Clean Heat Market Mechanism to incentivise heat pumps

[JohnMo]

40 minutes ago, Roger440 said:

Im on house 4. None have had WC. All have worked fine.  Its just another level of complexity to go wrong. Ive seen enough threads on here to know this has set up problems written all over i

Think you miss the point of WC, there is zero complexity you have a controller (either come with the boiler or cheap enough as an option) and within it a simple outside temp - flow temp curve. That's it, no thermostats, or radiator valves - a fully open system. You calculate the curve based on house heat losses, you balance the system, job done. People make a WC system complex by adding thermostats and make lots of justification of why their system needs it.

 

Biggest issue is installer training or lack of, S and Y plan rules the roost, because its easy for the installer.

 

 

[sharpener]

1 hour ago, Roger440 said:

Ive got a 5 year old oil boiler.

 

Missed that on first reading, thought you were referring to gas. You should be able to get WC for that now. I have a 25 y/o oil boiler which has an economiser in the flue, but won't run in true condensing mode because low temps would cause the flue to rot.

 

So my alternative to WC was a wireless TRV setup. Now that has been very problematic and I wouldn't recommend Honeywell Evohome.

 

The old oil boiler is now reaching end-of-support, and a new one will require an expensive new lined flue through the roof, so given the grant available an ASHP is a sensible choice. But as with many retrofits it has not been a straightforward project.

 

[Roger440]

17 minutes ago, JohnMo said:

Think you miss the point of WC, there is zero complexity you have a controller (either come with the boiler or cheap enough as an option) and within it a simple outside temp - flow temp curve. That's it, no thermostats, or radiator valves - a fully open system. You calculate the curve based on house heat losses, you balance the system, job done. People make a WC system complex by adding thermostats and make lots of justification of why their system needs it.

 

Biggest issue is installer training or lack of, S and Y plan rules the roost, because its easy for the installer.

 

 

 

I understand its purpose.

 

Saying there is zero complexity is a stretch. By your own admission you need to calculate a curve based on your house heat losses. 

 

Id suggest its now got quite complex. Who knows what the heat losses are? I dont. The installer definitely wont.

 

Given the homeowner certainly isnt going to do it, it falls to the installer. With best will in world, its unlikely to be right first time out. So repeat visits will be necessary to get it refined.

 

Not going to happen for your average budget installation.

 

[Roger440]

5 minutes ago, sharpener said:

 

Missed that on first reading, thought you were referring to gas. You should be able to get WC for that now. I have a 25 y/o oil boiler which has an economiser in the flue, but won't run in true condensing mode because low temps would cause the flue to rot.

 

So my alternative to WC was a wireless TRV setup. Now that has been very problematic and I wouldn't recommend Honeywell Evohome.

 

The old oil boiler is now reaching end-of-support, and a new one will require an expensive new lined flue through the roof, so given the grant available an ASHP is a sensible choice. But as with many retrofits it has not been a straightforward project.

 

 

Yes, im sure i can. But why would i want to? Ive twiddled with flow temps in so far as i can.

 

All im going to do is give myself a problem installing it and setting it up, as per post above. Or pay someone to do it. Which in itself, here, would be a major struggle to get anyone on site, never mind someone that could actually do it succesfully. 

 

Never mind the added complexity and therefore reliability that it will bring. Not saying it will fail, but has potential to. If its not there, it definitely cant fail. 

 

 

[Beelbeebub]

8 hours ago, Roger440 said:

 

I understand its purpose.

 

Saying there is zero complexity is a stretch. By your own admission you need to calculate a curve based on your house heat losses. 

 

Id suggest its now got quite complex. Who knows what the heat losses are? I dont. The installer definitely wont.

 

Given the homeowner certainly isnt going to do it, it falls to the installer. With best will in world, its unlikely to be right first time out. So repeat visits will be necessary to get it refined.

 

Not going to happen for your average budget installation.

 

This hits the nail on the head.

 

The installer needs to be a lot sharper with WC (and by extension a HP) and do more work than simply fit the unit and wind the flow temp right up.

 

HP manufacturers need to make the controls much simpler and more intuitive. Nobody needs to be fiddling with curves. A dial with desired temp should be be all they have to interact with. The box should learn the best temps to achive them in the most efficient way. 

 

We've got these fancy "AIs" that can compose music, paint pictures and write articles. Maybe one that could set up a heatpump efficiently isn't too much to ask! 🤖

 

 

[SteamyTea]

13 minutes ago, Beelbeebub said:

 

We've got these fancy "AIs" that can compose music, paint pictures and write articles. Maybe one that could set up a heatpump efficiently isn't too much to ask

I asked https://zzzcode.ai/

 

import numpy as np
import matplotlib.pyplot as plt

# Generate random weather data
np.random.seed(0)
temperature = np.random.normal(10, 5, 365)

# Generate heating system output based on weather data
heating_output = temperature * 0.8 + 20

# Plot the weather compensation curve
plt.plot(temperature, heating_output)
plt.xlabel('Outside Temperature (°C)')
plt.ylabel('Heating Output (%)')
plt.title('Weather Compensation Curve')
plt.grid(True)
plt.show()
 

 

[sharpener]

2 hours ago, SteamyTea said:

heating_output = temperature * 0.8 + 20

 

wrong sign I think, try

 

if (0AT<20) then heating_output = 20 + 0.8*(20-OAT)

else = 0

 

which simplifies to 36 - 0.8*OAT for OAT < 20

 

Edit:

Or in your specific case @SteamyTea (now I have read your other post) 46.8 - 0.8*OAT <g>

 

 

Edited December 18, 2023 by sharpener

[JamesPa]

13 hours ago, Roger440 said:

 

I understand its purpose.

 

Saying there is zero complexity is a stretch. By your own admission you need to calculate a curve based on your house heat losses. 

 

Id suggest its now got quite complex. Who knows what the heat losses are? I dont. The installer definitely wont.

 

Given the homeowner certainly isnt going to do it, it falls to the installer. With best will in world, its unlikely to be right first time out. So repeat visits will be necessary to get it refined.

 

Not going to happen for your average budget installation.

 

Weather compensation has been mandatory in Germany for many years, which is why many newer gas boilers have it.  Presumably German plumbers have found a way either to set it up in one go, return efficiently to tweak it, perhaps on a rota, or teach German house owners how to tweak it.  Are we saying that our plumbers are incapable of doing what the Germans do?  Presumably also Polish plumbers, who doubtless work just across the border from time to time now they have fled the UK, have also worked out how to do it.  Alternatively we can just sacrifice 10-25% in performance.  

 

Given that replacing our gas boilers with heat pumps is now an unavoidable necessity, we are either going to have to sort it out or suffer a performance penalty.

 

As it happens I am firmly of the opinion that, at least for homes with smart meters, a bit of AI (or even human intelligence) applied to the half hourly meter readings will both enable us to size heat pumps correctly (unlike the spreadsheet method as it appears to be applied in many cases) and create a weather compensation curve.  I find it impossible to believe that Octopus aren't working on this already.

 

Edited December 18, 2023 by JamesPa

[JohnMo]

Looking at gas boiler WC curves they start at 20 degs with an outside temp of 20 degs. So it should be really as simple as take note of outside temp set flow temperature at a reasonable guess, say 35 degs, how much hotter or colder than target temp is the house after 24 hrs, set appropriate curve by reading the supplied chart. All Atag boilers come with WC, they have a simple selection chart, low temp radiator or high radiators or UFH, well insulated etc. and points you towards a starting curve. You can then adjust the whole curve up or down on the controller like you would a thermostat.

 

Should all be pretty basic stuff for an installer.

[Beelbeebub]

7 hours ago, SteamyTea said:

I asked https://zzzcode.ai/

 

import numpy as np
import matplotlib.pyplot as plt

# Generate random weather data
np.random.seed(0)
temperature = np.random.normal(10, 5, 365)

# Generate heating system output based on weather data
heating_output = temperature * 0.8 + 20

# Plot the weather compensation curve
plt.plot(temperature, heating_output)
plt.xlabel('Outside Temperature (°C)')
plt.ylabel('Heating Output (%)')
plt.title('Weather Compensation Curve')
plt.grid(True)
plt.show()
 

 

Ha!  I was thinking more along the lines of putting the measured outside air temp, flow temps and inside temps into the system and it working out the best mapping for flow temp given outside temp, desired inside temp and actual inside temp.

 

So starts off pretty aggressively with high flow temps.

 

Watches how quick the internal temp rises.

 

Backs off abit, until it's controlling the internal temp at the lowest possible flow temps.

 

Given desired internal temp shouldn't change much or at least changes in a predictable manner (eg night setback), the system should be able to control to the desired temp profile at minimum flow temp.

 

So for the first few days the system will run at an inefficient high flow. Then back off progressively to the eventual best guess.  The extra cost of a few days of inefficient running are negligible.

 

The system would also be able to self adjust should something change eg new windows, more loft insulation, or an extension.

 

Such a system would be more a hybrid WC and Load compensation system. Not dissimilar to the TPI strategy but using flow temp rather than on/off

 

Should be possible 

[JamesPa]

54 minutes ago, Beelbeebub said:

Ha!  I was thinking more along the lines of putting the measured outside air temp, flow temps and inside temps into the system and it working out the best mapping for flow temp given outside temp, desired inside temp and actual inside temp.

 

So starts off pretty aggressively with high flow temps.

 

Watches how quick the internal temp rises.

 

Backs off abit, until it's controlling the internal temp at the lowest possible flow temps.

 

Given desired internal temp shouldn't change much or at least changes in a predictable manner (eg night setback), the system should be able to control to the desired temp profile at minimum flow temp.

 

So for the first few days the system will run at an inefficient high flow. Then back off progressively to the eventual best guess.  The extra cost of a few days of inefficient running are negligible.

 

The system would also be able to self adjust should something change eg new windows, more loft insulation, or an extension.

 

Such a system would be more a hybrid WC and Load compensation system. Not dissimilar to the TPI strategy but using flow temp rather than on/off

 

Should be possible 

Isn't that pretty much what Homely and the 'adaptive' control strategies which are can optionally be enabled on some heat pumps do.  We don't know exactly what algorithm they apply, but as I understand them the idea is to take a basic/default  weather compensation curve and slowly but surely tweak it to suit the house.

 

Of course this still depends somewhat on a tolerable balancing of the emitters (and not fiddling with them at least for a while) but the industry has (supposedly) been doing that for decades (or has it?). 

 

Either way its not that difficult and, as I observe upthread, the Germans have had WC for years, and we surely aren't claiming to be less capable than they are. 

 

 

Edited December 18, 2023 by JamesPa

[JohnMo]

A decent gas boiler comes with WC and load compensation and both can be selected to work together, great for radiators. Put controller in the hall or a place you want as a reference temp, lob in an approx curve, set target temp and you are done. Zero Reason this approach should not work on a heat pump. Controller just trims back or adds temperature to the output flow to match temperature required.

[Iceverge]

Back to the title of the thread. 

 

How about making electricity bills simpler by doing away with standing charges. Its a regressive charge anyway. 

 

Offer a free hour of electricity per day. People would soon figure how to maximise it. 

[SteamyTea]

42 minutes ago, Iceverge said:

Offer a free hour of electricity per day

At how many amps?

[Iceverge]

19 minutes ago, SteamyTea said:

At how many amps?

 

Lol. 

 

I wondered if someone was going to ask. 

As many as you can pull without burning out the fuse. 

 

Obviously you'd need to stagger the free hours. 

 

 

[Roger440]

5 hours ago, JamesPa said:

 

Either way its not that difficult and, as I observe upthread, the Germans have had WC for years, and we surely aren't claiming to be less capable than they are. 

 

 

 

Id suggest we are as a collective of installers. Of course there are a lot of very good capable insallers, but overall, no. 

 

The german approch to things is just different. Mostly in a good way.

[Roger440]

13 hours ago, SteamyTea said:

I asked https://zzzcode.ai/

 

import numpy as np
import matplotlib.pyplot as plt

# Generate random weather data
np.random.seed(0)
temperature = np.random.normal(10, 5, 365)

# Generate heating system output based on weather data
heating_output = temperature * 0.8 + 20

# Plot the weather compensation curve
plt.plot(temperature, heating_output)
plt.xlabel('Outside Temperature (°C)')
plt.ylabel('Heating Output (%)')
plt.title('Weather Compensation Curve')
plt.grid(True)
plt.show()
 

 

 

I knew it would get complicated. What does all that mean? 

[Beelbeebub]

1 hour ago, Iceverge said:

Back to the title of the thread. 

 

How about making electricity bills simpler by doing away with standing charges. Its a regressive charge anyway. 

 

Offer a free hour of electricity per day. People would soon figure how to maximise it. 

but with no standing charge that would essentially be free electricity.

 

IIRC there used to be tariffs with no or low standing charge (maybe it was regulated) but the first 5kwh or something were really expensive, effectively making a standing charge

 

On thing (away from heat pumps) I would do is make the prepay tariff the cheapest per unit with a very minimal standing charge.

 

 

[JohnMo]

@Roger440

 

This is how my gas boiler does (did) weather compensation, all pre programmed for any house

 

https://www.atagheating.co.uk/knowledge/external-controls-weather-compensation

 

No reason, why other manufacturers cannot do the same, including heat pumps

[JamesPa]

4 minutes ago, Roger440 said:

5 hours ago, JamesPa said:

 

Either way its not that difficult and, as I observe upthread, the Germans have had WC for years, and we surely aren't claiming to be less capable than they are. 

 

 

 

Id suggest we are as a collective of installers. Of course there are a lot of very good capable insallers, but overall, no. 

 

The german approch to things is just different. Mostly in a good way.

Oh dear, how sad.

 

Well we have to swap gas boilers for heat pumps, so we will just have to work it out or else the uk consumer will be paying 15-25pc more for heating than they need to.  Perhaps a 'collective of installers' could visit Germany!  

 

 

[Roger440]

8 minutes ago, JamesPa said:

Oh dear, how sad.

 

Well we have to swap gas boilers for heat pumps, so we will just have to work it out or else the uk consumer will be paying 15-25pc more for heating than they need to.  Perhaps a 'collective of installers' could visit Germany!  

 

 

 

It is sad, yes.

 

However, i dont think we need to worry too much, as we discussed on here before, to get this done by 2030, we need to do 70k installs a week. Which is about what we manage in a year. So the reality is, its going to be a very long time. So some will lean.

 

But ultimately, yes you are correct, many will be paying much more than they should. Indeed, plenty of people with heat pumps already are. I know a couple. To say they are not very happy would be an understatement.

 

 

[Roger440]

14 minutes ago, JohnMo said:

@Roger440

 

This is how my gas boiler does (did) weather compensation, all pre programmed for any house

 

https://www.atagheating.co.uk/knowledge/external-controls-weather-compensation

 

No reason, why other manufacturers cannot do the same, including heat pumps

 

Im unable to establish on the internet if my trusty worcester boiler is WC compatible. The gas ones are, but cant find a definitve answer on the oil variant.

 

I may need to go and look inside it. That can wait as its dark and wet right now!

 

Whilst you say its pre programmed, thats not quite true. You do need to set it up as per the video. Though clearly that simpler than calculating a house heat loss. But i guess most installers will go pessimistic to avoid comeback. For some reason on this thread, there doesnt seem to be much recognition that this is critical to the installer. Its got to be done, right, first time out. Nobody is going to pay for repeat visits, including the installer.

 

As suggested further up, quite why this cant be an entirely automatic function is beyond me.  So long as you have all the temp inputs, why is it so damn hard? Take it out of the hands of the installer or homeowner and let it sort itself out.

 

[Beelbeebub]

37 minutes ago, Roger440 said:

 

Im unable to establish on the internet if my trusty worcester boiler is WC compatible. The gas ones are, but cant find a definitve answer on the oil variant.

 

I may need to go and look inside it. That can wait as its dark and wet right now!

 

Whilst you say its pre programmed, thats not quite true. You do need to set it up as per the video. Though clearly that simpler than calculating a house heat loss. But i guess most installers will go pessimistic to avoid comeback. For some reason on this thread, there doesnt seem to be much recognition that this is critical to the installer. Its got to be done, right, first time out. Nobody is going to pay for repeat visits, including the installer.

 

As suggested further up, quite why this cant be an entirely automatic function is beyond me.  So long as you have all the temp inputs, why is it so damn hard? Take it out of the hands of the installer or homeowner and let it sort itself out.

 

Can anyone say why Weather Comp gets all the plaudits and Load comp seems to be ignored?

 

pure WC adjusts flow temp by using the external temp only.  It doesn't actually know what is going on inside.  Things like solar gain or open windows are beyond it.  If you have huge windows and it's a sunny day it will happily keep pumping heat into the house as you all cook.  Likewise, if your windows are open, it won't do a damn thing about it.  The saving grace is if you have a low temp emitter system (cough UFH) there is a built in regulation of the heat input due to the dT between emitter and internal temp being pretty small. 

 

My floor slab is about 23C for a room temp of 20C at the moment.  As soon as the sun comes out and the temp climbs to 21C my heat input (dT) has dropped by 1/3.  Likewise if my room dropped to 19C, my heat output will jump by 1/3.  So as long as I can keep my slab at a stable temp I'm pretty good.

 

But I don't think it would work if I had rads running at 40C,  there the difference in dT would be much smaller until i drifted uncomfortably high or low in temp.

 

It also (as has been noted) means the installer has to get the curve dialed in well.

 

With Load Comp, the HP/boiler adjusts the flow temp based on internal temp.  It doesn't really care about external temp, just staying at the same internal temp.  As internal temp drifts down, flow temp raises, and vice versa.  it would seem to be inherently self tuning. All you need to tell it is an acceptable time to get up to temp (5C per hour or 0.5C per hour?) and then it gets on with it. 

 

 

You can imagine a thermostat with a temp dial and maybe a boost button that  will kick the heating response up to a higher level if you want.  Installer sets up some basic parameters, which may normally be factory set but may be varied for edge cases, then leaves the homeowner to it.

 

 

[JamesPa]

1 hour ago, Roger440 said:

As suggested further up, quite why this cant be an entirely automatic function is beyond me.  So long as you have all the temp inputs, why is it so damn hard? Take it out of the hands of the installer or homeowner and let it sort itself out.

It's a reasonable question.  I think that getting the feedback loop to work stably, given the long response time of a house and the disturbing variables, is probably quite a challenge, but nothing a bit of electronics couldn't sort out over a period of a few weeks or certainly a season, as I understand Homely does.  I suspect also that too many heat pumps are directly descended from aircon at present with not enough market specific development, although given that its exactly the same problem as for a fossil fuel boiler, you might have thought the manufacturers who also manufacture boilers would have it sorted. 

 

Perhaps German plumbers (and plumbers in other countries where its mandatory on boilers) are adept at tuning wc so there isn't much of a market.  Its probably only since about the year 2000 that the technology to automate it became sufficiently cheap that it might be considered, and if by then plumbers had worked out how to tune it manually (which is very likely) the manufacturers would have little incentive to undertake a quite expensive development until it is adopted by backward countries like the UK.  Perhaps they still reason that, if the Germans can do it, then so can anyone else.  

 

 

 

Edited December 18, 2023 by JamesPa

[JamesPa]

47 minutes ago, Beelbeebub said:

With Load Comp, the HP/boiler adjusts the flow temp based on internal temp.  It doesn't really care about external temp, just staying at the same internal temp.  As internal temp drifts down, flow temp raises, and vice versa.  it would seem to be inherently self tuning. All you need to tell it is an acceptable time to get up to temp (5C per hour or 0.5C per hour?) and then it gets on with it. 

 

I fear that pure load comp wouldn't work very well because the response is too delayed.  Clearly one of the big problems with any space heating control is the delayed reaction of the house to both external stimuli (eg OAT changing) and internal stimuli (eg turning the heating up).  I'm not a control theory expert, but I do know that any delay in a control loop, particularly if you have disturbing and unpredictable features like solar gain, makes the design more challenging.  WC attempts to get ahead of that a bit by measuring the external temperature and adjusting the flow temperature, possibly before the house itself reacts, to counter the energy loss from the fabric which is bound to happen. 

 

I suspect that's why the more advanced heat pumps use a combination of weather compensation (to get a basic setting) and load compensation (to 'tweak' it a bit).

Edited December 18, 2023 by JamesPa