Some basic HP questions for those in the know

[Gordo]

hi i'm designing my HP system in preparation for the "professionals" design. 

 

Heatpunk design comes out at 5kw heat pump @ 45 degrees for an annual heat load of just over 11,000 kwh for space heating. My historical heating works out at 12,500 kwh, with a pellet boiler and includes DHW and some supplementary heating with a small log stove. Therefore im pretty content with validity of the heatpunk design output. its a reasonably well insulated 2003 house about 180m2 floor area

 

what im intrested to know is how the heat pump modulates its output.

 

I assume the flow rate is fixed?

 

do heat pumps modulate the flow temperature up / down or is that generally set (fixed) at the controller. then it just modulates the input electric to maintain the set flow temperature until it can go no lower and shuts down for a period (cycling)

 

unassuming the flow temperature is set (fixed) at controller is it a matter of playing with controller to come on the minimum flow temperature needed.

 

how does weather compensation come into the equation. will this then adjust the flow temperature automatically.

 

probably silly questions for those in the know. lol

 

Thanks in anticipation

Edited November 20 by Gordo

[JohnMo]

7 hours ago, Gordo said:

assume the flow rate is fixed?

Some will modulate flow don't.

 

Basics are

Target flow temp, dT, max and min flow temp (hysterisis).

 

So compressor starts, at min flow temperature, once started tries to establish dT at minimal flow temp, as dT decreases, more flow temp is added, up to target flow temp. Once there different manufacturer do different things, but it's all about power input, to maintain dT and flow temp.

 

WC is just a variable flow target temp based on outside temp.

[ReedRichards]

9 hours ago, Gordo said:

I assume the flow rate is fixed?

 

1 hour ago, JohnMo said:

Some will modulate flow don't.

I beg your pardon, @JohnMo?  I think he means "not necessarily".

9 hours ago, Gordo said:

do heat pumps modulate the flow temperature up / down or is that generally set (fixed) at the controller. then it just modulates the input electric to maintain the set flow temperature until it can go no lower and shuts down for a period (cycling)

Heat pumps have a target flow temperature.  But it may take them a while to reach that target.  They will indeed shut down if the target is exceeded (by some pre-set margin) at their minimum electrical input.

9 hours ago, Gordo said:

unassuming the flow temperature is set (fixed) at controller is it a matter of playing with controller to come on the minimum flow temperature needed.

You would try to establish the the minimum flow temperature needed for your desired interior temperature at each particular outside temperature.

9 hours ago, Gordo said:

how does weather compensation come into the equation. will this then adjust the flow temperature automatically

Once you have done that, you use weather compensation to render it automatic so you can stop playing around.   

[JohnMo]

47 minutes ago, ReedRichards said:

beg your pardon, @JohnMo?  I think he means "not necessarily".

Just reread

Meant to write 

2 hours ago, JohnMo said:

Some will modulate flow, some don't.

And as way of example - mine does modulate the circulation pump and I cannot change that. Others will have external pump(s) or some like Grant have an internal pump that runs at a fixed speed, the speed is adjustable, but remains fixed after that.

[sharpener]

For example by default Vaillant HPs modulate the compressor, fan and pump to maintain a dT of 5C. This requires a flow rate of 170 l/hr per kW of output. It is possible to define the building circulation rate separately for both HW and CH but I don't know of anyone who has done this or why they would.

 

I recommend reading the detailed analysis by Michael Podesta of how their weather compensation works. It is complicated, and spread over this and the three succeeding articles. Not helped by some errors in the Vaillant manuals in particular the basic graph illustrating how WC works(!).

[Gordo]

thanks guys that advice was helpful.

 

so most HP have a fixed flow rate.

HP all vary flow temperature based on how far off from the set target flow temperature

weather compensation will auto adjust target flow temperature based on inside/outside temperatures based on some selected algorithm (graph)?

 

think i got it.

[ReedRichards]

6 hours ago, Gordo said:

weather compensation will auto adjust target flow temperature based on inside/outside temperatures based on some selected algorithm (graph)?

 

It's usually a straight line; nothing sophisticated.

[JohnMo]

7 hours ago, Gordo said:

 

so most HP have a fixed flow rate.

I wouldn't go that far, some do have a fixed flow rate, some don't.

 

7 hours ago, Gordo said:

HP all vary flow temperature based on how far off from the set target flow temperature

Don't confuse house temp and heat pump target temp. The heat pump has zero care about what happens in the house, it's just a heat generator. It looks after it's self. The target temp referred too is heat pump target flow temperature, the heat pump is always trying to get to that temperature, while also managing dT. It manages dT first to protect itself from damage of the refrigerator circuits.

 

7 hours ago, Gordo said:

weather compensation will auto adjust target flow temperature based on inside/outside temperatures

WC just takes reference to outside temperature only. Doesn't care about inside temperature, just simply follows a straight line curve based on outside temp. You adjust the curve so it coincides with a house temp you like.

 

You have variations of the above, like load compensation, which does take references to house temp, these work well with radiators but not with UFH.

[JamesPa]

44 minutes ago, JohnMo said:

You have variations of the above, like load compensation, which does take references to house temp

Just to add to what @johnmo says, these variations typically work in addition to weather compensation not instead of.  Some more middle-market-upwards heat pumps feature a mode called something like 'adaptive' (or similar terminology), which automatically 'tweaks' the basic WC curve based on what the heat pump learns about the house over a period of time.  This is not the same as responding directly to room temperature (like a thermostat would), it had a much longer time constant for the response.  Various third party controllers, of which the best known is Homely, do something similar (and other smarts in addition) for heat pumps that don't feature this kind of supplementary control.

 

Note though

44 minutes ago, JohnMo said:

these work well with radiators but not with UFH

 

 

Edited November 21 by JamesPa

[Gordo]

6 hours ago, JamesPa said:

Just to add to what @johnmo says, these variations typically work in addition to weather compensation not instead of.  Some more middle-market-upwards heat pumps feature a mode called something like 'adaptive' (or similar terminology), which automatically 'tweaks' the basic WC curve based on what the heat pump learns about the house over a period of time.  This is not the same as responding directly to room temperature (like a thermostat would), it had a much longer time constant for the response.  Various third party controllers, of which the best known is Homely, do something similar (and other smarts in addition) for heat pumps that don't feature this kind of supplementary control.

 

Note though

 

 

Oh that’s new to me “load compensation”. I’m familiar with “homely”. So some HP don’t take internal temperature into account just outside with weather compensation. How do we find out which HPs don’t look at internal temperature? Look at controller data sheet maybe. 
 

do you think that “load compensation” is something I should want, or can WC generally take care of it when set up properly

[PhilT]

1 hour ago, Gordo said:

do you think that “load compensation” is something I should want, or can WC generally take care of it when set up properly

Yes to both, it depends on how much you enjoy the tech involvement. I paid extra for a smart thermostat so I can set to auto adapt, which incorporates load compensation, to take care of everything for me in the most efficient way possible, but I have rads, it's an unknown quantity with UFH. Others prefer managing the tech to achieve a "Goldilocks" WC setting. Either system can be made to work well.

Edited November 21 by PhilT

[Gordo]

35 minutes ago, PhilT said:

Yes to both, it depends on how much you enjoy the tech involvement. I paid extra for a smart thermostat so I can set to auto adapt, which incorporates load compensation, to take care of everything for me in the most efficient way possible, but I have rads, it's an unknown quantity with UFH. Others prefer managing the tech to achieve a "Goldilocks" WC setting. Either system can be made to work well.

I’m afraid I’m a bit of an obsessive tinkerer. Slightly concerned that a HP install could end up as another obsession that I spend disproportionate amounts of time playing with 😂 but I do have a small wood stove that I adore. So don’t really want to give it up completely. Want the HP to provide a background heat to say 19 or 20 degrees and walk the stove on on colder evenings. Thinking internal temperature sensor for HP would be advantageous?

Edited November 21 by Gordo

[JamesPa]

4 minutes ago, Gordo said:

I’m afraid I’m a bit of an obsessive tinkerer. Slightly concerned that a HP install could end up as another obsession that I spend disproportionate amounts of time playing with 😂 

I think that may be a given!

 

I'm obsessed already and have been for some time, and mine doesn't arrive until next Monday.  Whoever knew heating could be so all-consuming.  It's not just the ashp, that's just a trigger to try to understand the thermal dynamics of houses

[JamesPa]

Just now, JamesPa said:

I think that may be a given!

 

I'm obsessed already and have been for some time, and mine doesn't arrive until next Monday.  Whoever knew heating could be so all-consuming. 

 

It's not just the ashp, that's just a trigger to try to understand the thermal dynamics of houses, which appears to be stunningly poorly understood even though the underlying physics is extremely simple and extremely well understood.

 

[Gordo]

1 minute ago, JamesPa said:

I think that may be a given!

 

I'm obsessed already and have been for some time, and mine doesn't arrive until next Monday.  Whoever knew heating could be so all-consuming.  It's not just the ashp, that's just a trigger to try to understand the thermal dynamics of houses

I agree it’s the science / understanding that underlies it becoming obsessive trying to fine tune something to be as efficient as possible

[Gordo]

5 minutes ago, JamesPa said:

 

I work in the technical side of building. I can confirm that there is understanding out there. Few apply their knowledge in their specialty as it as it takes time and time being money which few are willing to pay for.

Edited November 21 by Gordo

[JamesPa]

30 minutes ago, Gordo said:

I work in the technical side of building. I can confirm that there is understanding out there. Few apply their knowledge in their specialty as it as it takes time and time being money which few are willing to pay for.

That's good to know thanks.

 

Obviously it's not cost effective to theorise/model each house individually, but is there any prospect that the behaviour of several broad classes of house could be described in a way that moves us forward for the purposes of understanding our heating? 

 

Also is there any prospect that those who could model their individual house (which I suspect includes almost any electrical engineer who has used some flavour of SPICE - which I would have guessed is a good modelling tool for the purpose) could get some insight from those who do understand houses on how to do so?

 

[Gordo]

14 minutes ago, JamesPa said:

That's good to know thanks.

 

Obviously it's not cost effective to theorise/model each house individually, but is there any prospect that the behaviour of several broad classes of house could be described in a way that moves us forward for the purposes of understanding our heating? 

 

Also is there any prospect that those who could model their individual house (which I suspect includes almost any electrical engineer who has used some flavour of SPICE - which I would have guessed is a good modelling tool for the purpose) could get some insight from those who do understand houses on how to do so?

 

Modern houses and their heat losses could be better predicted but the house building system in the uk with speculative developers offering minimum supervision over sub-contractors working to a bottom line and minimal statutory oversight doesn’t bode well for the actual houses performing to design assumptions.Developers have only a passing intrest in the ongoing running costs to the owner of the dwelling. I would like the industry to provide thermal imaging or such to buyers along with a AI report based on images on the predicted heat losses.

 

the as built SAP calculations which output the EPC certificate should be a good indicator of heat loss, but for the above reasons and inaccurate as built input data are often way off.

Edited November 21 by Gordo

[Roger440]

11 minutes ago, Gordo said:

Modern houses and their heat losses could be better predicted but the house building system in the uk with speculative developers offering minimum supervision over sub-contractors working to a bottom line and minimal statutory oversight doesn’t bode well for the actual houses performing to design assumptions.Developers have only a passing intrest in the ongoing running costs to the owner of the dwelling. I would like the industry to provide thermal imaging or such to buyers along with a AI report based on images on the predicted heat losses.

 

All very nice, but in reality, wishful thinking.

 

Developers are interested in profit, not the running costs. There is simply no incentive to be interested.

[JamesPa]

1 hour ago, Gordo said:

the as built SAP calculations which output the EPC certificate should be a good indicator of heat loss, but for the above reasons and inaccurate as built input data are often way off

Thanks.

 

I was thinking beyond steady state heat loss, which could, more or less, be measured.  I was thinking about the dynamic effects (eg what happens when the conditions, eg oat or iat, change).  I was also thinking about retrofit of older houses with composite constructions, not just new ones.  

 

Is any of this understood, even roughly?

 

This is a serious question.  I vividly remember a friend of mine embarking on a PhD, sponsored by the met office, to study the air flow over one isolated conical hill, followed up by a study of the air flow over two conical hills.  That was 40 years ago and, crude though it is, I suspect it's contributed to the fact that weather forecasting has advanced dramatically since then, albeit that it's still not perfect.  Weather forecasting is also a situation where the physics is well understood but the dynamics are complex (albeit much more complex than domestic heating).

 

Are we at the 'conical hill' stage in understanding thermal dynamics in houses, or are we further forward?  This isn't a trick question or a value judgement, just a genuine curiosity about how deep (or shallow)or understanding of houses is.

 

Edited November 21 by JamesPa

[Gordo]

19 minutes ago, JamesPa said:

Thanks.

 

I was thinking beyond steady state heat loss, which could, more or less, be measured.  I was thinking about the dynamic effects (eg what happens when the conditions, eg oat or iat, change).  I was also thinking about retrofit of older houses with composite constructions, not just new ones.  

 

Is any of this understood, even roughly?

 

This is a serious question.  I vividly remember a friend of mine embarking on a PhD, sponsored by the met office, to study the air flow over one isolated conical hill, followed up by a study of the air flow over two conical hills.  That was 40 years ago and, crude though it is, I suspect it's contributed to the fact that weather forecasting has advanced dramatically since then, albeit that it's still not perfect.

 

Are we at the 'conical hill' stage in understanding thermal dynamics in houses, or are we further forward?  This isn't a trick question or a value judgement, just a genuine curiosity about how deep (or shallow)or understanding of houses is.

 

Ohhh the general industry's understanding of thermal dynamics in a dwelling is virtually zero. beyond some professor or such. Thermal comfort is I believe much more complex than just degrees. I do know that humidity and air movement are significant factors which are usually ignored. Also heat loss calculations really only look at heat conduction again radiation and convection are generally ignored. for example the muli layered foil insulation products cant get any traction in the market because the cant prove the real world performance claims based on current testing procedures. Only conductivity performance figures are used based on a BS "hot box" test rig.

[Gordo]

1 hour ago, Roger440 said:

 

All very nice, but in reality, wishful thinking.

 

Developers are interested in profit, not the running costs. There is simply no incentive to be interested.

That is the problem. Statutory bodies only spot check work, developers and contractors and architects all profit (survival) driven. Time is money so lets cut that down to the bone. the only way to get some quality is to self build, arm yourself with knowledge (google) and supervise everything personally. Not a task to be taken on lightly and with a monumental learning curve with no dry runs. Obviously all this quality will come at a proportionate cost or time penalty.

[sharpener]

1 hour ago, JamesPa said:

I suspect includes almost any electrical engineer who has used some flavour of SPICE

 

22 minutes ago, JamesPa said:

Are we at the 'conical hill' stage in understanding thermal dynamics in houses, or are we further forward? 

 

I would imagine that the behaviour in the time domain could reasonably be modelled by RC delay circuits. I can't think what the analogue for inductance might be so probably not any second order differential equations.

 

But still room for oscillatory behaviour depending what is in the control systems. I don't think you would have much joy trying to prise that out of the mfrs. Presumably they have some theoretical models of house behavious to refine the control systems against.

 

2 hours ago, JamesPa said:

mine doesn't arrive until next Monday

 

You must be all agog @JamesPa. Look forward to reading the write-up, are you going to start a thread on this? PM me on the side if I can help.

 

9 minutes ago, Gordo said:

muli layered foil insulation products cant get any traction in the market

 

Was new to me when loft conversion ppl suggested it for my bathroom ceiling. No idea if it works, a better suggestion would have been to line the window wall with insulated plasterboard, there was just an inch to play with which would have made a significant difference.

 

Main objection is the blasted stuff is impossible to fix through bc it binds the drill bit.

[JamesPa]

1 hour ago, Gordo said:

Ohhh the general industry's understanding of thermal dynamics in a dwelling is virtually zero. beyond some professor or such.

Thanks.  Is anyone doing any work on it.  It's surely much less complex than weather forecasting but arguably at least as important.  Isn't that what BRE is for?

[JamesPa]

49 minutes ago, sharpener said:

I would imagine that the behaviour in the time domain could reasonably be modelled by RC delay circuits. I can't think what the analogue for inductance might be so probably not any second order differential equations.

That's my guess, but it would be good for the industry to confirm or not.  Maybe wind effects eg are such that a multiple RC model only is sufficient.