Re the chart that I posted.  Could someone comment as to why the subsequent defrost cycles do not perform as per the the first at 3.30pm?

Posted by: @trebor12345

↑

Just to add.  As the system is not that common I have been unable to find anyone local who knows anything about the system or what I can do to fix it.  The manufacturer has been involved, but whilst it's improved, there are still issues.

I appreciate that it's going to be difficult to find a Heat Pump that modulates down to less that 2kw.  The same applies to a gas boiler. 

The problem is that you need to modulate down to <1kW because thats where the temperature sits most of the time.  You wont find a heat pump or a boiler to do this, so whatever you do next will still involve the heat source cycling.

What you need is someone who firstly understands heating not someone who is just familiar with the particular heat pump, although of course the latter helps.  People who understand heating are sadly few and far between!  In the absence of that, people here who understand physics, control systems and heat pumps in general even if not your specific model can endeavour to help from first principles.

Can I request you read my post above and help with the answers to the questions.  Of course there is no obligation, but without some more info its not possible to help.

Posted by: @trebor12345

↑

Re the chart that I posted.  Could someone comment as to why the subsequent defrost cycles do not perform as per the the first at 3.30pm?

Sorry no, but it  is likely to do with the temperatures of slugs of air or water reaching the control sensors from time to time as circulation occurs and/or the OAT and/or the IAT, and the reaction of the control system.  Until its shown that it matters (and why it matters) why do we care what the heat pump is doing internally to its control circuit.  

If you are determined to replace the heat pump then you will need to be very careful to check min output, which is often difficult to find.  Many heat pump manufacturers produce models with lower max output by simply limiting it in software.  The min output remains the same as the higher capacity pump from which it is derived.  Daikin for example does this with it's 8kW model, anything rated smaller is simply the 8kW downrated.  Valiant does it with the 5kW model, the '3kW' machine is simply the 5kW downtated.  Other manufacturers do similar.  Looking at the basic spec will tell you almost nothing about how it might perform in your property.  I expect boilers are no different.  This is another reason I would caution against changing heat pump without understanding the actual problem.

All that said if you want a reasonably low output heat pump with reasonably user friendly controls the Vaillant 5kW is a good choice imho.  Vaullant manage cycling quite well and the main temperature control (a simple dial)  actually changes the WC curve rather than imposing a thermostatic limit.  Programmed setbacks do similar. There are lots of valiant engineers around.  I stress that it's the installer that matters more than the heat pump.

James

I understand your point about turning the heating off and on or up and down and the resulting effect.

Today the one room thermostat controlling the heating is set to 30C (effectively on all the time).  The Thermal Offs/Short Cycles are controlling the temperature of the house.

I have attached a chart of the flow and return temperatures for one of the cold nights/days.  The HP was running with a flow of 33C.  You will see from 3pm that the temperature starts to fall as a result of the defrost cycles.  Once the next day time temperatures rise, at 11am, we are out of defrost mode so the temperature rises.

Weather compensation is erratic, effected by the sun shinning on the rear temperature probe and also when the HP does it defrost cycles.

I should add that the graph is from some temperature probes I have attached to the pipes.

Posted by: @trebor12345

↑

James

I understand your point about turning the heating off and on or up and down and the resulting effect.

Today the one room thermostat controlling the heating is set to 30C (effectively on all the time).  The Thermal Offs/Short Cycles are controlling the temperature of the house.

I have attached a chart of the flow and return temperatures for one of the cold nights/days.  The HP was running with a flow of 33C.  You will see from 3pm that the temperature starts to fall as a result of the defrost cycles.  Once the next day time temperatures rise, at 11am, we are out of defrost mode so the temperature rises.

Weather compensation is erratic, effected by the sun shinning on the rear temperature probe and also when the HP does it defrost cycles.

I should add that the graph is from some temperature probes I have attached to the pipes.

 

All noted.

I can see whats happening but the drop in FT appears to occur just before the start of the plot and the onset of defrost, so is not obviously related to the defrost based on this evidence alone.  Comparing it to the room temperature it could just as easily be caused by a room thermostat with a 1C hysterisis (in fact I would say that is a closer correlation - note that the FT also starts to rise before the defrost cycles stop, just as the room temp hits 20C)

I would like to help but honestly cant tell whats going on here and dont think there is a chance of being able to do so unless you can answer my questions in full and provide a self consistent set of data (ie data all related to the same time period).  For example you have provided a plot above for one day, but said how the controls were being operated on a different day.  Given you say you have to constantly change things its impossible to make any deductions without knowing both for the same period of time

• You obviously have some data.  Can you provide several consecutive days of room temperature data (the week up to today would be good), flow temperature data (heat pump side of buffer) and, if you have it, OAT, and most importantly what you were doing to the controls (any and all controls) and when?
• What temperature do you want the heating to be at?
• If its not set to operate 24*7 what is it set to do or is that changed constantly also?
• Is the system in 'low noise' (quite/silent) mode
• have you tried to set up weather compensation and failed/abandoned, or have you never set it up.  This is mechanism built into your heat pump to look ahead to see what the weather is going to do to your house thus saving you from having to do so!  With low loss and UFH this becomes vital, because control on room temperature alone is almost guaranteed to fail!
• If you havent set up WC is it set to run at a constant flow temp and if so what is that (you have said 31 and 33 above, which or are you manually varying that?)
• Where is the thermostat located relative to where you have measure the room temp.  What make and model of thermostat or is it the internal one
• Where is the hitachi controller located and is is set to ignore its own temperature sensor?
• What controls the secondary pump (ie the pump on the buffer side of the UFH)?

If you can provide some consistent data it might be possible to work out whats going on and advise.

James

Thank you for your help.  I will work over the next week to try and produce the data requested.  This is still new to me so still a step learning curve.

Posted by: @trebor12345

↑

James

Thank you for your help.  I will work over the next week to try and produce the data requested.  This is still new to me so still a step learning curve.

Good.  No need to apologise, despite the fact that the principles of heating are extremely simple, the wide variety of houses and the constantly varying outdoor conditions makes it challenging to understand in practice.

Sorry to ask for so much but its very difficult to tell whats going on when things are changing a lot, particularly when one of the key changes which seems to be causing concern happened just before the period illustrated, and you are saying that you are frequently changing the controls.  

This week can I suggest you change as little as possible (preferably nothing).  If you do change anything keep a log!  Its really the data from last week that is most interesting because thats when defrosts, which you say are a problem, were occurring, but data from a period of stability is also useful too.

Even if you do end up changing the heat pump this exercise will likely identify features about your house which need to be considered in any new system design!

Others may of course have better ideas than I do!  I have a relatively high loss house so no direct experience of the situation where its going to be impossible to match heat output to demand even in the height of the season.  My comments are thus based largely on 'first principles' (basically physics) and what happens in my house towards the ends of the season.

@jamespa My last house was a low heat loss one. 120m2  +UFH with a 5 kW ecodan. So a similar situation to @trebor12345

Unfortunately I didnt have monitoring but ran the heat pump more like a coventional boiler, batch heating at certain times of the day. Defrosts were never a problem but that might be due to the climate in Swansea.

Posted by: @bontwoody

↑

@jamespa My last house was a low heat loss one. 120m2  +UFH with a 5 kW ecodan. So a similar situation to @trebor12345

Unfortunately I didnt have monitoring but ran the heat pump more like a coventional boiler, batch heating at certain times of the day. Defrosts were never a problem but that might be due to the climate in Swansea.

Thats really helpful.  The plot produced is odd and to me smacks of flow temp reduction triggered by a control not by defrost.  The temperature soon recovers after each defrost, the fall happens before the onset of defrost and the rise starts before the end of the defrost period.  Quite honestly it looks very like a thermostat (human?) with 1.5-1C hysteresis (typical of a human), but with only half a cycle and no explanation we cant be sure.

Comforting is the slow drop in room temp with, in reality, not much heating.  At an average FT of 26C the room temp takes 18hrs to drop 1.5C, and this is when OAT is pretty low.  That means the house wont follow any 'natural' cycling by the heat pump, other than cycling based on thermostat/room temperature sensor.  So with the right WC curve, sensible hysteresis settings, and, as you say, an element of batch heating, it should be possible to achieve adequate stability.

Can I ask, @bontwoody ,with your previous house apart from batch heating how were the controls set up.  Given the reaction time I cant see how OP's house can possibly be stable run on IAT thermostat (and certainly not human controls, which is what he says he is doing), how was your run?

Of course OPs is further complicated by that blessed, pointless, and wholly inadequate buffer and (presumably) a secondary control loop trying to do its own thing quasi independently of the heat pump.  Somehow, if batch heating it the way forward (and I agree it is), that needs to be synchronised to the heat pump, otherwise we will likely end up with the heat pump switching on according to its batch schedule, but the buffer/secondary control loop refusing to forward the energy to the UFH!  Its tempting to start with the secondary pump running 24x7 and then, once its refined (or the a decision taken that it wont work) try to finesse the secondary pump control.  That of course is assuming the buffer isnt located in a unheated garage. 

Grrh - I hate buffers!

@bontwoody 

Given your experience from your previous house and given op has  a 8kW heat pump for possibly a 2.5- 3.5kW house which takes 18hrs to cool by 1.5 deg, what do you think about this control strategy, at least as a starting point

Secondary pump on 24x7 until we find a way to sync it, thermostats etc to max

Heat pump on for 3 off 4 hour periods evenly spaced

Pure weather comp, all room comp disabled

FT of 22C at 15, 35 at -2

Adjust the last of these parameters to get the temp right, but do so  very slowly  no more than 1C every 24 hrs

Dhw heating scheduled outside the space heating windows to avoid confusion

Monitor iat and ft during setup period