@dons - the is quite a lot of info there, but not so much in the way of clear specific questions. It might help if you summarised your key questions, and then moved on to others as necessary.

One thing folks here will definitely want to know is the calculated heat loss at design conditions. If you don't already have that, then you should be able to get it from your installers. If they can't provide it, the red lights start flashing.

@cathoderay thanks for the reply. Just thought i’d try get everything in.

Safe to say the red lights have been flashing. Been involved for the majority of the build bar the plumbing & heating aspect. There definitely wasn’t an on site survey carried out as ASHP was ordered at wind/water tight stage. Engineer told me Freedom did the calculation for selecting the ASHP. I think the SAP report was provided to them for that. I’ve attached a page.

Main questions for me are;

1 - the schematics i’ve attached, will this be an improvement to my current system to make it more efficient?

2 - General question.. how does the Midea ASHP know how to modulate power? Does it look at flow temp out and return? As the heatmiser stats don’t talk to the ASHP.

3 - Has anyone used Homely with a midea pump and would recommend? Or should I consider leaving the controller in Weather Compensation and/or using it as an internal stat as it’s in a habitable room.

Posted by: @dons

↑

Just thought i’d try get everything in.

I appreciate that, and no harm in doing it, but it helps to have something to focus on, as you have now done.

Freedom, who also supplied my Midea heat pump via my installer, do have a rather comprehensive heat loss calculator/rad and heat pump selection tool, an Excel spreadsheet, but it needs more data than would have likely been available to them at the time the specified the heat pump. The other important thing is that for quite some time they guesstimated the Midea unit outputs, and got it a bit wrong, slightly over-rated at lower OATs, where Midea units have notoriously poor outputs compared to the 'badge' rating. For example, my '14kW' unit only managed to output a little over 11kW at a LWT of 55 (I have to have it high for various reasons) when it is around zero degrees outside.

Five to five on a Friday (actually half past six now) is never a good time to be chasing things, but I think as soon as the new week starts chasing up the actual heat loss calculations and system spec is an absolute must. The heat loss is the mother of all things heat pump, and if there isn't one, well, there should have been one. And there will be all sorts of consequences from not having one.

Another Freedom thing is they blow hot and cold about LLHs and plate heat exchangers (I have the latter). No one really knows why they do this, and the general view is they do more harm than good, though I am still dithering about removing mine.

The general way of running a heat pump is to set the room stats very high (I use 26 degrees) and then use the weather compensation curve (WCC) to control the output, leaving the room stat a solely a crude on off switch (eg set to 12 degrees C in summer). The question of exactly how a Midea unit modulates it's power output is a matter of much debate, all the more so as the outside air temp (OAT) sensor is in the heat pump box, and is thereby subject to local effects, yet it is the apparent main controller, via the WCC, of the Set Leaving Water Temp LWT. As the Midea units can't modulate down below a certain level, they appear to use cycling to do that at low outputs, but they also cycle alot, at least mine does, when on the face of it it shouldn't need to, another riddle to be solved. See my many posted charts from my system showing this cycling behaviour.

I don't recall anyone on the forum reporting experience with a Homely with a Midea heat pump. The basic Homely idea appears to come under the general heading of auto-adaption, as in adjust the WCC automatically to take in other factors beyond the OAT. I have a crude version implemented via a python script (small piece of computer code) that boosts the WCC a bit when the actual indoor air temp (IAT) is below the desired room temp (and vice versa when the actual IAT is too high).

As a general rule the first step is to open all the TRVs fully, make sure the system is balanced, ie all rads if present heat up equally using the lock shield valves (not sure what the equivalent process is for UFH as I don't have it) and then tweak the WCC to get it right. This takes a while, and needs the full range of weather conditions to happen during the tweaking period, for obvious reasons, to get it as right as can be. Only then think about adding other features/controls.

I did for a while try running with a fixed LWT (ie WCC off). The energy use went through the roof.   

@cathoderay thank you 👍 

i’m in a position where im going to try make the absolute best of what i can for the pump. I’ll try Freedom next week and see what was supplied to them for spec’ing and hopefully feedback.

I currently run it quite similar. I have all stats set to 24*c max with majority of the rooms not getting there with my curve settings. My only issue is if all rooms hit temp & closed the actuators then the only flow route would be the LLH or dead head against the actuators. So come summer i’m hoping to turn heating off fully or only have on to top up at night.

I may try play about with putting the midea controller to indoor thermostat mode and see what happens to energy consumption & ASHP temps. Homely still sounds attractive to me & is easily installed with my system layout so have asked freedom for a price to supply.

My system is pretty well balanced I think, i set the flow rates up by UFH loop length and then tweaked. Whole house temp comes up pretty similarly apart from when the suns out and its position. Only thing is i currently have 34*c going into the manifolds and roughly 31*c in the return side, which would tell me i need to cut the flow rates back or change the pump settings to a lower flow. How much of a drop should i be expecting to see here?

Im maybe overthinking the consumption but like to think the house is pretty airtight & would of been expecting 60-70kw max in very cold temps. I’m not sure how to measure SCOP to be honest.

Posted by: @dons

↑

I currently run it quite similar. I have all stats set to 24*c max with majority of the rooms not getting there with my curve settings. My only issue is if all rooms hit temp & closed the actuators then the only flow route would be the LLH or dead head against the actuators. So come summer i’m hoping to turn heating off fully or only have on to top up at night.

Ideally, you want the room stat set high enough to ensure the heat pump is always on, and all the other emitter controls fully open, to ensure the system is always running as an unobstructed circuit, and then only use the WCC to control output. Come the summer, you turn down the room stat enough to ensure the heat pump is always off. In practice, the room stat becomes just an on/off switch. That's the basic setup, to which other controls (Homely, setbacks, auto-adaption etc) can be added, if wished.

Sooner or later someone will tell you you need to remove the LLH. These (along with plate heat exchangers) are a 'Freedom thing', and no one really understands what they are for. The general view seems to be get rid of them.  

Posted by: @dons

↑

I’m not sure how to measure SCOP to be honest.

There are painfully complicated ways of working out SCOPs to compare different brands/models etc but I have always taken it that you get a good enough estimate for your system as it is now if you simply divide the total energy out for a whole heating season by the total energy in for that heating season. It does lump the DHW heating as well as the space heating, but on one level that is fine, insofar as you want to know your overall efficiency, and it will of course vary depending whether it is a warm or cold heating season, but for me it is 'good enough', even if it isn't for others. In practice, I am usually more interested in shorter time frames, eg how much of a hit do I get in cold weather (answer: a lot), is my system broadly as efficient now as it was when running in similar conditions last year (answer: seems to be, except for when I used a fixed LWT, when the COP fell) etc. If you do install a Homely, you might, for example, want to see how it affects your COP, if at all. 

I will be interested to hear how you get on with Freedom. My experience with them has shall we say been rather patchy. 

Ideally, you want the room stat set high enough to ensure the heat pump is always on, and all the other emitter controls fully open, to ensure the system is always running as an unobstructed circuit, and then only use the WCC to control output. Come the summer, you turn down the room stat enough to ensure the heat pump is always off. In practice, the room stat becomes just an on/off switch. That's the basic setup, to which other controls (Homely, setbacks, auto-adaption etc) can be added, if wished.

Sooner or later someone will tell you you need to remove the LLH. These (along with plate heat exchangers) are a 'Freedom thing', and no one really understands what they are for. The general view seems to be get rid of them.  

Because the stats are not linked to the ASHP system at all really, how will the heat pump know to switch itself off? Currently if all rooms get up to temperature & close all the actuators, the only outlet is for the water is via the LLH, the circulation pump downstream of the LLH will then either dead head against the actuators or cavitate the circulation pump. Does the heat pump measure the temperature returning to it (sees that it is what it is trying to produce) and tell itself to switch off?

Plan was to remove the actuator heads from the manifold for the bathrooms so there is always a flow path (and keeps the bathrooms warm for drying up the floors etc.)

There are painfully complicated ways of working out SCOPs to compare different brands/models etc but I have always taken it that you get a good enough estimate for your system as it is now if you simply divide the total energy out for a whole heating season by the total energy in for that heating season

Is there a simple way to measure this? I'm not sure how to measure the heat produced. I have a spare CT clamp kicking about that I could attach to my Zappi charger and set to monitoring only which would give me a very accurate measure of the energy the heat pump is using.

Posted by: @dons

↑

Because the stats are not linked to the ASHP system at all really, how will the heat pump know to switch itself off?

In an ideal world, it wouldn't switch itself off. It runs continuously, adjusting the LWT, based on the WCC, to deliver the right amount of heat to keep the house in balance at the desired IAT, whatever the OAT. Note Mideas appear to run with a reasonably constant flow rate (at least mine does) so most of the control goes into changing the LWT. That's the theory, anyway. It does need quite a bit of tweaking of the weather curve - trial and error - to get it right. Any emitter control needs to be set to always open (however you do that with UFH). The other thing is there are times when the unit can't lower output enough, and it resorts to cycling the LWT (rises and falls), but note the fluid still circulates all the time. How the Midea unit translates the feedback it gets from the RWT into the adjustments it makes is still a bit of a mystery... though some forum members have advanced various theories. 

Posted by: @dons

↑

Is there a simple way to measure this? I'm not sure how to measure the heat produced. I have a spare CT clamp kicking about that I could attach to my Zappi charger and set to monitoring only which would give me a very accurate measure of the energy the heat pump is using.

The simplest way to do this is via the SmartHome App (also available via the wired controller), which will give you lifetime energy in and energy out in kWh. Simple subtraction then gives you the energy in and out for any period you wan't. This isn't without its problems though. The numbers are whole numbers (rounding errors for short periods) and when it gets to 999 kWh the next value to appear on the wired controller is 1kkWh (massive rounding error)! I'm not sure what happens on the App, I stopped using it ages ago. Depending on how deeply you want to get stuck in, there are other ways of getting data out of a Midea system. But your first priority should be to get the WCC just so.

The CT clamp will be very useful as an independent way of verifying the Midea data, even if only usable on the energy in side of things. 

The simplest way to do this is via the SmartHome App (also available via the wired controller), which will give you lifetime energy in and energy out in kWh.

 

I can see energy consumed on the controller but can’t see energy out. Likewise I can only see power consumed on app and not produced. I’ve attached a couple screenshots. Are you able to point me in the right direction or am I going mad!

@dons - no, you are not going mad, it is just Midea playing language games! On the wired controller screen, you have POWER CONSUM, which is the total lifetime energy (kWh) (not power, kW) in, which you can use by subtraction to get periodic energy in eg current reading minus same time yesterday's reading = last 24 hours use. Above it you have HEAT PUMP CAPACITY which is the current (real time) power (kW) output. If it stayed at 6.28kW for an hour, thatis 6.28 kWh or energy out for that hour. A major problem is that without adding something, you can only read these numbers manually, and have to record them with pencil and paper. The OPERATIONAL PARAMETER pages also have useful entries eg WATER FLOW (flow rate in m^3/h), COMP CURRENT (compressor, not whole unit, amps in) and LWT/RWT which appear twice under the baffling names TW_O PLATE W-OUTLET TEMP and TW_I PLATE W-INLET TEMP.  

The SmartHome app is not much better, and again you are stuck with the data being in the app, or Midea cloud, with no way of downloading it. You want to use bar charts rather than pie charts, and find the History section which as I recall is accessed via the three dots at top right. Or you may already be in there, as you have time interval tabs at the top. You get get data on things that don't exist (TBH - tank backup heater?). You do get coloured charts which the wired controller can't do, but as I recall, it doesn't display your COP, you have to calculate that manually.

There are at least two ways of getting more useful data. The first is to snoop on the Midea app data, using a python module called midea_ac_lan. Because the Midea code behind the app is complicated and riddled with pointless degrees of security (we're not talking nuclear missiles here), the developers of midea_ac_lan only managed to get so much data, and then gave up. The module also needs somewhere to dump store and display the data, which usually means Home Assistant. I tried Home Assistant, and personally (others will disagree, strongly) I found it was too complicated and too much effort for what it could deliver. It was also ham-strung by the fact the midea_ac_lan only has limited data in the first place. I do however still have Home Assistant installed, because it does have one useful feature: it provides the easiest way of getting to the command prompt of a mini-pc that collects the Midea data over a modbus wired connection to the Midea wired controller (and a separate temp sensor for IAT). 

This alternative method, which I currently use, is a bit of a faff to setup, but once done it is reliable and useful. Using python's plotly module, it can produce useful charts like this (this is last 24 hours data, but it can display any period you want):  

The saw tooth pattern on the right hand side is cycling, which happens a lot on my Midea unit. Heat pump lore says regular cycling is bad news, but as it is usually relatively low frequency, once an hour or so, I just live with it. The only time it doesn't cycle is when the IAT is around 5 or 6 degrees, as can be seen on the left hand side of the chart. These charts also show defrost cycles when they happen (LWT drops below RWT). Here is my heat pump not quite coping a few days ago. Look at the energy in and COP:

I have partly documented how to do this in another thread, but ran out of steam as there was little apparent interest, and doing the write up was time consuming. The thread is here, and gets to the point where you have everything you need, but not the actual physical setup or code. If you are interested, I just might be persuaded to write and post the remaining sections... 

The other useful thing I have done is to write a 'auto-adaption' script (in python). It is a much simplified version of what Homely does, and Mitsubishi's Auto Adaption 'feature', in that it adjusts the WCC endpoints, and thereby the LWT, according to any difference between the set (desired) IAT and actual IAT. Basically, if the actual room temp is a bit less than it should be, then up the WCC a bit, if the actual IAT is a bit higher than it should be, then lower the WCC a bit. if the IAT is where it should be, then set the WCC to normal. This can be done over modbus, as you can write (set) values as well as read (get) values, and it does seem to improve reheat times, when otherwise the heat pump would struggle to recover from, for example, a setback.   

On the wired controller screen, you have POWER CONSUM, which is the total lifetime energy (kWh) (not power, kW) in, which you can use by subtraction to get periodic energy in eg current reading minus same time yesterday's reading = last 24 hours use. Above it you have HEAT PUMP CAPACITY which is the current (real time) power (kW) output. If it stayed at 6.28kW for an hour, that is 6.28 kWh or energy out for that hour.

Begs the question with the controller able to display so much data it can't give a 'home owner display' with simple nice to know information. I checked my controller every 10mins for an hour and worked out a very rough COP for that hour of 3.55 @ 4*C ambient... I take it that sounds about right, but still doesn't please the bank balance & usage. It's a waiting game for my Octopus bill (swapped over end of November & done a lot of tweaking/very low ambient temps) & using quite a lot during the cheap 7.5p tariff at night (EV, Battery Storage, washing etc.) Right now I'm pointing the finger at the heat pump for high usage but will be interesting to see what the actual bill is for the whole of December, which it really boils down to.

The SmartHome app is not much better

Yeah it's pretty awful bar remotely scheduling DHW for me personally. Don't trust the energy consumption.

I'm still in the process of setting up my home rack & yet to set up a Pi/HA, but will look into it further once I have the spare time. I'll take a look at the thread (Y)

It is a much simplified version of what Homely does, and Mitsubishi's Auto Adaption 'feature', in that it adjusts the WCC endpoints, and thereby the LWT, according to any difference between the set (desired) IAT and actual IAT. Basically, if the actual room temp is a bit less than it should be, then up the WCC a bit, if the actual IAT is a bit higher than it should be, then lower the WCC a bit.

 

What are you using to measure IAT? the midea controller or external? 

Some excellent information you've provided.

Posted by: @dons

↑

I checked my controller every 10mins for an hour and worked out a very rough COP for that hour of 3.55 @ 4*C ambient

That sounds in the right range, for a very rough COP at that ambient, perhaps a bit high, see 1200 to 1500 in the second chart I posted above. When the ambient goes a bit lower, the COP plummets, as again you can see in that chart.

Posted by: @dons

↑

I'm still in the process of setting up my home rack & yet to set up a Pi/HA, but will look into it further once I have the spare time

It sounds as though you are already familiar with this sort of stuff, you will I am sure get on just fine.

Posted by: @dons

↑

What are you using to measure IAT? the midea controller or external? 

Some excellent information you've provided.

The IAT comes from a separate sensor (XY-MD02) that is also modbus enabled and will work with 5V ie USB supplied voltages.

Thanks!