Combining an air-water heat pump and solar thermal

[Garald]

I am about to start renovating my new place (which is most of an old rowhouse, attic included). Climate zone Cfb, latitude 49° N.

I will almost certainly change the current heating source (an old gas-based water heater) for an air-to-water heat pump. Since the roof will probably be all available for my use (well, will need to talk about it the coop, but let us hope for the best), it seems logical to install solar. Obviously the energy from photovoltaic panels can be used for the heat pump (and other things), but wouldn't it be logical and efficient to use solar panels simply to heat water, which would then be heated further (at least in winter) by the heat pump? (That would be either in addition or instead of photovoltaic panels.)

I imagine this sort of hybrid system must either 
- be very common or 
- not make any sense for some reason I am not seeing.

It would be nice to get references in the first case, and a reason in the second one!

[JohnMo]

You can pick up solar thermal for a good price if you look around, as it's gone out of favour with PV getting cheaper in recent years.  Another option is PV, two ways to do that, normal grid connected with a diverter to immersion not that good for hot water except in summer.  The other is to use PV for direct water heating.  You can use DC from the panel to an MPPT tracker then to the immersion. Or direct to immersion, but have to careful matching everything up.

Edited May 11, 2022 by JohnMo
Grammar

[SteamyTea]

I would not bother with solar thermal, it needs maintenance and is a one trick pony.  Once the water is up to temp, it then just sits there waiting to go wrong.

PV is much more reliable.  It is easier to install as well.  Wire are easier to route than pipes.

[Garald]

Thanks - I will need to learn about the terms that JohnMo used (immersion, etc.), as I am newbie - what would be a good resource?

 

Here is what I am wondering (as someone with a background in pure science and no practical experience in this area). PV to heat involves two steps: solar power becomes electricity, which is then used to heat water. Solar Thermal involves just one step: solar power heats water. It seems completely counterintuitive (though not impossible) that the first process should be more efficient: it involves two steps.

 

What is the efficiency of the two processes nowadays?

 

(Also, when is the water ever up to temp? I may be eco-conscious, but I like hot showers!)

 

 

[SteamyTea]

26 minutes ago, Garald said:

Here is what I am wondering (as someone with a background in pure science and no practical experience in this area). PV to heat involves two steps: solar power becomes electricity, which is then used to heat water. Solar Thermal involves just one step: solar power heats water. It seems completely counterintuitive (though not impossible) that the first process should be more efficient: it involves two steps.

It is entropy.

Thermal energy, in the context of liquid water heating, is quite disordered, so higher entropy.

Electrical energy is much more ordered, so lower entropy.

Lower entropy to higher entropy sit in the better (more efficient) corner of the Carnot Cycle.

 

Efficiency calculations become a bit tricky when looking at renewable energy. If you calculate the area of land needed to supply just enough energy to heat a known quantity of water, a fixed number of kelvin, then ST looks impressive.

But if you look at time a ST system sits doing nothing, maybe 22 hours a day in the summer, and compare that that what a PV system can do i.e. heat the water in say 3 hours, then help run the rest of the house for another 10 hours, you get a much better utility return.

 

PV is now cheaper, simpler and more reliable. Roof integrated PV is a similar price to roof slating.

[Garald]

I should add that:

 

a) the plan is to keep the old-style radiators in the heating system (because cost), not to install low-temperature heaters.

b) we are talking about the Paris area, not a tropical desert.

Hence my scratching my head as to how the temperature could ever get up to temp, at least during winter.

[Garald]

4 minutes ago, SteamyTea said:

It is entropy.

Thermal energy, in the context of liquid water heating, is quite disordered, so higher entropy.

Electrical energy is much more ordered, so lower entropy.

Lower entropy to higher entropy sit in the better (more efficient) corner of the Carnot Cycle.

 

Efficiency calculations become a bit tricky when looking at renewable energy. If you calculate the area of land needed to supply just enough energy to heat a known quantity of water, a fixed number of kelvin, then ST looks impressive.

But if you look at time a ST system sits doing nothing, maybe 22 hours a day in the summer, and compare that that what a PV system can do i.e. heat the water in say 3 hours, then help run the rest of the house for another 10 hours, you get a much better utility return.

 

PV is now cheaper, simpler and more reliable. Roof integrated PV is a similar price to roof slating.

 

Thanks. But what about the ecological costs of production? I was having a chat with a friend who was trying to convince me to install solar panels (but he was preaching to the converted). He was talking mainly about PV, and he mentioned that their weak point was that, once you took into account production footprint, they were only about as green as nuclear (the main conventional source of electricity in the area). Surely the production footprint of thermal panels must be low, as they are essentially just big, black flat tanks? Or am I being naive?

[SteamyTea]

1 minute ago, Garald said:

He was talking mainly about PV, and he mentioned that their weak point was that, once you took into account production footprint,

That is the French for you.  They have been using this line for at least a decade.  Same line that petrolheads use about BEVs.

It is bollocks and I don't even enter it that conversation anymore with them.  Show they all the life cycle analysis and they will still not believe it.

2 minutes ago, Garald said:

essentially just big, black flat tanks

Not may people use them, most are evacuated tubes.

 

Go to PVGIS and play about.

 

Regarding 

16 minutes ago, Garald said:

the plan is to keep the old-style radiators in the heating system (because cost), not to install low-temperature heaters.

You could be setting yourself up to fail, then blame the ASHP.  Calculate it all correctly first, don't take a guess, or worse, an opinion.

[Garald]

How *do* I calculate whether the existing radiators (used with the gas heater I will scrap) will work well with the heat pump?

[SteamyTea]

56 minutes ago, Garald said:

How *do* I calculate whether the existing radiators (used with the gas heater I will scrap) will work well with the heat pump?

You can usually find manufacturers charts to show the power output at different temperatures. 

Or just halve it as a rule of thumb.

What you really need to do is a room by room heat loss calculation. Then you know the correct sizes needed.

 

What are if science did you study?

[Garald]

a) I have no idea of who the radiator manufacturer was. 

b) Primitive solution: if only it were still cold outside, I suppose I could just run the radiators at full blast and see whether they overheat the place (that would be a good thing, since it would indicate a lower temperature might be sufficient). I suspect they will; the place was well-heated during the first visit in a cold day in January, and insulation was and is terrible, so, once I insulate the place well...

 

(Of course sales in France take forever. My architect is hoping for the last cold wave of spring so that she can check for thermal bridges.)

 

I'm in pure maths (not strictly speaking a science). Did a double degree in maths and computer science (which *really* is not a science) before doing my postgraduate studies in maths. I would like to believe I am not completely illiterate in physics, but I'm not quite sure that's true, and at any rate it depends on the subarea. I *think* I am a quick study in physics (well, you'd expect it it from someone in math, particularly analysis and allied areas) but that may just be Dunning-Kruger.

[Garald]

PS. Knowing how to program is good for something in this context. Here's a simulation I did last weekend of sun exposure in my new library (the largest and most important place in the house):

 

https://webusers.imj-prg.fr/~harald.helfgott/simulcurie/21_6_2021.html   (summer solstice)

https://webusers.imj-prg.fr/~harald.helfgott/simulcurie/21_8_2021.html   

https://webusers.imj-prg.fr/~harald.helfgott/simulcurie/21_12_2021.html   

[Garald]

I was just talking a friend who has more experience than I of the practical side of things (well, same friend as above). He essentially said that he thinks of solar thermal as something for large families. Makes sense: who else needs lots of hot water during the summer?

[SteamyTea]

10 hours ago, Garald said:

I'm in pure maths (not strictly speaking a science)

Good stuff, just wait till @Adsibob reads this.

 

10 hours ago, Garald said:

I am a quick study in physics

Only GCSE stuff, just a case of learning the SI and SI derived units.

At least you will be able to work out thermal inertia and hopefully explain it better than I have.

[SteamyTea]

10 hours ago, Garald said:

Here's a simulation I did last weekend of sun exposure in my new library

I like the clock.

Did you take the reflectivity and absorption of the glass into account, or just treat the window as an opening?

 

You may find you are in great demand for thermal modelling.

Edited May 11, 2022 by SteamyTea

[Nickfromwales]

14 hours ago, Garald said:

Thanks - I will need to learn about the terms that JohnMo used (immersion, etc.), as I am newbie - what would be a good resource?

Get a glass of warm milk and have a bedtime read  

[Nickfromwales]

13 hours ago, Garald said:

How *do* I calculate whether the existing radiators (used with the gas heater I will scrap) will work well with the heat pump?

A very rough check would be to run them at 40oC flow temp from the gas setup. If the house won't get up to temp, you're going to need bigger rads. You may also need larger bore pipework to get a higher volume of the 'cooler' water circulating around the system.

[Garald]

2 hours ago, SteamyTea said:

I like the clock.

Did you take the reflectivity and absorption of the glass into account, or just treat the window as an opening?

 

You may find you are in great demand for thermal modelling.

 

Oh, I just treated a window as an opening. I'm just showing where direct light falls, not how much of it does! There has to be a tiny error due to refraction, but all mesures were taken by the architect to the nearest 5cm anyhow (the real estate agent was in a hurry), so I was not going to bother. We will probably make an opening at a particular place in the long wall (an opening to be hidden behind a cabinet door, that is) to let direct light into the bedroom on winter mornings.

 

For detailed modelling of light (with serious work on reflection), there are various front-ends using Radiance (https://www.radiance-online.org/). I found all the ones I tried to be kludgy, and the back-end itself a little archaic - there went an evening. I couldn't see whether there was any way to make it do what I wanted it to do (showing where direct light fell). I don't know what sort of professional software there is for thermal modelling.

 

In the end I found it easier to write my own program in Python/SageMath (using the Pysolar library to compute altitude and azimuth as a function of time, and the Three.js library to get the animation running). I put the clock on the wall because the Sagemath-Three.js interface is somewhat incomplete - it wouldn't let me put the time on the corner of the picture, as something extradiegetic, so to speak.

[JohnMo]

Not sure if you are improving the insulation of the house.  If you are the existing radiators would require less output, so the reduced temperature output of the heat pump, may work.

 

Solar thermal, most people will heat domestic hot water, hence the throw away term, one trick pony.  But there is no reason it cannot contribute to heating in the shoulder months also.  In fact it's yield at lower heating temperature should be better, than high temp DHW.  A drain back system for instance is very simple, it is not pressurised, it has a small vessel (10L) a small pump and a simple differential temperature controller and 2 or 3 temperature probes.

 

Oversize the system for winter performance, to displace some winter heating, most shoulder months heating and all DHW for 60% of the year.  When the system has done it job it drains the thermal panels, leaving no water in the panel to overheat or possibly freeze.

 

Your light simulations, watch for the low sun in September/October and March/April which can lead to room over heating.

[SteamyTea]

7 hours ago, Garald said:

I don't know what sort of professional software there is for thermal modelling.

I have played around with LISA and SolidWorks, but on here, it is usually a spreadsheet.

 

The basics of thermal modelling are pretty simple, just areas, material conductivities and temperature gradients, plus ventilation.  Where it starts to get complicated is where water vapour starts to condense, especially true when drying out a new building, or renovating an old one.

The other thing people are interested in, is energy storage within the fabric of the building.  This opens up a can of worms as some believe that adding more mass is better than lightweight structures where just the air is heated (or cooled).  The waters get muddied when energy prices and CO_2e are included.

Solar Gain, which in most of the UK at least, has not been much of an issue, is starting to be thought about because of better insulated and airtight buildings, and Architects love of large windows.  The usual mitigation method is to add blinds, reflective films or air conditioning.

 

I usually find it easier to do a statistical model and then decide if there is a serious problem, or a very small problem.  I would hardly want to pay out a few thousand to solve a problem that may only last a week, especially if it only happens every few years (had our very first extreme heat warning, where I am, 50°N, 5°W, ever last year, got to 26°C).  Hardly a real problem.

Edited May 12, 2022 by SteamyTea

[Gone West]

On 11/05/2022 at 00:39, Garald said:

I imagine this sort of hybrid system must either 
- be very common or 
- not make any sense for some reason I am not seeing.

It would be nice to get references in the first case, and a reason in the second one!

We moved into our house a year ago which has a 1.4kW ST panel and 4kW PV system. The ST system seems to work well. The immersion didn't work and the PV system was only being used as background supply so this year I installed a Solar iBoost which now supplies the immersion but a lot of the PV energy still goes to the grid. The ST currently heats the DHW to 60C and the immersion heats it to 75C so the heating oil isn't being used to heat DHW. I will sort out the DHW system this year but I would have thought the ST and PV would provide hot water for eight or nine months of the year.

[SteamyTea]

30 minutes ago, Gone West said:

We moved into our house a year ago which has a 1.4kW ST panel and 4kW PV system.

Would you be interested in putting energy meters on both the ST and the DHW side of the PV.

Chart the readings against your weather station.

 

(the ST thermal meter need not be an expensive one, just something that logs the times it is running and the flow and return temperatures, a RPi job)

[Gone West]

2 hours ago, SteamyTea said:

Would you be interested in putting energy meters on both the ST and the DHW side of the PV.

Not this year, too busy, but possibly next year. Haven't even got the weather station data logger working yet.

[Garald]

13 hours ago, JohnMo said:

 

 

Your light simulations, watch for the low sun in September/October and March/April which can lead to room over heating.

 

I wouldn't be worried about March or October - most people already have their heating on by then - or April,  since then only the lowest part of the long wall is exposed. (The left wall is occupied in part by a mantelpiece.) In September, the issue is mainly light falling obliquely between 7:30 and 9:30am, from the right window to the left part of the long wall; heat transfer should be less because of the sharp angle (together with the low altitude), no?

 

Closely related if somewhat off-topic: I am planning on installing floor-to-ceiling built-in bookcases. When I was a few days younger, and even more naive, I went around asking how should I go about choosing the material for the bookcases (and so forth) so that they would absorb a bit of heat during winter mornings (which is when they would be fully exposed to direct light head-on - low-altitude, but still) and release it during the day. I was told that I was being silly, in that it is a better idea to let the sun heat water rather than wood. It didn't seem like an either-or to me (do you make bookcases out of water?), but that obviously pointed me in the direction of solar thermal. 

 

Now I'm back to where I started. It's hard to get consistent information on the heat capacity of wood (in part because it depends on several different factors), but apparently it is descent - if I understand correctly, the volumetric heat capacity of oak is somewhere between half that of concrete (http://www.international-agrophysics.org/Thermal-properties-of-wood-and-wood-composites-made-from-wood-waste,142472,0,2.html) or about that of concrete, and the same is true of dense pine (the heat capacity per gram is about the same as that of oak). Of course direct light coming through a window on a winter night just doesn't carry that much power, but even if someone sitting or lying down in the reading nook were to gain 1° C, that would be something. Haven't done any computations that would tell me whether that's the right order of magnitude.

 

At any rate, maybe there's a point to this: bookcases of dense, dark-ish wood, with, say, cork behind the back-panels to provide a bit of insulation. But then perhaps the effect would be negligible, and I might as well choose light-wood bookshelves to get more light in the rest of the room. Can people eyeball this? As I was saying, an order-of-magnitude feeling for things should be enough to tell what makes sense.

Edited May 12, 2022 by Garald

[SteamyTea]

17 minutes ago, Garald said:

Now I'm back to where I started. It's hard to get consistent information on the heat capacity of wood (in part because it depends on several different factors),

How about this one.  Not read it myself, just a quick skim.  Does talk about thermal mass though.

https://www.irjet.net/archives/V8/i2/IRJET-V8I2307.pdf

 

I think the problem with using timber shelves is that the grain needs to be pointing towards the heat source (sun).

Books should be similar to cellulose insulation, a bit denser, with a higher volumetric heat capacity, but greater thermal conductivity as they have less air between the pages.

I find that books, when stored badly, go mouldy.  My first dissertation rotted away in my first house in 18 months.  Shame as I was quite proud of it at the time.

17 minutes ago, Garald said:

Of course direct light coming through a window on a winter night just doesn't carry that much power,

I once modelled this and then created an experiment.  I used 15 W.m^-2 and 25 W.m^-2.

There was very little rise in temperature at those low levels.  Newton's ghost was hanging around I think, then Fourier came and gave me a kicking.

Edited May 12, 2022 by SteamyTea