ASHP developments

[Iceverge]

1 hour ago, Marvin said:

This is the whole point that lots of people don't understand. Low (low water temperature) and slow (slowly warming up the building) the way to go(most efficient and therefore least expensive way to heat the home)! 

 

It's best method of getting a good COP from an ASHP for sure. It will deliver each kWh of heat to your house at the lowest price. 

 

However as a method of least overall cost use I don't think it applies to a house that has high losses and is intermittently occupied. It'll need to run far too long when nobody's in to get to a comfortable temperature when everyone's home from work. Much of this heat will be lost to outside with no benefit.

 

Unless you can convince your electricity supplier to drag a cable to your house that can power a 40kW heat pump I can't see ASHP's being a solution for most working people without tacking the house fabric first.

 

 

 

 

 

 

[billt]

2 hours ago, Iceverge said:

A nice video that. Quite accessible. 

 

Did you consider oversizing even further and lowering the flow temp to benefit from a better COP?

 

Because it's not practicable. Even UFH systems don't operate at much lower temperatures and the radiator sizes become very unrealistic.

 

Using weather compensation it won't give much of a performance increase anyway. Today it's running at a leaving temperature of 30 and a return of 26 so a mean of 28C. The Cop of for the last 30 minutes was 5.35.

 

CoP is a bit like insulation thickness, the law of diminishing returns applies. Going from 2 to 3 is a 50% improvement, 3 to 4 is a 30% improvement and 4 to 5 a 25% improvement. Chasing big numbers is amusing but largely pointless.

 

The radiators are barely perceptibly warm; Graham Hendra has a blog describing how he recommends setting heat pump systems up so the temperature is at least 37C because many customers were complaining that the system wasn't working as the radiators feel cold!

[Iceverge]

4 minutes ago, billt said:

CoP is a bit like insulation thickness, the law of diminishing returns applies. Going from 2 to 3 is a 50% improvement, 3 to 4 is a 30% improvement and 4 to 5 a 25% improvement. Chasing big numbers is amusing but largely pointless.

 

The radiators are barely perceptibly warm; Graham Hendra has a blog describing how he recommends setting heat pump systems up so the temperature is at least 37C because many customers were complaining that the system wasn't working as the radiators feel cold!

 

Agreed, think I wrote something very similar last week somewhere on here. Once you get to a COP of about 3-4 the maths really start to taper off. 

 

If your house is continuously heated I suppose you could get away with smaller radiators than an intermittently heated one.  Perhaps making ASHP's more suitable for retired people and work from home occupiers. 

 

My parent's rads often run at very low temperature dumping heat from the solid fuel cooker. It's made a big difference to the comfort of the house, even though they'll be cool to the touch. 

 

I think ASHPs, like almost every other appliance, baring the kettle and toaster, are too complex to ensure efficient operation by uninterested users.  

 

 

 

 

 

 

[Marvin]

3 hours ago, Iceverge said:

I think ASHPs, like almost every other appliance, baring the kettle and toaster, are too complex to ensure efficient operation by uninterested users.  

I think it's also what people are accustom to and expecting. We now have an expectation of instant results in life!

Light on. Kettle on. phone on. computer on. Heating what?!

[Roger440]

5 hours ago, Iceverge said:

 

I think ASHPs, like almost every other appliance, baring the kettle and toaster, are too complex to ensure efficient operation by uninterested users.  

 

 

 

This ^^^^^

 

The things people on this forum do to manage and optimize their ASHP are far far beyond what any normal people will do. Indeed, well beyond what a lot of installers will do.

 

If, when people turn on the heating, heat doesnt occur, then as far as they are concerned its not working.

 

Iceverges point about intermittently occupied houses is a very good point as well. If you have a pooly insulated house, and you only turn on your heating for a couple of hours in the morning, and again in the evening, then i could well see why a ASHP, even well specced and installed will either cost way more to get house comfortable during hours of occupation, or, operated like the gas system, then it will be rather cold as the heat wont arrive when its required.

[Marvin]

35 minutes ago, Roger440 said:

 

This ^^^^^

 

The things people on this forum do to manage and optimize their ASHP are far far beyond what any normal people will do. Indeed, well beyond what a lot of installers will do.

 

If, when people turn on the heating, heat doesnt occur, then as far as they are concerned its not working.

 

Iceverges point about intermittently occupied houses is a very good point as well. If you have a pooly insulated house, and you only turn on your heating for a couple of hours in the morning, and again in the evening, then i could well see why a ASHP, even well specced and installed will either cost way more to get house comfortable during hours of occupation, or, operated like the gas system, then it will be rather cold as the heat wont arrive when its required.

One has to wonder how countries like Norway cope with so many ASHP's!

 

It has to be because of their positive approach for over 50 years of introduction and their well insulated and air tight houses!

[HughF]

15 minutes ago, Marvin said:

One has to wonder how countries like Norway cope with so many ASHP's!

 

It has to be because of their positive approach for over 50 years of introduction and their well insulated and air tight houses!

It’s because they use air-air on retrofit and underfloor exclusively on new build.

[HughF]

56 minutes ago, Roger440 said:

 

This ^^^^^

 

The things people on this forum do to manage and optimize their ASHP are far far beyond what any normal people will do. Indeed, well beyond what a lot of installers will do.

 

If, when people turn on the heating, heat doesnt occur, then as far as they are concerned its not working.

 

Iceverges point about intermittently occupied houses is a very good point as well. If you have a pooly insulated house, and you only turn on your heating for a couple of hours in the morning, and again in the evening, then i could well see why a ASHP, even well specced and installed will either cost way more to get house comfortable during hours of occupation, or, operated like the gas system, then it will be rather cold as the heat wont arrive when its required.

Air-air has to be the answer to both these issues. Push button, heat comes out.

[billt]

1 hour ago, HughF said:

Air-air has to be the answer to both these issues. Push button, heat comes out.

 

That is rather dependent on the house design. Here we have 12 heated spaces. Doing that with A2A would not be simple. A2A is arguably less comfortable than radiator provided heat and they definitely make noise which may or may not be irritating.

 

I've installed an A2A unit for one room which has 3 and a bit external walls and no insulation, but it would not be sensible to use A2A in the rest of the house.

[HughF]

7 minutes ago, billt said:

 

That is rather dependent on the house design. Here we have 12 heated spaces. Doing that with A2A would not be simple. A2A is arguably less comfortable than radiator provided heat and they definitely make noise which may or may not be irritating.

 

I've installed an A2A unit for one room which has 3 and a bit external walls and no insulation, but it would not be sensible to use A2A in the rest of the house.

So, what’s the alternative? R290 heat pumps that support load compensation and high flow temps, along with rads?

[HughF]

1 minute ago, HughF said:

So, what’s the alternative? R290 heat pumps that support load compensation and high flow temps, along with rads?

Actually, the more I think about it, he more I think that this is the way forwards for mass adoption. R290, and proper load compensating stats. Turn the stat up, rads gets hotter… people understand that.

[PhilT]

Just now, HughF said:

Actually, the more I think about it, he more I think that this is the way forwards for mass adoption. R290, and proper load compensating stats. Turn the stat up, rads gets hotter… people understand that.

totally agree - that's what Octopus are leading with, and my bet would be that the rest of the energy suppliers and hp makers will follow

[HughF]

I genuinely don’t know why these  heat pump manufacturers don’t support opentherm for control of flow temp…. Then you could easily hook up a nest or tado and let that do the flow temperature calculation/adjustment.

[JamesPa]

16 minutes ago, PhilT said:

totally agree - that's what Octopus are leading with, and my bet would be that the rest of the energy suppliers and hp makers will follow

 

20 minutes ago, HughF said:

So, what’s the alternative? R290 heat pumps that support load compensation and high flow temps, along with rads?

 

+1 (mostly). 

 

My personal view is that FT=55 (mandated now for new buildings even if they have gas boilers, and likely  enough in many retrofits - eg mine - because rads are already oversized due to fabric improvements) will turn out to be the practical sweet spot for radiators.  The ability to go to 65/70 will be used for DHW, likely with legionella cycle done by the Heat pump for simplicity - Vaillant already do this. 

 

A bit of adjustment of the relative prices of gas/electricity (which are politically driven anyway) will make this no less affordable than gas, and using a higher flow temp for DHW will eliminate the need, in many cases, to replace cylinders.* 

 

I don't however accept that weather compensation will go, I think it will stay but simplified and augmented by load compensation, so that it doesn't have to be too accurately to set up - and is thus easy.  Several heat pump manufacturers already have this and of course this is exactly what Homely does.

 

I did actually say more or less exactly what @HughF says about a year ago on this forum, but was shouted down by many at the time.  

 

 

*It may be necessary to retrofit a plate heat exchanger to improve recovery time, but that's cheap by comparison with total replacement.  There is still the problem of persuading installers not to insist on cylinder replacements - because that's where they hide the mark up, but hopefully the market will eventually fix that.

 

Edited January 3 by JamesPa

[HughF]

1 minute ago, JamesPa said:

 

 

+1 (mostly). 

 

My personal view is that FT=55 (mandated now for new buildings even if they have gas boilers) will turn out to be the sweet spot for radiators, with the ability to go to 65/70 used for DHW, probably with legionella cycle done by the Heat pump (Vaillant already do this). 

 

A bit of adjustment of the relative prices of gas/electricity (which are politically driven anyway) will make this no less affordable than gas, and using a higher flow temp for DHW will eliminate the need, in many cases, to replace cylinders.* 

 

I don't however accept that weather compensation will go, I think it will stay but augmented by load compensation so that its easy to set up.  Several heat pump manufacturers already have this and of course this is exactly what Homely does.

 

I did actually say more or less exactly what @HughF says about a year ago on this forum, but was shouted down by many at the time.  

 

 

It may be necessary to retrofit a plate heat exchanger to improve recovery time, but thats cheap by comparison with total replacement.  There is still the problem of persuading installers not to insist on cylinder replacements - because that's where they hide the mark up, but hopefully the market will eventually fix that.

 

I agree completely, why has it taken homely to figure out that automatic weather comp setting, based on user feedback (turn up thermostat to make rads hotter, that’s how most people think) is necessary. Why do the heat pump manufacturers think they can get away without room temperature feedback built in? 

[JamesPa]

3 minutes ago, HughF said:

Why do the heat pump manufacturers think they can get away without room temperature feedback built in? 

I have absolutely no clue.  Whilst WC is pretty good, its unlikely, for many houses, to be good enough on its own and even if it is in principle good enough, to set it up sufficiently precisely is too much of an ask for the average person.

Edited January 3 by JamesPa

[PhilT]

2 minutes ago, HughF said:

I agree completely, why has it taken homely to figure out that automatic weather comp setting, based on user feedback (turn up thermostat to make rads hotter, that’s how most people think) is necessary. Why do the heat pump manufacturers think they can get away without room temperature feedback built in? 

Mitsubishi does auto adaptation very well indeed, but most effectively with rads, ufh is an unknown quantity. I stopped using WC months ago, full auto being so much better, the numbers suggest around 15% more efficient than WC mode.

[JamesPa]

3 minutes ago, PhilT said:

Mitsubishi does auto adaptation very well indeed, but most effectively with rads, ufh is an unknown quantity. I stopped using WC months ago, full auto being so much better, the numbers suggest around 15% more efficient than WC mode.

Interesting.  Do we know how Mitsubishi Auto Adaption actually works.  I can think of at least three options

 

Use the WC curve as a starting point and 'tweak' but without long term learning

Use the WC curve as a starting point and 'tweak' but with long term learning, so that it might eventually depart from the WC curve entirely

Use nothing as a starting point and 'tweak' but with long term learning, ie work it all out from scratch

 

I am pretty sure it must use a feedback loop with at least 2 time constants, a short term one and a long term one.

Edited January 3 by JamesPa

[HughF]

It’ll be interesting to see how octopus end up doing it with their new heat pump.

[Roger440]

2 hours ago, HughF said:

Air-air has to be the answer to both these issues. Push button, heat comes out.

 

But helpfully excluded from any of the schemes.

 

[Roger440]

3 minutes ago, HughF said:

It’ll be interesting to see how octopus end up doing it with their new heat pump.

 

Will be interesting to see if reality matchres the hype?

 

[PhilT]

46 minutes ago, JamesPa said:

Interesting.  Do we know how Mitsubishi Auto Adaption actually works.  I can think of at least three options

 

Use the WC curve as a starting point and 'tweak' but without long term learning

Use the WC curve as a starting point and 'tweak' but with long term learning, so that it might eventually depart from the WC curve entirely

Use nothing as a starting point and 'tweak' but with long term learning, ie work it all out from scratch

 

I am pretty sure it must use a feedback loop with at least 2 time constants, a short term one and a long term one.

It uses WC as a start point then learns the curve of roomstat temp vs heat output. On Melcloud you can clearly see the flow heat curve flattening out more and more as it approaches target room temp, over the course of a few days, until it achieves a consistent control pattern. The end result is steady long runs, typically in the high 30s in these mild temps, with rads. It also manages much gentler starts than WC, I.e. no power spikes, and automatically compensates for defrost cycles by upping the flow temp slightly post defrost.

Edited January 3 by PhilT

[JamesPa]

25 minutes ago, PhilT said:

It uses WC as a start point then learns the curve of roomstat temp vs heat output. On Melcloud you can clearly see the flow heat curve flattening out more and more as it approaches target room temp, over the course of a few days, until it achieves a consistent control pattern. The end result is steady long runs, typically in the high 30s in these mild temps, with rads. It also manages much gentler starts than WC, I.e. no power spikes, and automatically compensates for defrost cycles by upping the flow temp slightly post defrost.

Interesting.

 

You would have thought, given this, that it would suffice to tell the heat pump

 

a) the design flow temp

b) the design OAT and

c) whether it was dealing with rads or UFH (so it knows to use dt^1 (UFH) or dt^1.3 (rads))

 

all of which the installer should know. 

 

These three pieces of information are sufficient for the heat pump itself (or an app provided by the manufacturer or a third party) to work out a basic WC curve to start from, and then Auto Adaption takes over from there.  Surely its not unreasonable to expect a plumber to program that basic information in, which they must have.

[HughF]

2 hours ago, JamesPa said:

Interesting.

 

You would have thought, given this, that it would suffice to tell the heat pump

 

a) the design flow temp

b) the design OAT and

c) whether it was dealing with rads or UFH (so it knows to use dt^1 (UFH) or dt^1.3 (rads))

 

all of which the installer should know. 

 

These three pieces of information are sufficient for the heat pump itself (or an app provided by the manufacturer or a third party) to work out a basic WC curve to start from, and then Auto Adaption takes over from there.  Surely its not unreasonable to expect a plumber to program that basic information in, which they must have.

Hence the reason why Graham Hendra came up with the Lizzie curve. 37@15, 50@0.

 

It’s a good starting point, but ideally the room stat would adapt those numbers if the room ended up too hot or too cold. Sadly, that’s left down to the installer to come back and adjust… and we know how well that works out.

[Beelbeebub]

55 minutes ago, HughF said:

Hence the reason why Graham Hendra came up with the Lizzie curve. 37@15, 50@0.

 

It’s a good starting point, but ideally the room stat would adapt those numbers if the room ended up too hot or too cold. Sadly, that’s left down to the installer to come back and adjust… and we know how well that works out.

That's the key bit, the installer having to come back and adjust. 

 

With gas boilers the installer can wind it up to max, and then the thermostat on/offs to control the temp.

 

Which installer love because they don't have spend an hour to drive back to your house to twist a dial on the boiler. 

 

With HPs that leads to poor cop and low output which gives high running costs and not enough heat output. 

 

Ideally, the HP woiod be set up perfectly from day one, but realistically it will require a little tuning to get right. 

 

Users are used to a thermostat so we need a similarly simple interface. 

 

Maybe a simple dial with no temps just "warmer/colder" and the controller working things out from there. The controllers goal being to flow a temp sufficient that the user doesn't feel the need to fiddle the dial.  If it consistently being turned up then in should adjust for a higher flow temp for those internal and external temps and vice versa 

 

Edited January 3 by Beelbeebub