@jamespa Regretting I attempted this now. A very small amount of dirt in the filter, but L9 error now and won't start.

Posted by: @davidalgarve

↑

@jamespa Regretting I attempted this now. A very small amount of dirt in the filter, but L9 error now and won't start.

L9 = Low Flow error

Posted by: @davidalgarve

↑

@jamespa Regretting I attempted this now. A very small amount of dirt in the filter, but L9 error now and won't start.

Oh dear, sorry to hear that, isn't l9 low flow?

@f1p Yes understand that but why. Investigating air in system

Back working again now and problem was air in system. Flow rate is 17 l/min

Answer to question re flow rate put to Mitsu UK (Portugal is useless):

For a 12 kw unit the target flow rate is 32 lpm with a minimum of 14. Which would put this in the operating range that we set out.

 The installer will usually lower the flow rate form the target until the system is balanced and pushing heat into the property.

Now where, in the menu, would they have that ability so that I can check?

Tried "Heat Pump Settings" and that is on 5 min 100 max so that is not limiting it.

It has gone up to 18 l/Min so perhaps the system is slowly winding it up (hope!)

With the input from @f1p and @jamespa I am persuaded that my problems are all related to low available flow rate, particularly as I have a record of a higher figure. However, I am uncertain why this is the case, e.g. is the pump being regulated down or are there other factors?

I had come to the conclusion that my problems started when I added DHW, which necessitated the addition of a three way valve and some additional pipe fittings that would have increased pipe friction. At that time I still had a second pump in the circulation to the radiators, so the available flow would be that from the two pumps in series and the flow would still have been sufficient for the radiator supply.

Before the onset of winter, the buffer tank was converted to volumiser and the second pump was removed. I had checked that the characteristics of the primary pump were similar to the secondary pump, but I ignored the extra capacity from two pumps in series. I also failed to take into account the extra pipe friction, particularly in the line to the radiators. 

This led me to think that the reduction in flow rate was due to increased  pipe friction and I hoped to confirm this by checking the flow rate to the DHW tank,  which is much shorter and almost certainly has less resistance.

I was very surprised to find that the flow rate to the DHW tank was almost the same as that to the radiators, i.e. Setting 3 > 11 l/min and Setting 5>15 l/min

So is the pump defective or is the flow rate being regulated down?

It would be useful to compare flow (l/min)/ power (kW), against the published performance of this Grundfos pump and I wonder if this data is collected by the MelPump app?

Posted by: @davidalgarve

↑

So is the pump defective or is the flow rate being regulated down?

 

You can possibly check the latter if it has a PWM connector which is separate from the pump power supply.  Just pull the PWM cable and most PWM pumps default to max speed.

When you had a buffer you didn't exactly have two pumps in series (thats the point of the buffer), but there was more pump power.  Its possible that could have made a difference.  Opinions seem to differ on whether two pumps literally in series (which is what you would have if you reinstated the second pump but not the buffer) is acceptable ; personally, so long as they aren't widely mismatched, I cant see why not and neither have I heard a rational explanation.  Assuming you still have the original pump then simply replacing it into the circuit may be an option.  

@jamespa Thanks for your comment James. Of course you are correct and I think I was suffering a little brain fog in suggesting that the pumps were both contributing to the flow. Astonishing rain and wind here and some flooding so have only just got around to making further checks.

I had to remove the circuit board to get to the pump and I find it is a Grundfos UPM3 15-75 130. As you suggested I pulled the PWM connector with no effect on the pump speed as shown on the 'Commissioning Wizard' i.e. 17l/min.

The Grundfos pump is showing a green LED which on some models of the UPM3 indicates "No external control", but I can't think that applies in my case where the flow changes when the pump speed is changed from e.g. Setting 3 to 5.

Consequently, I am still perplexed as to why the installed pump is not delivering anything like its rated flow.

To reinstall the second pump would require some plumbing and reliable contractors are hard to find. Additionally one reason I was keen to remove it, was that although it is a PWM pump, the PWM controller was not supplied, i.e. no speed control available.

I am not sure where I go from here.

Posted by: @davidalgarve

↑

I am not sure where I go from here.

I have been wondering that myself.

The fact is that, even with the buffer (meaning that the secondary pump doesn't have to push water through the pipes to the HP and the heat exchanger) your flow rate was hardly spectacular.   So before leaping to any conclusions I think a bit more thought is needed.  Is the heat pump well oversized, is the emitter system simply under specified in terms of its pipework (or correctly specified, but not for an oversized heat pump).  Those are the next questions to answer IMHO.

Can you (a) remind me what the previous flow rate was and was that primary or secondary

Describe your emitter system particularly with reference to things that will cause friction 

Have any measurements of house loss (eg from consumption/heat delivered as measured by the heat pump

Also can you remind us the model of heat pump

and the model(s) or better still performance curves (I think you said you have them) of the water pumps

Also re

Posted by: @davidalgarve

↑

of a three way valve and some additional pipe fittings that would have increased pipe friction

Can you be more specific and I presume its a diverter valve (ie either DHW or space heating but not both)

These may point in a direction.  There are probably 5 choices

• refit the buffer, accept and work around the limitations
• leave as is, accept and work around the limitations
• add an additional water pump
• replace the existing water pump with something beefier
• do something with the heating system to improve throughput

Currently there is insufficient information to determine which is the best option.

@davidalgarve I have deleted your post made in error, I will delete this one also once I know you have read it! 

@jamespa

Thank you James, both for the comments and removal of my post made in error.

Before answering your questions I think I should say that the heat pump was installed at a time when LPG prices were increasing at an alarming rate and the gas boiler was failing. I used a previously reputable dealer, but subsequently discovered that the founder had moved back to UK. I have no information about what, if any, calculations had been made re the heat loss or the radiators.

The installation was made just before Christmas 2021 and I subsequently found out that it was set for fixed temperature (50C) and consequently, it was a very expensive winter. I played around with Tapo’s on most rads to get heat to the most important rooms and to get a facility for setback.

After learning about weather compensation on the forum, I made the necessary changes and we had two comfortable winters in the house and the elimination of mould in cold corners.

Continuing to study the forum, I was persuaded that I should remove the buffer tank and to a lesser extent, change from weather compensation to adaptive and both were adopted before the onset of this winter. I now know that adaptive does not work for me, but I am perplexed at the inferior performance of weather compensation following conversion of the buffer tank.

So, to answer your questions:

 I have a record of 23l/min at Setting 5,  but this is in primary flow to the buffer and was before the addition of DHW

I have 12 Roca aluminium radiators and 2 towel rails all connected with copper pipework, but must stress that before the buffer tank conversion and removal of the secondary pump, it was adequate most of the time on setting +0 to +3 on the FTC override.

I don't have any direct heat loss calculation but now have MelPump data if that can give any useful information.

Heat pump is PUHZ-SW120YHA which I had previously quoted as 16kW but is actually 12kW

Primary pump is a Grundfos UPM3 15-75 130 and it is PWM controlled from FTC6. Brochure for this is here:

Performance curve is (I think) here:

Mitsubishi have confirmed that "the target flow rate is 32 lpm with a minimum of 14."

The secondary pump (removed) is a Grundfos UPM2 25-70 180 and performance curve is here:

Note that this is also a PWM pump, but no facility was ever provided for changing speed.

I can't help feeling that the low flow output from the UPM3 is being controlled by the FTC6 in some way but can't find what. This makes me reluctant to reintroduce a second pump, especially the one removed, which does not have speed control.

The 3 way valve is either space heating or DHW, here:

Once again, I am most grateful for your consideration and would appreciate any further comments,

@davidalgarve 

Thanks, thats great data on the pumps and as you say different but not much.

I have some ideas forming both by way of explanation and next steps, but first will ask a couple more questions so we can do some very basic calculations.

Posted by: @davidalgarve

↑

I have 12 Roca aluminium radiators and 2 towel rails all connected with copper pipework, but must stress that before the buffer tank conversion and removal of the secondary pump, it was adequate most of the time on setting +0 to +3 on the FTC override.

Do you know the pipework diameters (there may be several at different points) and are the rads connected in parallel or series (I believe the French sometimes connect rads in series, I dont know about Portugal)

Is there any significant pipework noise either now or before the buffer conversion?

Posted by: @davidalgarve

↑

I don't have any direct heat loss calculation but now have MelPump data if that can give any useful information.

Yes definitely.  What can you easily get hold of.  Energy in/energy out as a time plot would be good, preferably through the coldest month(s) with OAT

Posted by: @davidalgarve

↑

I can't help feeling that the low flow output from the UPM3 is being controlled by the FTC6 in some way

Some of the UPM3 variants have a display on the pump, can you see yours and if so does it have a display (just some coloured LEDs) and what is it saying.   Did you say that pulling the PWM connector made no difference?  If thats the case then the controller isnt modulating it.

I think its likely that, for some reason, one pump doesn't have sufficient oomph to push 32l/min through.   The questions above, if you can answer them, will help ferret out the reason.  By splitting the circuit in two and introducing a second pump the buffer helped this a bit (but not entirely, your primary flow rate was still low hence why Im questioning pipe sizes!).  We dont so far as I am aware know what DT or flow rate it achieved at the emitters but if you have any data that would help too  Also what pump speed was set on the primary when the buffer was connected?.

To determine most likely route to improve things we need a bit of a handle on pipe sizes and actual loss and, ideally, what was previously happening in the secondary.  Hence all my questions