ASHP/ buffer tank issue

[ReedRichards]

11 hours ago, Rupe30 said:

I spoke with the UFH providers who had designed the system, they informed me the mixing valves should be set to 4 with my ASHP flow temp set to 50 degrees. 

 

 

If that is true then it's a terrible design.  The whole point of running UFH with a heat pump is that UFH should not require a high flow temperature from your heat pump so you achieve the most economical operation.  My retrofit system was designed for radiators and the maximum flow temperature is 50 C but if I had UFH I would be looking for a maximum flow temperature of 40 C or less.  Also the water circulating in the UFH would make your floors uncomfortably hot if it was anywhere near 50 C.    

[Rupe30]

On 22/12/2022 at 08:50, ReedRichards said:

 

If that is true then it's a terrible design.  The whole point of running UFH with a heat pump is that UFH should not require a high flow temperature from your heat pump so you achieve the most economical operation.  My retrofit system was designed for radiators and the maximum flow temperature is 50 C but if I had UFH I would be looking for a maximum flow temperature of 40 C or less.  Also the water circulating in the UFH would make your floors uncomfortably hot if it was anywhere near 50 C.    

With the mixer valves set to 4 that could well be dropping the flow down to a max of 40. Am I better off setting the heatpump to 40 and leaving the mixers at 6? 

 

I have found sitting on the floors uncomfortable upstairs so interesting you say that. 

[ReedRichards]

21 minutes ago, Rupe30 said:

Am I better off setting the heatpump to 40 and leaving the mixers at 6? 

 

Financially yes, that would be much better.  Make the heat pump flow temperature as low as you can get it whilst still keeping the rooms warm enough.

[Nickfromwales]

Leave the mixers alone? Just drop the ASHP flow temp down, and the mixers will always, then, be the 'safety' cap on the max input into the floors. Opening the mixers to 6 is unnecessary.

[Rupe30]

Little update, my buffer tank is actually 180 litres.. (apologies!)

 

The main issue I have is when I have a large proportion of the house calling for heat (approx 200sq/m), the heat pump is unable to reach the 50degree target I have it set to. It usually gets to 41 degrees max which by the time it hits the manifolds is 38.  I currently heat between 4am-7am and 1pm-4pm as I've joined the octopus cosy tariff, which I realise isn't ideal for an ashp but the house seems to be coping well if outdoor temps are above 5 degrees or so. I then have two zones on from 7pm-10pm in areas that have higher heat loss. When I just have the 2 zones on the target flow rate is achieved. And the rooms heat up 1.5 degrees in that 3 hour period. 

 

Looking at my electricity usage, my heat pump rarely draws more than 6kw for heating, it spikes to 10-12kw for DHW. Why is it not drawing more power to get the flow rate up to 50? 

 

 

 

 

 

 

 

[Nickfromwales]

1 hour ago, Rupe30 said:

Little update, my buffer tank is actually 180 litres.. (apologies!)

 

The main issue I have is when I have a large proportion of the house calling for heat (approx 200sq/m), the heat pump is unable to reach the 50degree target I have it set to. It usually gets to 41 degrees max which by the time it hits the manifolds is 38.  I currently heat between 4am-7am and 1pm-4pm as I've joined the octopus cosy tariff, which I realise isn't ideal for an ashp but the house seems to be coping well if outdoor temps are above 5 degrees or so. I then have two zones on from 7pm-10pm in areas that have higher heat loss. When I just have the 2 zones on the target flow rate is achieved. And the rooms heat up 1.5 degrees in that 3 hour period. 

 

Looking at my electricity usage, my heat pump rarely draws more than 6kw for heating, it spikes to 10-12kw for DHW. Why is it not drawing more power to get the flow rate up to 50? 

It'll be down to the settings / commissioning. The heat-pump clearly has permission to work harder for DHW, so it could be given the same permissions for CH too I assume. The curve for the CH operation may also be set too low, so a bunch of options are probably open to you here.

[markocosic]

A 16 kW OUTPUT heat pump will never draw 16 kW electrical input.

 

I'm surprised that you're seeing 10-12 kW during DHW production. Are you running an immersion at all?

 

If the heat pump is not reaching the design output in space heading mode I would be looking at what the system flow rates are. Primaries to top up that buffet vessel that all is fed off may be too small.

 

If you're using a time clock for the heat pump then stop. The poor thing will not be able to keep up with the buffer vessel demand because you're sucking it dry by trying to reheat a giant concrete slab from cold each time.

 

I would ditch all mixers; set the maximum heat pump flow temperature to the maximum that your floor can dissipate without damaging it (e.g 35C); set the weather compensation on the heat pump so that flow temperature reduces to say 23C at 15C external; and allow it to run 24/7.

 

A 16 kW unit won't work with a 13 kW design heat loss if you're trying to run only a couple of hours a day.

 

It's not crazy to try to do that and to rely on charging a big slab on the cheap rate electricity; but it will only work if your best losses are low enough that the slab can keep the whole house warm (yours is to unbalanced) and if your heat pump is oversized enough to deliver 13*24hrs of heat in say 8 hours. 1ph 16 kW > 3ph 40 kW would be needed to do that.

[ReedRichards]

48 minutes ago, markocosic said:

If you're using a time clock for the heat pump then stop. The poor thing will not be able to keep up with the buffer vessel demand because you're sucking it dry by trying to reheat a giant concrete slab from cold each time.

 

On 16/12/2022 at 14:28, Rupe30 said:

We have underfloor heating but only a thin layer of screed (6mm I believe) so the temps need to be fairly high to get any heat into the rooms.

 

 

Doesn't "a thin layer of screed" indicate that @Rupe30is not  trying to heat a giant concrete slab but actually quite a small one?  

[markocosic]

My read:

 

Downstairs is a hulking great slab that's rinsing the buffer vessel.

 

Upstairs has almost nothing.

[Rupe30]

Yes it's when downstairs comes on that the flow rate drops and doesn't recover. The odd thing is downstairs doesn't take long to heat up so I'm surprised it's cooling the buffer tank down so much. Today downstairs was up to temp within an hour, my thermostats do keep heating on for a while even when target temp has been reached, so I've turned the thermostats off manually to allow upstairs to get to 50 which will allow it to heat.

 

If I set the whole system to 35 upstairs will lose too much temperature, where as downstairs would probably be fine at that. 

 

Would turning the mixer down on the downstairs manifold take any pressure off the buffer tank? 

 

I basically can't heat upstairs and downstairs at the same time. 

 

Perhaps I should heat downstairs 24/7 on low flow temp then switch downstairs off and just in heat upstairs at a higher flow temp during the cheap rate hours. Although in I don't think I can't schedule flow temp changes on my LG controller.

 

couple pictures below to see inputs and outputs into buffer tank.

 

 

[Rupe30]

7 hours ago, markocosic said:

A 16 kW OUTPUT heat pump will never draw 16 kW electrical input.

 

I'm surprised that you're seeing 10-12 kW during DHW production. Are you running an immersion at all?

 

If the heat pump is not reaching the design output in space heading mode I would be looking at what the system flow rates are. Primaries to top up that buffet vessel that all is fed off may be too small.

 

If you're using a time clock for the heat pump then stop. The poor thing will not be able to keep up with the buffer vessel demand because you're sucking it dry by trying to reheat a giant concrete slab from cold each time.

 

I would ditch all mixers; set the maximum heat pump flow temperature to the maximum that your floor can dissipate without damaging it (e.g 35C); set the weather compensation on the heat pump so that flow temperature reduces to say 23C at 15C external; and allow it to run 24/7.

 

A 16 kW unit won't work with a 13 kW design heat loss if you're trying to run only a couple of hours a day.

 

It's not crazy to try to do that and to rely on charging a big slab on the cheap rate electricity; but it will only work if your best losses are low enough that the slab can keep the whole house warm (yours is to unbalanced) and if your heat pump is oversized enough to deliver 13*24hrs of heat in say 8 hours. 1ph 16 kW > 3ph 40 kW would be needed to do that.

 

Based on running it 24/7 is it best I keep the same temp in the house all the time? As i'm probably going to run into the same problems trying to increase temperature. 

 

I've set outdoor temp to -4 18 and LWT to 20 42 that should run 23degrees at 15 external yes? 

[markocosic]

3 hours ago, Rupe30 said:

Yes it's when downstairs comes on that the flow rate drops and doesn't recover.

 

I assume by "flow rate" you mean flow temperature out of the buffer vessel

 

 

3 hours ago, Rupe30 said:

If I set the whole system to 35 upstairs will lose too much temperature, where as downstairs would probably be fine at that. 

 

You have tried this already? With the heating on 24/7?

 

 

3 hours ago, Rupe30 said:

Would turning the mixer down on the downstairs manifold take any pressure off the buffer tank? 

 

It should, yes.

 

Try that. 

 

Limit what the slab can draw on startup by limiting the flow temperature to the slab.

 

 

3 hours ago, Rupe30 said:

I basically can't heat upstairs and downstairs at the same time. 

 

You can / should / must heat both at the same time. But it looks like you're got systems designed for different flow temperatures which is unhelpful.

 

 

3 hours ago, Rupe30 said:

Perhaps I should heat downstairs 24/7 on low flow temp then switch downstairs off and just in heat upstairs at a higher flow temp during the cheap rate hours. Although in I don't think I can't schedule flow temp changes on my LG controller.

 

Your mixer suggestion is better. Heat the whole lot at the minimum temperature that the highest temperature zone can get away with. Blend down where you have to.

 

 

3 hours ago, Rupe30 said:

 

couple pictures below to see inputs and outputs into buffer tank.

 

Yikes.

 

What a load of spaghetti vomited into the attic space.

 

Is that space supposed to be insulated at rafter level? It isn't. Not with those gaps in the insulation for the air to stack effect up through.

 

If the insulation between the habitable space and the attic is similarly poor then your issue might also be that upstairs simply isn't insulated well enough for the underfloor to deliver enough heat into it.

 

 

3 hours ago, Rupe30 said:

Based on running it 24/7 is it best I keep the same temp in the house all the time?

 

When it's cold out it'll need to run 24/7 at the comfort temperature

 

When it's less cold out you can get away with a modest setback at night

 

When it's mild out you can get away with running it on the time clock

 

Who designed this setup and to what spec?

 

 

3 hours ago, Rupe30 said:

I've set outdoor temp to -4 18 and LWT to 20 42 that should run 23degrees at 15 external yes? 

 

That's gibberish to me sorry. I don't speak LG. 

 

You want say 42C when -4C  and 20C when 18C? If that's how you set it in LG speak then yes. 🙂

[ReedRichards]

5 hours ago, Rupe30 said:

I've set outdoor temp to -4 18 and LWT to 20 42 that should run 23degrees at 15 external yes? 

 

So that would mean one degree down in LWT for every degree up above -4^oC.  So 15^oC is 19 degrees above -4, 42-19 = 23^oC so yes, I agree.

 

 

[Rupe30]

Thanks @markocosic really useful info.

 

The attic space pictured is really poorly insulated. The rest is done properly with  superquilt over the top, Shame they didn't just carry this insulation on into the attic but at the time it was just going to be empty attic space. The glass gable has a pretty hefty heat loss though which is a little tricky to get around.

 

I used to run the heating 24/7, tried a few weather compensation modes, but our landing and master bedroom (pictured) would lose temperature. 

 

I have recently had cavity wall insulation pumped in the landing wall and external wall insulation put on the majority of the house. I've also insulated the ceiling of the master bedroom which links to the poorly insulated attic. So I'm going to try again with the weather comp mode with the settings mentioned previously. 

 

 

I've set the thermostats to stay at one temperature so we'll see in the morning if the landing and master have been able to hold temp. If not I'll increase the weather comp temperatures and turn the mixer down on downstairs' manifold and go from there.

 

The system was designed by the company that did our thermodynamic report, which was organised by the company that installed our heatpump (who are no longer in business). It was designed to operate at 50 degrees flow temp which I now realise is a poor design. 

 

 

[markocosic]

Apart from superquilt (a gimmick AFAIK) what else is in the roof insulation wise?

[ReedRichards]

Weather Compensation makes the Leaving Water Temperature less warm when it is warmer outside.  With your settings the LWT will be 42^oC if the outside temperature is -4^oC or less.  Given that we have a cold spell at the moment Weather Compensation will set your LWT to about 42^oC.  If that is what you use without Weather Compensation then it will make no difference at all for the present.  I find it hard to see how it could ever change the balance of temperatures around your house.

 

Can you measure the flow and return temperatures across your radiators?  Given that you have a buffer tank with an unusually large volume of water I would have thought the buffer tank temperature must approach the LWT in which case I would expect that the temperature of the water entering your radiators should be close to the LWT and the water leaving your radiators should be a maximum of 5^o less than the input water.

 

If your house is of timber-framed construction then there will be a small cavity between the insulated timber frame and the exterior cladding.  This cavity is, I believe, open to the outside air.  If so, then adding exterior wall insulation will make little if any difference to the effective insulation of your house.   

Edited March 8, 2023 by ReedRichards

[markocosic]

The buffet will be cooler than LWT if heat is being drawn faster than it is replenished.

 

Timing (run 24/7 with weather comp to set heat input to dwelling; rather than running for short periods with air temperature thermostats to set heat input to dwelling) is the main change here.

 

I fear there may be no effective insulation in the roof space here if magic quilt is the only thing present.

[Rupe30]

Last night it was -1 outside, inlet outlet was showing as 35 37. 

 

The house kept the temperature apart from the landing which dropped 0.5 degrees. Downstairs heating hasn't come on much so I'll turn the mixer down on that manifold. but looks like a great base to start from. 

 

40 minutes ago, markocosic said:

Timing (run 24/7 with weather comp to set heat input to dwelling; rather than running for short periods with air temperature thermostats to set heat input to dwelling) is the main change here.

I am still using air thermostats, should I be using another way?

 

43 minutes ago, markocosic said:

 

I fear there may be no effective insulation in the roof space here if magic quilt is the only thing present.

 

It's latest building regs so there's 150mm celotex before the superquilt. I have a thermal camera and roof looks fine. There are still cold spots on the landing ceiling though so I'm hoping once I rectify that, the landing will hold temperature.

 

2 hours ago, ReedRichards said:

 

 

If your house is of timber-framed construction then there will be a small cavity between the insulated timber frame and the exterior cladding.  This cavity is, I believe, open to the outside air.  If so, then adding exterior wall insulation will make little if any difference to the effective insulation of your house.

 

It's not timber-framed. Genuinely surprised the house kept temperature last night at the lower flow temps, so the insulation has made a big difference.

 

 

2 hours ago, ReedRichards said:

Can you measure the flow and return temperatures across your radiators?  Given that you have a buffer tank with an unusually large volume of water I would have thought the buffer tank temperature must approach the LWT in which case I would expect that the temperature of the water entering your radiators should be close to the LWT and the water leaving your radiators should be a maximum of 5^o less than the input water

 

I recently measured flow and return temps on the manifolds (we don't have any radiators) and increased the flow rates so that return was 7 degrees less than input. Most of them were 10-15 degrees cooler on the return (set to manufacturers spec). I'm sure I read on another post there is an actuator that automatically adjusts to keep flow and return within 7 degrees, anyone know of the product?

 

Water entering manifolds is generally about 5 degrees cooler than LWT.

[markocosic]

3 hours ago, Rupe30 said:

Last night it was -1 outside, inlet outlet was showing as 35 37. 

 

Ok.

 

We're getting somewhere. 🙂

 

 

3 hours ago, Rupe30 said:

I am still using air thermostats, should I be using another way?

 

Yes.

 

For heat pumps you really want no thermostats. The flow temperature is set so that the heat into the building matches the heat leaving the building.

 

If you have underfloor heating you can usually just set the slab temperature ~1C above room temperature.

 

Slab at 21C room at 20C then "1C worth" of heat goes into room. Slab at 21C room at 19C then "2C worth" of heat goes into room. Slab at 21C room at 21C then "0C worth" of heat goes into the room. This will usually self-regulate pretty well without any thermostats whatsoever.

 

If you find that you need slab at 23C in the cold and at 21C when it's mild then you can use the heat pump flow temperature to achieve this.

 

 

It looks like your system doesn't need 40C+ at design condition if the air state were never calling for heat downstairs.

 

Try 35C at coldest outdoor temperature / design condition then drop to 20C at 15C. 

 

 

3 hours ago, Rupe30 said:

It's latest building regs so there's 150mm celotex before the superquilt.

 

Ah, ok!

 

If there's airtightness and 150 mm of celotex you'll be ok.

 

 

3 hours ago, Rupe30 said:

I recently measured flow and return temps on the manifolds (we don't have any radiators) and increased the flow rates so that return was 7 degrees less than input. Most of them were 10-15 degrees cooler on the return (set to manufacturers spec).

 

10-15 degC is far too high fr a heat pump setup.

 

Aim for perhaps 5C dT on your loops at design condition.

 

If you find certain areas are a touch cool and others are a touch hot then you can narrow / widen the dT on those. Narrowing the dT increases average slab temp. Widening it reduces average slab temp.

 

Actuators should be...ideally in the bin along with the air stats and the mixer pumps. You do the whole lot direct off the heat pump. Buffer can probably go in the bin or be taken out of circuit too once you've sorted your UFH setup.

 

Then you can stand the DHW cylinder upright in the remaining space to resolve that issue. 🙂

 

 

 

 

3 hours ago, Rupe30 said:

 

 

 

I'm sure I read on another post there is an actuator that automatically adjusts to keep flow and return within 7 degrees, anyone know of the product?

 

Water entering manifolds is generally about 5 degrees cooler than LWT.

 

[Rupe30]

Ok, as Im unsure about how to go about setting temperature based on slab temp, would a workaround be to set all my thermostats to 30c then use weather comp to keep the house at the right temperature. That will stop the zones turning off and on & keep the buffer tank at a constant temperature. I'd probably need to turn my mixers down to 2 on the downstairs manifold. 

 

I'm not sure there's enough height for a DHW cylinder where the buffer tank is, as much as I'd like to put the buffer in the bin! 

 

On the subject of DHW, are any horizontal cylinders any good? Or do I need to relocate to get a vertical?

[markocosic]

15 minutes ago, Rupe30 said:

Ok, as Im unsure about how to go about setting temperature based on slab temp, would a workaround be to set all my thermostats to 30c then use weather comp to keep the house at the right temperature. That will stop the zones turning off

 

That ought to work generally.

 

You would lose the "trimming" of slab temperature zone by zone though.

 

 

15 minutes ago, Rupe30 said:

I'd probably need to turn my mixers down to 2 on the downstairs manifold. 

 

Perhaps.

 

You could also have mixers set on 11 then widen the dT on the downstairs manifold and narrow it on the upstairs manifold in order to set the average temperatures.

 

 

15 minutes ago, Rupe30 said:

I'm not sure there's enough height for a DHW cylinder where the buffer tank is, as much as I'd like to put the buffer in the bin! 

 

On the subject of DHW, are any horizontal cylinders any good? Or do I need to relocate to get a vertical?

 

How not sure are you?

 

A 250 L cylinder is ~1550 mm tall. A custom fattie could also be made.

 

Horizontal cylinders don't really work. Their effective capacity is far smaller than their actual capacity because stratification within them during discharge is poor and you're more prone to get a bucket of lukewarm blended water.

[Rupe30]

22 minutes ago, markocosic said:

 

You could also have mixers set on 11 then widen the dT on the downstairs manifold and narrow it on the upstairs manifold in order to set the average temperatures

 

I was going to ask if that was a better way of doing it. 

 

22 minutes ago, markocosic said:

 

You would lose the "trimming" of slab temperature zone by zone though.

 

Enough to cause a problem or worth a go?

 

22 minutes ago, markocosic said:

 

A 250 L cylinder is ~1550 mm tall. A custom fattie could also be made.

 

Horizontal cylinders don't really work. Their effective capacity is far smaller than their actual capacity because stratification within them during discharge is poor and you're more prone to get a bucket of lukewarm blended water.

 

Lukewarm blended water is absolutely what we get. Even set to 55 degrees it doesn't feel close to that. 

 

I'll measure up tomorrow I think we do have that height. Is it worth having a smaller buffer than losing it all together?

 

[andyj007]

7 hours ago, markocosic said:

 

 

 

Actuators should be...ideally in the bin along with the air stats and the mixer pumps.

 

 

 

 

 

 

please explain this..   i understand the mixer valves  , and removed from manifolds,  but to remove individul room stats confuses me..  bedrooms you may want typically at 18-19, bathrooms at 21, office may not need any if many computers on,  how do you regulate that ? 

surelly having those rooms that require less input heat, turn off when desired temperrure reached allows other rooms to be heated with less ashp energy or running time  therfore more efficency..  what am i missing ?

 

 

[markocosic]

19 minutes ago, andyj007 said:

what am i missing ?

 

You:

 

 

- Design the loops to meet the heat loads for the rooms (bedrooms will be designed to a lower temperature than bathrooms)

 

- Adjust the (fixed) flowrate in the real world to adjust the slab temperature and therefore the temperature in each room (you can vary the average slab temperature by widening or narrowing the deltaT across the slab)

 

- Adjust the overall input according to the weather (reduce the flow temp as it gets warmer to globally reduce the slab temperatures proportionate to the heat losses

 

 

This is a "steady state" method of setting the room temperatures to work with the "steady state" preference of the heat pump; putting in the amount of heat required constantly.

 

The other option is "pulse width modulation" or "running at full burn for long enough each day to put in the amount of heat required that day; as preferred by boilers.

 

 

Heat pump efficiency will be highest if ALL the zones are as active as possible for as long as possible in order to meet the heat demands (as this results in the lowest flow temperatures

 

The office situation can be handled partly by self compensation in a high quality build (if you are setting the average slab temperature then the slab stops transferring heat to the air if the room gets too warm) and partly by averaging (if you know that computers are always dumping in heat then you dial down the input to that loop)

 

 

Crappy buildings really need thermostats. An insulated building with UFH and sensible glazing/shading generally needs stuff all.

 

 

If your office equipment is overheating the room it's probably time to buy some new office equipment. The days of this being a major power draw are as over as the days of 15p/kWh electricity.

 

A laptop will happily do 10 hours on 100 Wh; so 10W average

https://support.apple.com/kb/SP854?locale=en_GB

 

A desktop will do web browsing at about 10W and only peaks at about 50W if rendering video etc

https://support.apple.com/en-gb/HT201897

 

 

Humans are the main issue. Humans and ovens.

[andyj007]

1 minute ago, markocosic said:

 

You:

 

 

- Design the loops to meet the heat loads for the rooms (bedrooms will be designed to a lower temperature than bathrooms)

 

- Adjust the (fixed) flowrate in the real world to adjust the slab temperature and therefore the temperature in each room (you can vary the average slab temperature by widening or narrowing the deltaT across the slab)

 

- Adjust the overall input according to the weather (reduce the flow temp as it gets warmer to globally reduce the slab temperatures proportionate to the heat losses

 

 

This is a "steady state" method of setting the room temperatures to work with the "steady state" preference of the heat pump; putting in the amount of heat required constantly.

 

The other option is "pulse width modulation" or "running at full burn for long enough each day to put in the amount of heat required that day; as preferred by boilers.

 

 

Heat pump efficiency will be highest if ALL the zones are as active as possible for as long as possible in order to meet the heat demands (as this results in the lowest flow temperatures

 

The office situation can be handled partly by self compensation in a high quality build (if you are setting the average slab temperature then the slab stops transferring heat to the air if the room gets too warm) and partly by averaging (if you know that computers are always dumping in heat then you dial down the input to that loop)

 

 

Crappy buildings really need thermostats. An insulated building with UFH and sensible glazing/shading generally needs stuff all.

 

 

If your office equipment is overheating the room it's probably time to buy some new office equipment. The days of this being a major power draw are as over as the days of 15p/kWh electricity.

 

A laptop will happily do 10 hours on 100 Wh; so 10W average

https://support.apple.com/kb/SP854?locale=en_GB

 

A desktop will do web browsing at about 10W and only peaks at about 50W if rendering video etc

https://support.apple.com/en-gb/HT201897

 

 

Humans are the main issue. Humans and ovens.

 sounds good thats unlikely  happen in the real world, the chance of having a system designed as you have described by the leading underfllor people,   and then set up by a so called experianced plumber,   and then tweaked and adjusted by average joe publicsadly  is close to zero..   even if it were greater than zero, throw in some heat recovery , log burners and the entire sytem becomes unbalanced ....        be inetersted in your own set up and running costs to see whats possible of a well set up system..  do you have a thread somewhere   im never to old to learn..