1. Assuming you have just installed the heat pumps, why did you install the Aerona 3 with the antiquated controller, rather than Aerona R290 with the new Smart Controller?  
2. I am surprised that you need two heat pumps.  I am living in a 5 bedroom detached house built in 1968.  The floor areas is 350m2 spread across ground and first floor.  The house has reasonable, but not top notch insulation - e.g. 30% of the loft insulation is below current building regulations.  It is rated EPC C.  We installed a Grant Aerona R290 15.5kW air source heat pump in August 2025.  It has coped comfortably through the winter with temperatures as low as -4.8C.  

no progress on the flow rate valve windows, thy remain dirty so the flow rate valves are still left wide open.   From what I have read and speaking to Grant Tech too much flow rate (remember both pumps are set to min) is far less likely to cause a problem than too little flow rate.  I think that’s in keeping with comments above too.   So for now this remains on the list for another day.

I’ve spoke to Grant Tech 3-4 times over the last few weeks each time to tune and resolve one issue or another.

1) One of the key things is that the slave pump started running at 5C outdoor air temp regardless of the slave engagement settings.

now this might not sound like a big problem at first but given I have master and slave on open loops and plumbed in parallel it meant that when the salve pump was running it was pumping out equal volumes of return flow temperature water back into the outbound flow and in effect halving the temperature the master thought it was heating the loop up to.    So for instance the master would be sending out at 39C, returning at 35C but the slave was then pumping through 35C and thus diluting the total outbound flow at 37C.   This explained why I was seeing a dip in room / floor temps during 5C and 3.5C when the slave actually engaged its compressor.

It turns out this was due to frost protection kicking in, specifically the one enabled by DIP switch 1 which is factory set to UP which is ON.    Turning this DIP switch off on the slave and also changing params 43-00 = 0 and 43-10 = 0 (which also dsiables frost protection on room and outside air temps) solved this problem completely!

2) once resolved this enabled me to focus more clearly on the impact of the defrost doom loop.  One night in particular was bad when the outside temp dropped from 5C to -1C in just 3-4 hours.   This drop would have resulted in my WC curve adding around 3C to the target loop temp and combined with the falling outside temp was simply too much too quickly for them to catch up.   
Again after speaking with Grant Tech the advise chopping down the WC curve increase.  They suggested simply reducing the top end but I decided to tackle both ends, raising the bottom end a little too.   In total I changed my WC curve change from 12c to just 7c.

old settings were 30c@23c and 42c@-3c

new settings are 33.5C@23c and 40.5C@-3c.

a couple of days later and everything was much more stable.  Small gradual changes are much more likey not to cause the defrost doom loop than sudden rapid increases.

3) then I noticed that the slave was still being slightly affected whenever the master went into a defrost cycle, whereas it didn’t happen the other way around.  Specifically what was happening was the very briefly the master would interrupt the power to the slave at the start and end of the defrost cycles, but this meant the slave then went into a 3min compressor shutdown followed by a 3min compressor start up, all in all meaning I was loosing 12mths of slave running time at the critical moment when I needed it most I keep running.    I sorted these issue myself by reading the Chofu manual on backup heater (param group 46-xx settings).

Specifically I found that Grant had advised 46-05 =2 seconds and it should have been 0 seconds.   This paramater is the backup heater startup delay in seconds, so everytime there was a change in status for the backup heater (the slave) the power would be interrupted for 2 seconds causing a compressor cycle!

Next was setting 46-23 =24C default.    Grant gave no advise on this setting just as they had not given any advise on setting 46-20 to enable the slave during defrost.  Turns out 46-23 is the min loop return temp to engage the backup heater during defrost.   So I set 46-24 to 41C meaning that it I’ll always engage the slave as my loop temp would now always be below 41c.

The overall results have born out to be quite significant.   No more periods of the floors mysteriously getting colder and then suddenly warming up again.   It’s now a much smoother translation and constantly gentle.

looking at power use on my battery / inverter I can also see that the spikes are way less dramatic during cold spells.  Previously power would average 1-4kw over any given 5 min period, whereas now it’s between 2-3kw.  By ensuring the slave runs continuously through the master defrosts (unless of course lightnight strikes twice) then the impact on the loop temp is far less dramatic.    The fact I only have half a house completed and plumbed in means my total water volume is lower than the design so it makes sense that I was much more susceptible to defrosts than I should be once the house is finished, hence fine tuning this was critical.   It’s clearly a good thing to do regardless Id say though.

thats it for now - I’ll keep monitoring and tinkering as necessary over the next few weeks of cold weather.

Thanks again to the excellent advice from both on here on this forum and from Grant telephone Tech support.😊

Power use yesterday was around 70kwh, so I would say a good +10kwh on even the prior bad days.  That said looking at the power use graph on my inverter there was far less saw toothing, which is good thing and no doubt manifest itself as a much more stable temperature.   At least theres some return now on the investment, sinking a lot of electric and watching the temps fall at the same time was a double whammy.

I suspect the high usage is largely down to the warm roof lacking any insulation around 45% of the perimeter edges behind the soffits, yet so we must be loosing a fair old huge chunk of heat through that, not to mention every time an internal door between the new and the old parts of the house is opened, warmth rushes into the unheated area.  The roof will be resolved by mid Jan, the rest of the house, maybe with some luck, before winter'26.

So for now I suspect its likely as good as it can be, and the WC curve isn't far off the mark for a half finished house.   I dunno if I could be getting away with using less electricity but I'm at a loss as to what else to tweak now, other than maybe tinkering with turning down the flow rate, albeit Im also struggling to see how that would help really. 

I have a busier week ahead next week, prepping for the roof works so likely have less time to tinker after today.   By Tuesday we're over the worst of this cold snap so I suspect it will be far less on my mind anyway.

That all sounds very sensible and I agree with your analysis.  Like you I turn on the immersion on the very coldest days, because I run low noise overnight there is a bit of a deficit and this helps mitigate.  Let's see what the next days being but the fact the master and slave are now defrosting in antiphase is a major step forward imho.

I also managed to witness a slave defrost cycle and could see the master continued to run through it.  I knew theoretically it had to but it was nice to see the mirror half process to what I witnessed yesterday.   Again the impact on loop temp drop was dramatically less than before so this change alone has made a significant improvement i feel.

lets see how today goes…

i appreciate it will be heating less water per min, but having to increase the water temp by more C per litre surely offsets any savings made?

anyway I think today’s job has to be to try clean out those filters.  It’s dark in that area as the lighting hasn’t been fitted yet and the pipes are all double lagged but I cannot see anything resembling a tiny mesh filter, just two large Grant Mag-One magnetic filters, one on each of the master and slave pipework. Fortunately the plumber has left the removal tools and spare o-rings in a very obvious place so thumbs up to him for that.

so I’m going to read the manual for them, watch a couple of vids on YouTube for similar mag filters and give it a go at cleaning them over lunchtime when the sun is out and we’re getting a bit of solar gain into the rooms to offset the fact I’ll need to turn the ASHP’s off whilst I isolate and clean the filters.

in terms of power use which you were asking before.  On a good day when the temp is around 4C or more and we are just before the defrosts start to happen the ASHP’s have been using around 38kwh during a 24hr cycle for both heating and HW.

on bad days when the defrost doom loop has been happening that is more like 60kwh during a 24hr cycle.

i made one more change yesterday too given that the next 5-7 days look extremely cold.  I’ve disabled the HW cycle on the ASHP’s and enabled an immersion heater timer in its place.  That way the ASHP’s will now run for an uninterrupted 24hrs a day.  

My plan is that this will give me a better run at understanding any changes I’m making is having and also that the additional power used by the immersions for a couple of hours in place of the ASHP HW cycle will be offset in part by the smoother running of the UFH not having to recover.   TBH an extra 3-4kwh per day for the immersions is reality small fry compared to what the ASHP’s are doing!

As far as I can remember the reason why it caused more defrosts was that it kept the compressor running constantly which caused it to ice up.   Keeping a bigger differential meant it would have periods of recovery meaning the coils didn’t frost up as quickly.

Either complete Bow Locks or at best a marginal effect.

The amount of energy it needs to produce is fixed by your house.  If you force it to produce that amount in a shorter time then it will have to work harder.  A bigger differential wont result in shorter running, it will result in a lower average flow temperature.  If anything that is likely to increase the running time.

The only argument I can see is that if you are forced to operate at a higher FT because of the higher DT, the COP is poorer so the amount of energy that is extracted from the air is less (and the amount you have to supply more) so the fins cool a bit less for any given amount of energy supplied to the house.  Thats defeating the purpose of having a heat pump and taken to its logical extreme is an argument for resistance heating alone.

Remember that Grant don't make heat pumps, they resell far eastern heat pumps, until recently with some add-ons of their own design which cripple the performance of an otherwise perfectly good underlying piece of hardware.

googling it does seem to support that as common knowledge that too high or too low flow rate is really bad.  I’ll give it a go for now.

Yeah, and google can tell you all sorts of untruths these days. 

Too high is bad because it reduces the average emitter temp for any given FT thus reducing the efficiency because you have to jack up FT to compensate.  Simple physics.  Too low - no evidence to support this that I have seen and, more importantly, no explanation for the mechanism, unless its so low that you are really working the water pump too hard.  Thus suspect BS.

I would also give it a go just so I can tell Grant I had done so, but with little expectation of success and a recognition that it will probably increase running costs.

next step then is how to clean the filers in that case.   I’ll google it, presumably there’s plenty of hand vids on youtube…

Note there may be two filters - a largish magnetic one about 200mm high cylinder, and a small mesh filter perhaps only about 50mmx50mm.  The latter will be closer to the heat pump than the former, mine is outside immediately behind the pump and I didnt know it existed until it blocked up!

will cleaning the filters then cause the tung to slowly fall off meaning I can view the flow rates again or will they need dissembling and cleaning properly?

Maybe. maybe not, impossible to tell TBH.

to be honest it’s really quite dry here at the moment albeit is cold.  Outdoor RH is 67% according to the inlet sensor on my MVHR which is a couple of metres away from the ASHP’s.   I had never actually seen any frost on the compressor fins but I havnt closely inspected them either.

anyway yes I really couldn’t understand why it would help turning the flow rate down so I asked him to repeat it so I would understand.   Made little sense to me either but googling it does seem to support that as common knowledge that too high or too low flow rate is really bad.  I’ll give it a go for now.

next step then is how to clean the filers in that case.   I’ll google it, presumably there’s plenty of hand vids on youtube…

will cleaning the filters then cause the tung to slowly fall off meaning I can view the flow rates again or will they need dissembling and cleaning properly?

their advice was that my flow rates are likey too high which is what is causing excess defrost cycles.  

I presume he didnt say why?  I dont personally understand how this can be the case (so suspect BS) but of course willing to listen to arguments.

Ideally I should have 2-4 defrost per day whereas I can currently having more like 12-18 of them!

Interesting.  2-4 would be great but certainly not what I and others see on the damper days.  12-18 more likely as far as I am concerned and some report 1 every 45 mins (not with Grant, but the physics is identical). 

Both ashp pumps are already set to min, so he suggested turning down the flow rate valves a little at a time so that the differential between flow temp and flow return was min 5C ideally 6-7c.  Currently my differential is around 3C.

Stupid idea to impose unnecessary restrictions in the pipework IMHO or to deliberately increase DT, unless he can explain the mechanism by which this will improve defrost.  By all means try it but my advice would be to reverse it out if it doesn't make a difference!

If you increase DT from 3 to 7 you will, at least in principle, need to jack your FT up by 2C to compensate so that the average emitter temp, and thus output, stays the same.  This costs 6% in efficiency.

any ideas as to why a 12mth old system would be gunked up so badly?

possibly not properly flushed, possibly just residual cr*p.  My filters blocked after 6 months. 

he also talked me though a parameter 46-20 which is used to ensure the slave (or backup heater as it was intended for) remains active during master defrosts.  Mine was set to disable (0) so we’ve changed it to (2) which ensures the slave keeps running when the master defrosts.  

Now that does sound sensible!

Good luck, please post back results.