Observations Commissioning ASHP

[JohnMo]

New build originally designed around a gas boiler and UFH, flow temp are low as could be, due to 300mm pipe centres.

 

So previous setup was, gas boiler, big buffer, mixer and pump on the manifold. The flow temp out the buffer was consistent, but filled based WC (Weather Compensation) curve. Flow temps varied between 26 and 34 degs (at -9 OAT).

 

The heat pump is installed without a buffer, all UFH loops open, with no actuators, mixer or pump on the manifold, so as simple as it gets. Have a single thermostat, just to give a start permissive for the heating.

 

Couple of observation

1. The min heating temperature of the heat pump is 25 degs, dT is 5.4 degs, so a return temp would be 21.5.

2. After one heat cycle at a low flow temp (26 degs), the floor doesn't cool down enough to allow to ASHP to start up again. So heat pump would not restart. That didn't work

3. At low loads heat pumps cycle, I typically have a 10 to 20 minutes run time and 45 minutes off.  The recirculation pumps continues running at all times. If the heating load decreases (OAT, solar gain etc), the heat isn't transferred out the flow as quickly, so the off period of the ASHP increases.

4. To compensate for the on/off run nature of the ASHP, and to get the required heat into the floor, you need to flow at higher temperatures at low loads, than you would expect.

5. So to make up for the on/off nature I increased the flow temp to 29.  The flow temp pattern through the floor is typically (8 deg day), heat pump ramps up to set temperature and when it cannot modulate to control dT the heat pump shuts down. The flow temp and return temperature stabilise down to about 24 and flow will slowly cool over over the next 40 minutes, down to about 23 and the heat pump starts.

6. Very small increments in flow temp have an impact on house temp, less than 0.5 degs change in target temp, can be felt in the house.

7. Been tracking the temp on a Shelly H&T today, temperature drift in the house is around 0.2 degs.

 

Have tested the 3 basic modes of operation so far, operation controlled by thermostat, batch charging overnight on cheap rate, and continuous running 24/7 on straight weather compensation. So far leaving it to run all the time seems the cheapest to run and easiest to implement.

 

Have a simple Shelly H&T set up in the hall, to record how temperature moves over time and OAT changes, this will allow fine tuning of the WC curve.

 

 

[ReedRichards]

How warm is it inside the house?

[JohnMo]

26 minutes ago, ReedRichards said:

How warm is it inside the house?

 

We have it at 20 degrees, the temp over the last 24 hr has been very stable. Found that any hotter is too hot with the UFH, much cooler you are looking for a jumper. However there was zero sun yesterday, as it was peeing down all day, 8 degs in the day. Outside temp in the evening and overnight dropped to 4 degs for about 10 hrs.

 

More of a brighter day today, so will get some solar gain, so will see how things react in the house and the ASHP.

 

This was the cycles of the heat pump on a 7 deg night (couple of nights ago)- short runs long off time. (Green spikes are the ASHP)

 

Last night at 4 degrees, run time extends with one run over 2 hours. Some of the bigger spikes are the dishwasher, the one 7am is kettle.

 

House temps, the big spike was the H&T in my hand.

 

 

[JohnMo]

Still using the weather compensation without any thermostat intervention to see what happens on different days. So yesterday was foggy in the morning and sunny in the afternoon but 8 degrees and by 6pm was down to 2 degs. We had some solar gain in the afternoon which warmed the front rooms of the house.

 

So a couple of plots to show what happened

 

This is the H&T in the hall, the heat from the solar gain moving into the hall, and slowly going back to the starting temperature over night.

 

So couple of busy views, this is yesterday evening.  The afternoon plot shows a slow down of the heat pump when the sun was out. The yellow is solar being generated.

 

A 2hr window when the ASHP didn't add heat to the house.

 

 

The evening when outside temp is 2 degs

 

And this morning, you can see the run times getting longer.

 

So although the ASHP isn't stopping when the room temp goes higher, it is sensing from the return water what is happening in the house.

 

As for energy consumption the last 24hrs we used 12kWh on the ASHP and all auxiliary equipment.

 

The previous 24hrs, with no solar gain and slightly warmer temperatures 9.5kWh

 

Against a calculated heat loss of the garden room and house of close to 60kWh. So all things being equal a CoP of around 5. Which is inline with what to expect from the technical manual, running temp and

 

[Jenki]

Would a volumizer not help in this situation?  It sounds to me like the water volume is not enough, and you are getting short cycling?  - but what do I know, I can't get mine to work😥

[JohnMo]

The unit is really bigger than I need, but at the moment it's just outside short cycling and quite controlled. Short cycling for an ASHP is anything less than 10mins. Generally getting a good 15 min or more. Doing batch heating it was running for 4 hrs non stop some days. But the CoP wasn't any better.

 

The big reason for running 24/7 is to displace the standby heating I was doing on our summer house, this required an electric heater set at 14 deg 24/7 and was using around 5 to 6kWh per day on its own. All of yesterday and so far today it's only used standby power for the thermostat 10Wh.

[JohnMo]

Starting get cold nights, was down to -1 last night. So using the H&T sensor to track temperature inside the house.

 

 

The house temperature gradient is a downward slope, so to compensate, have increase the end of the WC curve up 0.2 degrees. We have another cold night, tonight expecting it to be -2, so will see if the change has made any difference.

 

Also with colder temperatures the run time of ASHP has increased. Now running just over an hour at a time overnight, with around 30 mins off.

[ReedRichards]

On 05/11/2023 at 09:43, JohnMo said:

Short cycling for an ASHP is anything less than 10mins.

References?  I sometimes wonder if short cycling, which is a valid concern for fuel-burning boilers, has been carried over as a concern to heat pumps without justification.  I doubt that anyone worries about short cycling with an electric boiler so what goes wrong if a heat pump?  

[JohnMo]

8 minutes ago, ReedRichards said:

References

It was from a government sponsored report, no longer have the file, basically gas boiler and ASHP are around 10 mins, GSHP, 5 mins. Meaningful run time when you take into account the heat input required to heat the metal work pipes and a slug of water enough to transfer heat to the house.

 

An immersion heater isn't the same at all.

 

Ignore or do what you will with it, I basically made sure, I could achieve a run time of 10 mins or more without a buffer.

[ReedRichards]

What you have achieved is excellent and I'm sure it's best to play safe and have long cycles rather than short ones if you can.  But in the same way that some concepts that apply to internal combustion engine cars don't apply to electric vehicles (such as optimum mileage per unit of fuel at around 56 mph) I wonder if short cycling really is bad for heat pumps?  

[JohnMo]

20 minutes ago, ReedRichards said:

What you have achieved is excellent and I'm sure it's best to play safe and have long cycles rather than short ones if you can.  But in the same way that some concepts that apply to internal combustion engine cars don't apply to electric vehicles (such as optimum mileage per unit of fuel at around 56 mph) I wonder if short cycling really is bad for heat pumps?  

This isn't the report I read but very similar

 

7389-effects-cycling-heat-pump-performance.pdf

[sharpener]

Might also be a hangover from the age of non-inverter drives where the high electrical and mechanical stress of frequent direct-on-line motor starts will shorten the life of motor and compressor. Wouldn't think this a problem now.

 

The above report's main conclusions (in 2012) were

 

Both ASHPs and GSHPs should be designed for a minimum run time of circa six
minutes, which will avoid the worst excesses of detrimental performance caused by
short cycling...run times of ~ 8 minutes for the ASHP gave COP
values close to catalogue steady state values.

 

which is significantly shorter than the 15 mins often quoted. So perhaps this problem is less severe than many people seem to think?

[ReedRichards]

Also the report is testing heat pumps with "single fixed speed compressors".  I take it that means heat pumps with non-inverter drives that were incapable of modulation.

[sliderule]

When you have to get a new heat pump after repairs to the heat exchanger and the compressor failing, as we have had to after 12 years of use, you will realise that short cycling is bad. It also means that your UFH is periodically getting colder flow water which slows down the warming of the room. I discovered that ours was short cycling when we got a smart meter and I tried to prevent it. I'm sure it caused our problems. I did the heat loss calculation and got a lower power unit. Now it only cycles off once or twice a night, but it's early days yet.

[JohnMo]

41 minutes ago, sliderule said:

Now it only cycles off once or twice a night

 

I am thinking about reducing the flow rates to the floor loops. A quick calculation shows between the summer house and the house

 

House manifold flow 12.4

Summer house manifold flow 2.

 

With ASHP dT, I am pushing around 5.2kW into the floor. But have a peak demand around 3.5kW.

 

This should allow the ASHP to run at a lower output and cycle less.

 

[sliderule]

If your ASHP is already running at it's lowest power, reducing the flow will cause it to cycle. In that situation I think it's better to run it at a flow rate that avoids cycling, and let the floor heat to your required temperature and shut down for a few hours.

[JohnMo]

It's a 6kW heat pump, with a flow and dT that was demanding 6kW. Reducing the flow has allowed the heat pump to modulate down further. 

 

Overnight it was down to -2 OAT and the heat pump ran for 7 hours straight, except to defrost.

[JohnMo]

Had DHW heating and cooling running in the summer, once commissioned, no issues. So thought when I installed the UVC there would be no issues when heating instead of cooling. But this hasn't been the case.

 

It looked like the DHW heating cycle finished and the heat pump didn't start on house heating again. First couple of times I just shutdown the heat pump at the isolation switch and all reset and worked fine. However yesterday I actually watched at the heat pump, to see what was happening.  The heat pump was trying to start in heating mode, after a very brief off period when it completed DHW heating. But during the restart it was overshooting the set temperature point by quite a margin, and then seemed to be locking itself out.

 

So issues with my install, the heat pump is around 14 to 15m from the cylinder, so although the temperature in pipes drops fairly quickly, at the end of the heating cycle, there was still enough heat being carried that adding a small amount of extra heat was enough to overshoot.

 

So have implemented a few changes

 

1. Found in the controller the PI regulator integral time, was set to 255 seconds (max setting). This setting apparently is good for large flow rates and large volumes. So ok on a big heat pump, not ideal on not a small one.  I reduced this figure to 60. You actually hear the difference it makes as the compressor starts, as it rains in the temperature rise very quickly.

2. I added a forced 10 minute break, in the underfloor heating timer of 10 mins after the DHW heating period.

3. Increased the shut down hysterisis from 0.2 to 0.4. This allows the heat pump to run at a slightly higher temp than set point.

4. Increased the UFH flow rates to the previous ones.

 

Now UFH resumes around 30 mins after DHW, which fine. But the run time of the heat pump has increased significantly. Previously it would do an on off cycle once per hour, now it seems to run for around 3 to 4 hrs straight. This has allowed me to reduce the flow temp slightly.

 

 

[JohnMo]

Final update

Been faffing about in the background, to get the house heating and summer house heating to work together, one is high inertia (thick screed UFH) and one very low inertia (low volume fan coil).

 

Final solution is, have dumped WC altogether and now flow set to a fixed demand temp of 35. That suits the fan coil in the garden room and a slow batch charge of the floor. Observation the heat demand of 35, the discharge temp from the heat pump never actually gets to 35 for the 10 to 12 hrs it runs, at the moment. Most days it gets to about 33 by the end of the heating cycle. Zero cycling occurs, except for defrosts.

 

There are two thermostats, one in house and one in the garden room.  The house thermostat can call for heat and changes set point of an electronic UFH mixing valve, the garden room thermostat can call for heat.

 

The electronic mixer valve in the house, is selectable to two different adjustable set points, currently set to either 35 Degs or 27 Degs. The 35 Deg set point ensures mixer is fully open and zero mixing occurs.

 

The system volume isn't large enough to support the heat pump driving a single fan coil on its own, so the house floor is always available to act as buffer via the 27 set point on the mixer valve.

 

Batch charging is simply controlled by a single room thermostat, with a hysterisis set to 0.1 - at 00:30 it is set to 20.5 (starts heating if at or below 20.4), at 07:30 it is set to 20 (switches heating off when above 20.1) and at 12:00 it is set 19.5, makes sure heating doesn't restart.

 

DHW heating is timed to allow heating any time between 1pm and midnight.

Edited February 12 by JohnMo

[JoeBano]

How’s the myson Ivector? Would you say it’s a good piece of kit? 

[JohnMo]

55 minutes ago, JoeBano said:

How’s the myson Ivector? Would you say it’s a good piece of kit? 

 

It's not bad, nice and good quality build, seems good quality overall, works well, now I have my head around flow temperature. Pretty quiet overall, you can program so it doesn't boost the fan at the high speed ever and only does the lowest fan speed overnight.

 

It has an adjustable threshold for water temperature (for cooling and heating) where it switches off the fan. If you set this too low it cools the room instead of heating, if you are not careful.

 

Water temp target temp need to at about 33 to 35, so the water temp doesn't drop below about 28 in any heat pump off cycles, otherwise the fan just drops room temperature.

 

It comes as standard equipped for cooling.

 

Comes with its own programmer, user defined schedule programming etc.

[JoeBano]

Are you planning on using the myson Ivector for cooling in the summer? Love to know how it effective it is, I’m planning a loft conversion hopefully end of this year. I’m going to need some cooling for the hot summer days. 

[JohnMo]

On 12/02/2024 at 12:48, JohnMo said:

Final update

Been faffing about in the background, to get the house heating and summer house heating to work together, one is high inertia (thick screed UFH) and one very low inertia (low volume fan coil).

 

Final solution is, have dumped WC altogether and now flow set to a fixed demand temp of 35. That suits the fan coil in the garden room and a slow batch charge of the floor. Observation the heat demand of 35, the discharge temp from the heat pump never actually gets to 35 for the 10 to 12 hrs it runs, at the moment. Most days it gets to about 33 by the end of the heating cycle. Zero cycling occurs, except for defrosts.

 

There are two thermostats, one in house and one in the garden room.  The house thermostat can call for heat and changes set point of an electronic UFH mixing valve, the garden room thermostat can call for heat.

 

The electronic mixer valve in the house, is selectable to two different adjustable set points, currently set to either 35 Degs or 27 Degs. The 35 Deg set point ensures mixer is fully open and zero mixing occurs.

 

The system volume isn't large enough to support the heat pump driving a single fan coil on its own, so the house floor is always available to act as buffer via the 27 set point on the mixer valve.

 

Batch charging is simply controlled by a single room thermostat, with a hysterisis set to 0.1 - at 00:30 it is set to 20.5 (starts heating if at or below 20.4), at 07:30 it is set to 20 (switches heating off when above 20.1) and at 12:00 it is set 19.5, makes sure heating doesn't restart.

 

DHW heating is timed to allow heating any time between 1pm and midnight.

At the end of the heating season, I started testing - did I really need a buffer, and did I really need a mixer.

 

The answer was no.

 

I also challenged myself did I need a set point temp of 35, again over a few weeks I reduced to 32.

 

Thought after those finding I would simplify the items not needed in the system.

 

So buffer deleted out of system and while I was at it I added a heat meter to flow from the heat pump. Can use it to get a proper idea of CoP, next heating season.

 

Next I looked at the mixer, so this was removed. Within the same pipe work is a diverter valve for UFH or cylinder and a return pump from the cylinder (long pipe run). So reconfigured to get some proper hydration separation, so configured as a close couple tee.

 

So flow is normally to 3 port diverter, then into UFH. Demand for cylinder heat, CCT pump and 3 port diverter are energised. Diverter gives a flow path back to the heat pump.

 

Towards the end of March I had the house thermostat set as per my original setting, but increase the summer house target temperature so it had no setbacks. The house floor acting as buffer, whenever the summer house called for heat. The house temperature remained quite stable as the floor was getting regular top-up of heat. In fact the house stopped calling for heat at all even with outside temp below zero.

 

[JohnMo]

Further to the last post, several updates, all simplifications - less stuff.

 

Problem 1. I could never get the heat pump to flow at low temperatures without either, short cycling, or the compressor never switching back on again after it's first cycle. The min flow temp I could achieve in heating mode was 29. Now set to 26.

 

Solution is discussed here

Problem 2. Was heating the summer house effectively without high flow temps typically above 30. This is somewhat fixed by the above (problem 1), I can now get a much lower target flow temperature. And use the house floor as a buffer and rob the heat in the house floor to heat the summer house. Have set the fan on the summer house fan coil to operate at 21 degs water flow temp. This keeps the summer house at a steady 20 degs.

 

Problem 3. Wife confusion with all the buttons and thermostats. What did what? Basically a house thermostat, summer house thermostat and a different selection switch for heating or cooling. So have simplified this also.

 

Heating is now always enabled and on full WC. No thermostat in the system, and does all control from flow/return temperature and outside temperature. Aim keep house reasonably stable at 20 to 21 degs when in heat mode.

 

When not actually heating/cooling circulation pump runs at min speed (0.6m³/h). This moves heat around the house and into summer house. At lowest flow temp (10 degs OAT and above) ASHP compressor will not start until the return temperature to ASHP drops below 21.1 degs.

 

House thermostat is set in cooling mode. When house is at 23 Deg and above,  instead of switching heat pump off it changes mode from heating to cooling. In cooling mode the ASHP looks for a return flow temp above 19.4 degs before actually cooling.

 

No thermostat in summer house or house except cooling and switching thermostat.

 

Modes are any temperature below 23 deg heating enabled, above 23 cooling enabled.

 

So now almost in set and forget mode, just may need to adjust WC curve once it gets cold.

 

Problem 4. DHW have changed this a number of times. Currently set to heat twice, once at 0600 while in E7. This gives wife plenty of hot water for morning shower and lunch time. Then at 3pm it's heated again (gives PV time to heat if available) so plenty of hot water for evening washing up and my evening shower 

 

Been quite a steep learning curve.

 

 

[JohnMo]

Forgot to include the weather compensation curve.

 

It starts at 10 degs OAT at a target flow temperature of 26 degs and goes down to -5, where flow is 31 degs. So flow temperature increases 0.33 degs with every degree drop in OAT.