HA42NKA1. The blower actually has the same core part numbers and ratings. The unit I got has slightly better efficiency numbers and more importantly was available at the time and qualified for rebates that the P series one did not.
This might be the hidden reason: these are multizone outdoor units so I think might be software locked to prevent modulating lower. Even though it's connected to a single indoor ducted unit, the installation instructions seemingly require a branch box, which would add another layer of potentially missing communication between indoor and outdoor units. My guess is the 54k indoor unit might also be affecting the ability to modulate lower.
It is interesting the AHRI submittal and tests show a low modulation around 10k, but there are not enough details about the multi-zone configuration tested. Additionally, the 10k number might unfortunately be based on theoretical capabilities of indoor units and not an actual real-world test result.
I think an article on how to assess the modulation performance of these units would be really cool. I've never seen one (but point me if I've missed it :)
Something that explains how the manufacturer numbers and the NEEP numbers were calculated and then best practices for gaining insight in the field (from energy meters to what is described in the OP, supply temp, CFM etc).
Then a final bit about why IN THE HELL ;) these units don't provide their own audit logs that are easily readable, with a final final piece suggesting that a regulatory body (or at least a building certification program) require these data to be available. Sure, the manufacturers will say "there are lots of installation issues that we aren't responsible for" and the installers would hate it too. But both end-user comfort and public policy needs this data.
Completely agree, these units have sensors throughout that could easily be used to determine real time performance specs and COP and the sensors could also be used to flag anomalous behavior or even just verify that everything is working to spec.
I did receive a quote for one heat pump that does this with real time over the air performance monitoring https://www.ecoer.com/ Quote I got was high and capacity a little lower than the Mitsubishi and what I needed at low temps but that monitoring information would have likely saved and incredible number of hours for me / installer / support line / local reps... etc
Interesting! I checked their website and it doesn't have many details. I called and talked to tech support and they were chatty. Sounds like they use combination of sensors (including supply temps, amp/voltage, and maybe refrigerant levels) to create an estimate of current output. I asked if they had any demo of their system and interface and they took an email and said they'd check their training material. So we'll see what happens.
There website is very much a "call your dealer" brochureware thing, which is a missed opportunity.
Seems kinda unique, but hell yes I'd like to know if the system is leaking higfh global warming potential refrigerant (and therefore get a heads up before it fails). Seems like a no brainer (and again, something that should be required by regulation).
I wish I could add the monitoring to my Mitsubishi unit. If I had a do-over opportunity I would switch based on my experience.
Also onboard with the regulation requirements for the refrigerant leaking and the efficiency verification needs. The is especially true because there seem to be very few people who fully understand how these work - or at least I cannot find them to help.
There are some dip switches in the indoor unit to set min fan speed (low vs very low) and capacity.
Not sure what the capacity dip switch will do but might be worth trying to play with it.
There is also one labeled "Heating 4-deg up" without any description of what it does.
Reducing the static pressure setting will also reduce effective airflow which should improve the min capacity.
Most other Mitsu units have an installer menu on the remote, not sure if this has. There might be further settings there.
The outdoor unit also has a bunch of dip switches, quick look there didn't show any major adjustments that would help with heating. If the unit is oversized for the load, you can look at changing the derate settings which should help. If you have summer time humidity control issues, decreasing the cooling evaporator target works.
@Akos much appreciated and interesting suggestions, will double check those switches. I also could not find any information on what the "heating 4-deg up" option was. I think the capacity dip switch is the size of the indoor blower - maybe increases target sub cooling as capacity goes up but having the largest size likely no help after verifying correct.
From the fan characteristic curves there only looks to be a 10% CFM difference across the three static pressures settings but agree that should move modulation down some and maybe the impact is more than what is shown on the fan curves. Interestingly, the rated CFM is the peak value on the curves or close to it.
I also didn't see anything in the outdoor dip switch options that would help nor anything in the remote installer menu. Thanks!
To my ear it sounds like a control problem not an oversized problem.
What are you using for a thermostat?
Do all 3 computers have the most current software?
Is it a coincidence that we often see posts complaining about Kumo cloud with poor modulation?
Walta
I have the Mitsubishi MHK2 thermostat and have been keeping the setpoint unchanged and no Kumo cloud so that is not it. I also don't think its an oversized problem and setup was verified by local reps.
I am monitoring my heat pumps through a current transmitter with emporia vue. It is obvious by amp draw that it is varying the load. Seems like a more reliable solution compared to temps and air flows.
I have house power draw with 5min resolution and do see high energy usage indicating the lack of modulation that correlates with what I see on the temps and airflow. Lowest draw to the unit I see is over 3000W when in 40s out and about 2-2500 W when in the 50s.
IMHO, Akos is on the right track about reducing the fan speed.
Run as low as possible fan speed, don’t touch the +4 degree setting and let it do its thing. A 4 ton outdoor at 2kw is about as low as it will go. At full tilt that unit uses about 6kw so this is a 1/3 max turn down- about spot on for those units… also keep in mind they use a self- learning algorithm so in time they will learn your setup/ demand and will fine tune themselves. Turning the outdoor off resets the learned blocks.
@greenright - can you please expand upon this self learning algorithm on the Mitsu units?
Is this in the outdoor unit brains? So if there is a power outage the thing resets the learned blocks? What is it learning?
Also, you posted about a 4 ton outdoor unit - can you share the min & max on a 3 ton outdoor unit (I presume we are talking about the Mitsu hyper heat multizone outdoor units)
Thanks