Mits Air Hyper Heat 3.0 ton heat pump – excessive defrost cycle and water discharge
We recently had a Mits Air Hyper Heat 3.0 ton Side Discharge heat pump installed. We have a Sensi smart thermostat.
The unit goes into defrost mode 10-13 times a day, and is discharging 25+ litres of water a day. We have to gather the discharge water in buckets to prevent it from seeping down into our foundation. The unit is noisy. If we adjust the thermostat 1 degree our auxiliary heat comes on and continues to come on until we manually set the system back to heat pump.
We live in eastern Ontario and these issues are happening with temperature ranges from +5 to -10. We haven’t even experienced cold temps yet.
Our HVAC contractor states that the refrigerant levels are correct and that all measurements they have taken indicate that the unit is operating properly. He feels that the humidity in our geographic area is triggering the unit to operate in this fashion.
Any insight would be much appreciated.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
This near freezing weather is when heat pumps will produce the most amount of condensation. A defrost cycle per hour is normal.
25L does sound a bit on the high side though. If there are restrictions around the outdoor unit or if there is not enough airflow around it which can cause recirculation, which will significantly increase the amount of frost buildup.
Some units also have the option to configure defrost operation, I would check the installation manual for your unit to see if there is something you can adjust.
Post a picture of your outdoor unit to see if airflow is the issue.
Thank you. We will look at the installation guide. Here is a pic of the unit. I also posted a broader pic below.
I think it is a mistake to connect any third-party thermostat to any variable compressor system.
The third-party connection only has commands of OFF, LOW and HIGH. You paid for the 5000 or so other speeds that would perfectly match your homes heat loss keeping your home at exactly at it set point. As apposed to the third part that must over shoot the set point turn off until it falls below the set point only to over shoot again. Yes, the swing is small less than 2° but it is real and noticeable.
My factory thermostat logs every defrost cycle with 13163 hours of run time and 3175 defrost cycles or 4 hours of run time per cycle. The outdoor humidity and temp are huge factors in how often a defrost is required.
The other thing to consider is that excessive defrost cycles are a symptom for an under charged unit so you may have a small leak in the system. What is the temperature difference between the supply and return air? If less than 20° that would be another symptom.
Walta
In this humid weather around freezing your outdoor will defrost often. Very often. This is normal. Yours will defrost even more often because judging by the pic it is in somewhat enclosed space.
And an on/ off thermostat works just fine with a modulating mini splits because the modulation decisions are made by the indoor unit based on the refrigerant temp drop across the coil (in heating)… and the thermostat plays no role in that.
The modulation decision is made based on how well the unit is maintaining the setpoint. A temperature sensor gives this information, the on/off of a thermostat does not.
Some units use a relatively smart controls and you can get pretty good modulation using a regular thermostat.
The nice way is to slowly ramp up power for each heat call and than hold or slowly ramp down the output after the heat call. This is not as smooth as continuous modulation but is good enough. Kind of like using the thermostat output as a PWM signal and setting the power level based on the mean value.
Don't know if that is the case here, but sure hope there are some smarts.
The key thing is that there are two levels of modulation -- the fan speed in the indoor head and the compressor speed outdoors. When you have an indoor sensor you can modulate the indoor fan speed to keep the temperature in a very narrow band. Then you modulate the compressor to match the output of the indoor unit.
With an on-off thermostat, or a high-low-off thermostat, the most you can do is try to lengthen the cycles. When the thermostat is on all you know is the temperature is below the setpoint, you don't know by how much. And when it's off you know it's above the setpoint, you don't know by how much. You can look at the length of the on and off cycles and try to deduce the actual load, but you can't actually match it because then you lose your source of information, the cycling of the thermostat, and are unable to respond to changes. So you still have to overheat by a bit for a while, and then cycle off. All you can do is make the cycles longer and less frequent. This takes away a big benefit of the variable speed technology, the ability to run for long times at close to constant outputs.
… yes and no… If you use a dumb thermostat you will lose the variable indoor fan speed (but there are usually only 3 speeds anyway), but modulation is largely unaffected. The compressor step is determined by the pressure drop across the indoor coil (in heating all these do is target specific high pressure) - the compressor speed drops as the indoor temp raises and the coil drop becomes less thus allowing the compressor to slow down and still
maintain target high pressure. This as the room temp raises rhe compressor slows down.
On indoor fan speed- lower speeds are strictly for comfort. For efficiency you want to run the indoor fan at the highest tolerable for comfort speed anyway.
So the biggest advantage of modulation- matching the load (drop across the indoor coil) to the compressor step is still functional with a dumb thermostat. The only advantage of a communicating thermostat is that the indoor unit knows what the set point is and can control the fan- but that has very little if anything to do with efficiency as anything other than max fan speed is less efficient anyway
And I agree the location is actually perfect- sheltered from the elements but with enough airflow.
It is somewhat possible for a manufacture to do some of the stuff you are describing. My question is do you have any evidence that any manufacture is putting that much thought and effort into supporting thermostats they did not sell? sales literature, theory of operation, patent filing or just wishful thinking?
“So, the biggest advantage of modulation- matching the load (drop across the indoor coil) to the compressor step is still functional with a dumb thermostat.”
The outdoor unit has received a command for low-speed heat. It then runs the compressor at what ever RPM it is programed for and holds that RPM until the command changes. The outdoor unit will adjust the expansion valve to get the best superheat or load drop for the current set of conditions. The outdoor unit has no data to understand how close the set point it is + or- the only data it knows I have a command at this moment.
“The only advantage of a communicating thermostat is that the indoor unit knows what the set point is and can control the fan- but that has very little if anything to do with efficiency as anything other than max fan speed is less efficient anyway”
Many miny owners report discharge temps about 120°. The indoor fan speed is about clearly set to maximize comfort over efficiency.
My Rheem communicating thermostat gives me tons of data about what my systems is doing and history about what it has done.
I think not buying the manufactures communicating thermostat is silly.
Walta
"The compressor step is determined by the pressure drop across the indoor coil (in heating all these do is target specific high pressure) - the compressor speed drops as the indoor temp raises and the coil drop becomes less thus allowing the compressor to slow down and still maintain target high pressure. This as the room temp raises rhe compressor slows down."
But the whole point of a thermostat is it keeps the room temperature constant, so you can't use fluctuation in the room temperature to modulate the output of the compressor.
"The only advantage of a communicating thermostat is that the indoor unit knows what the set point is and can control the fan- but that has very little if anything to do with efficiency as anything other than max fan speed is less efficient anyway."
The spec sheet for one of the Mitsubishi M-Series is here: https://ashp.neep.org/#!/product/34583/7/25000/95/7500/0///0
At 47F, at minimum output (14.7 kBTU/hr) the COP is 4.4. At max output (34kBTU/hr) the COP is 3.29. You get significantly higher efficiency being able to run continuously at lower output than cycling on and off at max output.
"But the whole point of a thermostat is it keeps the room temperature constant, so you can't use fluctuation in the room temperature to modulate the output of the compressor."
Yes - but remember one very good representation of the room temp fluctuation is the pressure drop across the indoor coil - and the compressor responds to that (the system aims roughly at 420f high side)... so as room temp goes up the compressor step goes down as it takes less compression to reach the target high pressure which as mentioned is fairly static.
"At 47F, at minimum output (14.7 kBTU/hr) the COP is 4.4. At max output (34kBTU/hr) the COP is 3.29. You get significantly higher efficiency being able to run continuously at lower output than cycling on and off at max output."
Correct. With a communicating thermostat the system will run longer if in fan auto mode. Comfort will be higher too as it will run lower fan speeds when approaching setpoint. This should be more efficient, right? Higher COP after all means just that. However, keep in mind that the only time a system is efficient is when it is off. So if you actually have it blasting at high fan speed and compressor stepping up to keep the desired drop across the indoor coil -you-will-actually-use-less-energy-to-REACH-setpoint. It has to do with two things - one - higher air speed across the coil = better heat transfer. Two- the outdoor will run less time and most of the defrost cycle algos are based (mostly) on time once temp drops below 40f outside or so.
Also once a communicating thermostat is close to setpoint it will slow down the fan (that is all it does - the thermostat does not control the compressor step- the drop across the indoor coil (which is affected greatly by the indoor fan speeed) determines the compressor step) and the btu output.. and will run.... and run... and run.. ideally all the time if it manages to match the btu output to the heat loss of the heated space... the indoor temp will be rock solid (or close to setpoint), but it will run.. and run.. and run some more... and that uses energy... more energy than if the unit just blasts until it hits setpoint and... shuts off...
So hard and fast - better heat exchange and less defrosts (but less comfort obviously and more indoor temp swings). which results in lower electric bill. This is a well known truth especially in the commercial world (this is what I do) - large warehouses where comfort is irrelevant dont even call for modulating equipment- history has shown lowest bills are achieved with simple on/off high btu units. Dont believe me? If you have an energy monitor on your ASHP corelate your consumption to heating degree days for several days on "auto" and then turn fan on "High" and do the same. I guarantee you that you will see a minimum of 10-15% less consumption with fan on "High" and that is accounting for the higher energy use of the indoor fan. Hope this helps
The unit I linked to has a minimum output of 14,700 BTU/hr at COP 4.4 and maximum of 34,000 BTU/hr at COP 3.29.
If what I need is 14,700 BTU/hr, I could either run at minimum output all the time, or I could run at full output 43% of the time, or 26 minutes per hour. If I do the former I get a COP of 4.4, the latter 3.29. The fact that it's off for 34 minutes per hour does nothing to improve efficiency.
This is perhaps a better pic to explain the location and airflow potential. The unit is under a small deck, wide open on two sides
That looks like a pretty good location for it.
I would try to adjust land slope to channel the water away from your basement. If water off your AC unit is making its way into the house so will rain water, you want to fix this and then the water of the AC unit becomes a non-issue.
DC - I dont seem to be able to reply to your latest post (I guess it is nested deep enough) so I will reply here.
" The fact that it's off for 34 minutes per hour does nothing to improve efficiency."
It sure does- because if you run X minutes more you will hit more defrost cycles (as mentioned the defrost algo is driven mostly by run time) and over the course of the day you will hit 33-50% more defrosts and all your "gains" will be more than negated. So that alone makes the theory not like the real world. Another question- when you will get more snow/sleet/freezing rain on the outdoor coil and pack it causing a non-time-based defrost? when you run for 30 mins or for an hour? (hint- most units have same outdoor fan rpms at any temp below 40 (when direct vapor injection starts on the cold weather units).
But lets say there are no more defrost cycles for the sake of argument when running low and long. You are missing the point that "efficiency" has nothing to do with ability to heat. It has to do with energy used per btu (over time). In a perfect world your communicating thermostat will adjust the output to match the heat loss and will run non-stop. Best scenario for high efficiency- correct? Yes. Best scenario for money spent per day? Not really. Because you lose heat over time. Pop quiz- when will you lose more heat- over the course of 30 mins or the course of an hour? The hard and fast on/off system will satisfy the setpoint in 30 mins and then rest until the indoor temp drops 1-2 below set point and turn on again. This is where it wins- not running during that temp drop of indoors. The low and long system will never shut off because it will just replenish the heat loss but never be satisfied (or be satisfied much longer). So one runs-stops, run-stops and the other runs-runs-runs... Now add the fact that higher flow across the coil = higher pressure drop and thus less pumping losses and the run-stop system usually comes ahead moneywise even if its COP is lower at any point in time.
Again- you dont have to believe me- do the experiment I asked you and let us know how it works. I for one can reproduce the results I've mentioned at will... at home and at the several dozen commercial projects I have exposure to.
Lastly- pros know too well that modulation is mostly done for comfort. I know this is against the residential "grain", but just the facts.
I am done here. Thank you
I refer you to the article, "The Hunt for the Most Efficient Heat Pump in the World." https://www.wired.com/story/heat-pump-maximum-efficiency/
It's a modulating air-source heat pump. In the real world people are achieving seasonal COP's of over 5.
Modulation is not about comfort -- although HVAC is fundamentally about comfort, and maintaining a setpoint is fundamental to comfort. Modulation is about efficiency, because a heat pump is most efficient when it runs continuously at the lowest output needed to meet demand. In order to modulate it's not enough just to modulate the speed of the compressor, you also have to modulate the output of the air handler. The way to do that is by modulating the fan speed. I guess in theory you could modulate the output by modulating the temperature, but heat pump efficiency is inversely proportional to the temperature delta, so increasing the temperature hits efficiency and gives you nothing that increasing the fan speed wouldn't give you without the efficiency hit.
>I am done here. Thank you
Bye.