Fixing mistakes of the past, Thermal Store and Fireplace with water jacket
Location: South Poland, climate zone 6a
Hello all,
4 years ago we renovated a wooden \ brick house and did our best to make the house airtight (not tested) and as insulated as possible. I had a company in the UK perform a Sustainability Review and Thermal Modelling that you are welcome to read in full attached, but the key takeaway for this question is the heating requirements that were suggested…
* Wood Stove on the ground floor for room heating and hot water ~10kW
…connected to a Thermal Store.
For the last 4 years, we have had a 8kW (6.4kW water \ 1.6kW room) fireplace as our main heat source and topped up as needed with some electric heaters, mainly in the bathroom when washing the kids. A few notes on this experience…
Positives:
1. This has mostly worked well with heating needed for approx 5 months of the year Oct to March.
2. 6kW Immersion heater added to the Thermal Store that uses spare Solar energy to heat hot water allowed us to have hot water in summer and benefit from sunny days in winter
3. I have sensors added to the UFH, Thermal Store, and Fireplace that allow me to monitor and automate the heating systems and UFH with great flexibility (with Home Assistant)
Negatives:
A. With the store at around 50% 75c (167f) we have enough hot water to heat the upstairs, downstairs our concrete slab soaks up all this heat in around an hour, therefor we mostly have a cool floor downstairs (and lots of hassle from my partner)
B. No mixer was added for Direct Hot Water so the water depending on the Store temp can come out of a tap at around 70c+ which is very dangerous
C. I noticed overnight we lose around 10c from the store and I believe this is due to thermosyphon happening around the UFH heating feed and return pipe, I can see this at the UFH Manifold temps
D. Feed for UFH is taken from the thermal store top (75c) and then mixed down at the manifolds (40c) so a lot of unneeded heat is stored in the UFH feed pipe
E. When we have the Immersion heater active we get a thermosyphon happening in the fireplace feed and return pipes in reverse (hot water goes return > feed) that heats up the fireplace water jacket and activates the pump. We have to manually close a valve on this loop to stop this.
F. The Minimum return water temperature for the fireplace is 55c, but as this is taken from the bottom of the store it is normally around 26-31c
Attached are the schematics of the system to give an idea of the setup.
I would like to resolve the negatives above, I do not have high confidence in the installer and would like to have a well-informed understanding of how to improve these before I talk to anyone or may undertake some of the work myself, my suggestions are:
Improvements
Resolve A: As per the design documents add a 10kW Ground Source Heat pump, this should bring the bottom half of the Thermal Store up to around 45c, this is enough for the UFH in the winter.
Resolve B: Add a mixer for DHW at the store to bring tap water down to 50c
Resolve C: Add a solenoid valve at the store on the UFH feed pipe that is closed when not in use.
Resolve D: Add a mixer for UFH feed at the store to bring water down to 45c. I still want to take the water from the top and not for example the middle of the store so I can tap into all the hot water in the store, sometimes it is better to heat the house than have hot water for washing.
Resolve E: Add a solenoid valve at the store on the fireplace return pipe that is closed when not in use.
Resolve F: Move the fireplace feed pipe from the bottom to the middle of the store where the temperature will be around 45-50c when the Immersion heater is not in use.
Other points:
1. With the suggestions above I will have the Immersion heater and Fireplace managing the top of the store, both will not be in use at the same time and the Heat Pump will manage the bottom of the store for months when heating is needed.
2. For F I could also have a mixer valve added to take any available heat from the top of the store and mix it with the feed to the fireplace to help get closer to 55c. But keen to not add extra complexity
3. I still have a spare coil at the top of the store for another heating system if needed I can add later
You can see the proposed schematics with these suggestions also attached.
Im keen to get feedback on my suggestions and would love to hear from others who have similar setups their experience. Sorry for so much content but I’m keen to provide as much info as possible for those interested 🙂
Thank you in advance!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
You can't have a solenoid between the expansion tank and the Thermastore. The best location for the solenoid would be on the supply line to the stove after the pump. If the two floor heat pumps have check valves in them (or can be installed), you should not need the 2nd solenoid.
As shown in your current diagram, all house heat can only come from the stove or the resistance element, none of the heat from the heat pump would be used.
To fix it, you need to move the cold water feed of the mixing valve to the middle of the tank bellow the resistance heater. This way if the water at the middle location is hot enough, it will only use that for heat and only use the top of the tank if that runs out.
Air to water heat pumps are readily available in UK, I would not bother with GSHP. If you get split unit (not monoblock) you can connect it straight to the Thermastore instead of the coil which reduces delta T thus operating costs. Make sure to get a vapor injection cold climate unit.
I really hope you have insulation under the main floor slab. I know lot of slab on grade there is missing it. If you don't have any, you will still have issues, probably even using up the 10kW heat pump output, as you are mostly heating the soil under the house.
Hi Akos, thanks for your reply.
"hope you have insulation under the main floor slab" - for sure, 20cm EPS for the main slab (60% floor area) and 10cm for the rest.
"can't have a solenoid between the expansion tank and the Thermastore" - this makes sense, thank you.
"best location for the solenoid would be on the supply line to the stove after the pump" - interestingly in this location I do have a check valve, but this does not stop the flow of heat, I am assuming the movement of heat is not strong enough to "trigger" the check valve. Before the pump is the current manual valve we use to close this loop so makes sense why this was added here.
"As shown in your current diagram, all house heat can only come from the stove or the resistance element, none of the heat from the heat pump would be used." my assumption here is that the heat pump heated water would go to the top of the store, so for example, if the fireplace is not used and we have no sun, the heat pump would heat all water above it. But, I can see this logic may be misplaced and the heated water from the heat pump would stay in the bottom area of the thermal store where the heat pump coil is located?
"To fix it, you need to move the cold water feed of the mixing valve to the middle of the tank below the resistance heater" I may have misunderstood how the mixing valve works here, if I attach the cold feed to the mixing valve to the middle of the store would it not only use that if the hot feed is too hot, so it will use the top of the store as the primary source of heat still?
"Air to water heat pumps are readily available in UK, I would not bother with GSHP" same here, its the most popular option for installation and most installers will only do ASHP. Im very tempted to go with ASHP but also keen on GSHP to take advantage of running the system in reverse and having cooling in the summer from the pump.
" I do have a check valve, but this does not stop the flow of heat"
I've seen this happen as well. Usually spring check valves better than a swing check but can still leak when debris builds up on the seat. Solenoid is a guaranteed off.
" heat pump heated water would go to the top of the store"
Heat pumps usually produce colder water ~45C. With your stratified tank, if the top is heated by the element or stove to say 55C, no heat from the heat pump will make it there.
You need to pull the supply water for the floor heat from a location in your tank that is bellow the output temperature setpoint of the heat pump.
"f I attach the cold feed to the mixing valve to the middle of the store would it not only use that if the hot feed is too hot"
Say the middle is at 40C, top at 70C and you want 45C water. It would use about 85% 40C water and 15% 70C water. So most of the heat would be sourced from the heat pump.
If you set the mixing valve to 40C, than 100% would come from the heat pump. Only if the tank drops bellow 40C would any of the top hot water be used.
Air to water heat pumps cool no problem. The issue is you would have to do some extra plumbing to isolated the heat pump and the UFH from the tank during cooling season. The simplest is to use a set of 3 ways and connect the UFH directly to the heat pump supply/return. You'll also need separate controls to control the cooling, might be simplest to set the UFH pumps to ON and adjust cooling by changing the setpoint of the heat pump output water temp. Make sure to keep this water temp setpoint above indoor air dewpoint.
"Solenoid is a guaranteed off." yep, this is my thinking, I have read about many options to stop thermosyphon but Solenoid just feels like it will work
Thank you for your example with the mixing valve, so if I understand correctly the valve mixes up to the desired template (and not down as I had been thinking). So, for example if the full store was at 45C and I wanted 40C it would only take this from the cool input of the valve.
"The issue is you would have to do some extra plumbing to isolate the heat pump and the UFH from the tank during cooling season. " - I believe this would be the same with the GSHP as well as I would not wish to store any cold water, while I have been dreaming of a GSHP for the last 6 years, I think you and DCcontrarian have killed my final justification for it!
>"Air to water heat pumps are readily available in UK, I would not bother with GSHP" same here, its the most popular option for installation and most installers will only do ASHP. Im very tempted to go with ASHP but also keen on GSHP to take advantage of running the system in reverse and having cooling in the summer from the pump.
A second vote in favor of ASHP. Because GSHP's are a niche product innovation tends not to arrive in their design as quickly. The latest modulating ASHP's are so much more efficient that they end up using about the same energy as a GSHP.
The comment on cooling puzzles me. ASHP's can produce cooling. If you're using hydronics it's a question of having an emitter that can capture condensation, some sort of fan coil unit. That unit would be the same regardless of the type of heat pump.
Thank you for your feedback, we have a large field next to the house where I was planning to place the GSHP, but your feedback is very valid and more and I'm nudged towards a ASHP. As for ASHP for cooling I will read more into this and will discuss this with the local installers so thank you for the suggestion.
Reading the energy modeling, they say the house is 200 m2 and the heating load is around 20 kW.
For those used to American units, that's about 2000 square feet and 70,000 BTU/hr. This sounds way out of whack for a house built to modern standards.
And my eye was drawn to this section:
"We suggest it may be possible to include some more thermal mass in the first floor (The
ground floor has a solid floor so there is good thermal mass there, though more would be a
benefit). One option we suggest considering would be if some internal walls could be
blockwork instead of studwork. For example, on initial inspection and as an example to
consider, the walls either side of the landing could perhaps be blockwork? "
Long-time readers know my views on the subject, I'll just say I consider those to be the words of a charlatan.
Hi, I think the concern was in regards to how the wooden log side of the house would be insulated, from outside: 20cm wooden logs, air barrier, 20cm mineral wool, vapour barrier, and plasterboard. So there was not as much "mass" on the inside to retain heat, we used fermacell boards for the floor and walls that gave us some additional mass.
But agree that many things can be mass (what I believe your point is) and we have not noticed any issues.
Comment about this statement: " With the store at around 50% 75c (167f) we have enough hot water to heat the upstairs, downstairs our concrete slab soaks up all this heat in around an hour, therefor we mostly have a cool floor downstairs (and lots of hassle from my partner)"
It took me a long time to realize this, so even though it seems obvious in hindsight I assume it's not obvious: the heat transfer from a heated floor to the room is entirely determined by the temperature difference between the floor and the air in the room.
The report says heating load for the house is around 4kW at a typical winter temperature of 0C. Total area is 200 m2. I'm going to use US units because that's what I'm more comfortable, then I'm going to ask ChatGPT to translate into metric. That's 14,000 BTU/hour at 32F over 2000 square feet. I'm going to assume that 60% of the heating load is on the second floor and 40% on the first, so the heating load on the first floor is 5600 BTU/hour over 1000 square feet. That's 5.6 BTU/hour per square foot. The rule is that the heat output of a floor is 2.0 BTU/hour per square foot for every Fahrenheit degree, so that equals a temperature difference of 2.8F. So if your room is at 70F that would put the floor at 72.8F. I find a floor temperature below about 78F to be undetectably heated.
Note that the above calculation is independent of water temperature or flow rate or any of the other details. It also ignores any contribution to the heating of the ground floor from the wood stove.
ChatGPT's response:
"That's approximately 48,752.92 Watts at 0°C over 185.80 square meters. I'm going to assume that 60% of the heating load is on the second floor and 40% on the first, so the heating load on the first floor is approximately 32,501.95 Watts over 92.90 square meters. That's approximately 350.73 Watts per square meter. The rule is that the heat output of a floor is approximately 4.4 Watts per square meter for every Celsius degree, so that equals a temperature difference of 1.27°C. So if your room is at 21.11°C, that would put the floor at 22.38°C. I find a floor temperature below about 25.56°C to be undetectably heated."
ChatGPT got it backwards, 1kW=3400 BTU.
Hi, agree that the issue we have is that we do not heat the ground floor often so the slab is very cold, when we do it takes heat quickly and releases it slowly as we would expect.
The temp sensor for the slab is around 60cm to the closest UFH pipe, when this is at 19c we find that the full house is a comfortable temperature to wear a short-sleeved t-shirt in winter.
"I find a floor temperature below about 25.56°C to be undetectably heated" and that's fine for us, we do not wish to feel the floor is heated, just not cold. A heat pump keeping the floor at a steady temperature I feel should meet our needs.
If you have any suggestions on the proposed changes to the plumbing would be most welcome :)
It's not something changes to the plumbing is going to fix.
"The temp sensor for the slab is around 60cm to the closest UFH pipe, when this is at 19c we find that the full house is a comfortable temperature to wear a short-sleeved t-shirt in winter." This suggests to me that the floor is contributing zero heat to the room. If it's really at 19C and the room is at, say, 21C, the floor should actually be absorbing heat from the room. Something else is contributing enough heat to meet the heating needs of the room and even warm the slab somewhat. My guess would be the wood stove.
>"I find a floor temperature below about 25.56°C to be undetectably heated" and that's fine for us, we do not wish to feel the floor is heated, just not cold.
If the room is already t-shirt warm with the floor at 19C, making the floor warmer is only going to make the room warmer. If you have multiple loops of tubing in the floor you can turn some of them off, which will make part of the floor warmer and part cooler. That's the best you can do.
I think this sensor is a little misleading and the slab is warmer, especially closer to the UFH pipes. A better example would be the supply temperature to the manifold that is around 35-40c with a delatT of around 10c for the hour or so it's on. If I can maintain a supply of heated water from a heat pump to this slab I believe I can maintain a comfortable level of heating without the need of the fireplace, if needed I can adjust flows, stop the manifold pump or close off circuits as you suggested (we have 4 in the main room for example).
The fireplace most definitely contributes to the heating of the space, too much if we have it on constantly (coldest days we find two burns meet our needs), heating the slab with water heated from excess solar also has the same effect of making the house feel comfortable.
The house is reasonably comfortable most of the winter with just the people and activities you have in the house and firing the fireplace once a day if no sun, but it would be nice not to have a cold floor that you need to wear slippers on and hot water for heating available on demand so a heat pump and modestly heated slab I believe would meet our needs. (it's a right pain leaving the house for more than a few days in winter as it takes two days to get back to a comfortable temperature and we cannot remotely activate the fireplace)
I think you're getting closer to where you're eventually going to end up: run the heated floor all the time in the winter, only run the fireplace when it's so cold that the heated floor doesn't provide enough heat.
I want to repeat what I said earlier:
The heat transfer from a heated floor to the room is entirely determined by the temperature difference between the floor and the air in the room.
This is so important, I want to make sure you understand it. If you don't, re-read it until you're sure you do.
You should be thinking in terms of a target floor temperature. If you don't have a contactless thermometer, get one. The question you should be asking is, what floor temperature is comfortable to me? Both at the lower end and the higher end. Then you need to figure out how cold it has to be outside where if you run the floor as hot as is comfortable it doesn't put out enough heat. That's the point where you need to start using the fireplace. What you then want to do is figure out a control mechanism where as it gets warmer outside, the floor temperature drops so that the room doesn't overheat. If you have multiple loops you can also choose to heat a smaller amount of the floor, if that works for how you live.
Notice I'm not talking about water temperature or flow rates or zone valves or mixing valves or any of that stuff. That will all be important eventually, but it all exists in the service of floor temperature. When we get to the point where you can say, "I want the floor to be at 27C and it's only at 24C" then we can start talking about those things. But we're not there yet.
Hello all,
I have been distracted with other projects so this was put on hold, as the heating season approaches I'm switching my focus back to this project.
Attached is an updated diagram with the suggestions from here - thank you all. I have decided I will go with a monoblock heat pump as I have limited space in the house and such a unit I will be able to setup myself as I will not need to touch any coolant. All installers I have spoken to are not interested in setting up cooling support and based on my experience to date I feel more comfortable making the changes to the plumbing and installing the heat pump myself.
The only thing that may change from the diagram is that I may not need the Expansion vessel on the heat pump loop if the heat pump itself has one that is large enough - I have not yet decided on what heat pump to get.
"I want to repeat what I said earlier:
The heat transfer from a heated floor to the room is entirely determined by the temperature difference between the floor and the air in the room"
@DCcontrarian, sorry for not responding sooner. Yes, I do understand that knowing the target floor temperature is key. I will need to discover this once I have the system working and can experience different heat levels from the floor.
Thank you all again for your feedback, has been very helpful. If there are any other comments or suggestions, they are most welcome!
Is the monoblock on a glycol loop? If yes, you can't use a 3 way valve to connect it to your underfloor heat, it needs a plate heat exchanger between the two.
Hi Akos, no it is using normal water hence the need for the auto defrost valves. I like this solution as it keeps things simple
I that case, no need to connect it to the coil, this just adds delta T thus lower COP.
I would T the cold mixing valve supply and return of your UFH and connect the heat pump there. This way if the UFH and heat pump is running at the same time, the heat is directly supplied without having to mix in the tank. This also means you only need a single 3 way instead of 4 as it only needs to select if the monoblock output goes to the mix valve or directly to the UFH. It also saves the 2nd expansion tank.
Hi Akos,
Im struggling to picture this, I think I am misunderstanding the suggestion:
"I would T the cold mixing valve supply"
With a 3 way value to the HP output?
"return of your UFH"
UFH return to HP supply? Would this not cause an issue bypassing the heatstore?
I understand that having the HP on its own coil will lower its performance, but I like the idea of separating the water from the heat store. So, for example, if the anti-freeze valves activate, I only lose the HP loop and not the full store. Also, this means I can use cooling in the summer by bypassing the thermal store and adding the UFH to the HP loop with cold water.
I know the aim is best COP possible, but is the impact of having the HP go via a coil into the store significant?
Connect it like this.
This will give you cooling and it won't move any cold water trough the thermal store.
For freeze protection, you don't want the system to drain down the water. What you want is for tap water from your automatic makeup water feeder to continuously flow through the system as this will keep the Monoblock from freezing. This way when power returns and the Monoblock needs to restart, it is full of water and ready to run.
For cooling, the pumps for your UFH needs to run continuously when the heat pumps is enabled. The way I have done this with mine is by setting them to a high heat temperature as my thermostats don't support cooling. The way you control the house temperature is by adjusting the weather depended cooling curve of the heat pump. This will automatically adjust the amount of cooling based on outside temperature. Very effective. One thing to watch is radiant cooled underfloor is extremely slow, mine took about 6h before I noticed much change. Your UFH pumps must flow enough in water to match the rated flow of your heat pump.
Hi Akos,
"automatic makeup water feeder to continuously flow through the system"
That's an interesting comment, I never thought of it working like this. It's just a valve I open to top up if needed, I guess leaving it open would make it automatic, it has a check valve.
Thank you for the picture, that's interesting, few questions...
1. Running in heating mode there would be no need for the mixer valve to take water from the lower part of the store as the only hot water there would of been from the heatpump. So the setup would be directly from heat pump and mix with top of the store if needed
2. In cooling mode I have a solenoid valve on the ufh hot feed I can close to disconnect the thermal store, would I need one also on the cold return back to the store?
3. Bypassing the thermal store does raise the question is there value in still having it if it's only for heating water to taps, as almost 50% of it will not be used. Would it not be beneficial to keep the connection with the store so that I have a volume of water already heated for ufh, so when ufh is activated I have a buffer as it takes a large amount of heat to warm up. Again I know the best COP is ideal, but this being slightly lower to allow a buffer of heat in the thermal store a reasonable compromise?
I know no one system if perfect and there are many ways to achieve this, finding this conversation very insightful, thank you
1. Yes and no. The UFH water flow will generally not match the heat pump. If you have say only one zone calling, most of the flow from the heat pump would be into the thermal store.
Most monoblocks run in output temperature control mode and will modulate heating power based on return water temp.
So if only one zone is calling for heat, the heat pump will supply heat to this zone and the rest of the heat will go into the thermal store. This will happen until the thermalstore heats up enough so the return water temperature to the heat pump is high. At that point the heat pump will shut down (but continue to circulate water). Your UFH will continue to run and now take heat from the tank which will cause it to eventually cool down. At that point the cycle repeats.
2. Generally you only need to disconnect one side of the loop. Look at a zone manifold, they only have one actuator. No need to add extra plumbing and controls.
3. You are not bypassing the thermal store, it is still needed to help match heat load and heat pump output during heating season. Even in the summer, you can switch from cooling and run the heat pump in heating mode to heat the water up for your domestic hot water. This is a standard function built into most units.
There are many many permutation especially if you add in 3 heat sources.
PS. "It's just a valve I open to top up if needed, I guess leaving it open would make it automatic, it has a check valve." Normally there is also a pressure regulator to reduce the tap pressure to heating pressure, just making sure there is one there.
Great thank you for your feedback, this all makes sense.
Sorry for the delayed reply. I have been talking to a local engineer about doing the job. At first, this seemed very positive as they understood the setup I have and the results I wanted, but they have rejected the job saying...
"The installation is not very complicated but we do not install heat pumps in such a system and we will not give you a guarantee that you will be able to control this heat pump well"
Their main point is that they do not wish to add a cooling function and insist on an additional heat exchanger between the thermal store and the heat pump. They also say that the return water from the back boiler of the fireplace can cause issues with the heat pump. I said I'm happy to go with a different suggestion if it can be explained why, then the job was rejected. I understand companies wish to keep installations similar, but I do not agree they should insist on setups without justification and I do feel they punch in details into a program that outputs a proposed configuration without understanding it.
So, back to Plan A. I hope to do the installation myself next week. My final diagram is attached, based on feedback, and slightly adjusted to reflect where the store's connections are located.
I have one last question:
Should I be worried about water from the fireplace back boiler being "contaminated" and causing issues with the heat pump? This is the first I have heard about this, I will have a Magnetic Filter installed on the Heat Pump supply. Maybe its best I switch the fireplace back boiler to the coil at the top of the store so its water is isolated from the rest of the system if there is a risk?
The return from the heat pump needs to T onto the return of the floor heat if you want to use it for cooling otherwise you are flowing cold return through the storage tank during cooling season which you don't want.
With the heat pump T to the UFH return, it will still use the buffer tank when needed for heating, so it doesn't change the operation.
Otherwise, I think it should work. You'll have to play with the mix valve temperature setpoint to get it dialed in and you might need some controls to disable the heat pump if the thermal store is too hot (say on a sunny day winter day and after running the wood stove) so the extra heat up top can be used up.
I don't have any experience with water jackets on stoves, but I can't see it being much different than the water jacket around a gas fired boiler. The cast iron in those can react with oxygen in the water and create sludge. Usually not a big issue if the system has the right oxygen barrier piping though. The in-line magnetic filter should catch any of this sludge in the pipe but I would also add a simple Y strainer as well.
Also don't forget to add isolation and priming valves as needed. Getting air out of these complex hydronic systems is not easy.
My first impression is that you would be best served to try and isolate the heat pump from the rest of the system. Heat pumps work best when their output is matched as closely as possible to the actual heating load. That way the compressor can modulate to match the load. Having that 800 liter storage tank is going to confuse the heat pump, it's going to tend to run full out and then short cycle, neither of which are efficient operation modes. A heat exchanger, as the local engineer suggested, would just make the problem worse.
I would expect that the heat pump would only run when there's no fire in the fireplace and no solar going on. The reason for that is that the water temperature off of the heat pump is going to be much lower than those other two, and there's no good way to mix devices of quite different output temperatures and have them "help" each other. Certainly not with a heat pump that is trying to match its output to the load.
So I would connect the storage tank only to the solar and the fireplace. I would have a thermostat in the tank, and when it hits a level below the operating point of the heat pump I would switch 3-way valves on both the input and output of the UFH so it's directly connected to the heat pump and turn on the heat pump.
This also allows you to isolate your cooling from the fireplace and solar, which don't want to be connected anyway.
If it turns out that the water coming out of the storage tank is unsuitable for the heat pump -- too dirty or too high pressure are the problems I can see -- you could put a heat exchanger between the storage tank and the UFH to isolate them.
The domestic hot water coil complicates things. If you're not OK with having hot water only when something other than the heat pump is providing heat then I would eliminate the coil and just have a standalone heat pump water heater. Heating domestic hot water is not a good application for air source heat pumps, I don't know why the manufacturers insist on promoting it.
Also, one of the features that makes heat pumps work really well with heated floors is outdoor reset. With the storage tank the outdoor reset on the heat pump gets defeated.
Thank you both for your feedback, in reply...
@Akos
"The return from the heat pump needs to T onto the return of the floor heat"
- Sorry your diagram was very clear and I missed this, have updated my diagram
"need some controls to disable the heat pump if the thermal store is too hot"
- Yes I have sensors to monitor temperature and also will have full control of the heat pump via Home Assistant, so my configuration should be very flexible
"I would also add a simple Y strainer as well."
- Yes this comes with the heat pump and will be after the magnetic filter
"don't forget to add isolation and priming valves as needed"
- Insolation, yes of course.
- priming valves, the thermal store has one on top and UFH manifolds each have one, I assume these would meet my needs as they are all above the Heat Pump, or would you suggest an additional one on the Heat Pump circuit to be sure?
@DCcontrarian
"Heat pumps work best when their output is matched as closely as possible to the actual heating load"
- From my perspective, I see the bottom of the thermal store as part of the heating load, I want this temperature to be around 40c (to be played with). When the UFH first comes on there will be a high demand as we heat our concrete slab so the buffer of 40c water in the store will meet that, and then the heating load will reduce as the UFH gets to temperature. I can see what you mean in that the heat pump may short cycle keeping the store at 40c, but I could reduce this by setting a target range say 45-35c. I may even have the Heat Pump configured to review the forecasted temperature and preheat the bottom of the store in advance and not to come on at all if we believe heating demand will be low or not needed.
"I would expect that the heat pump would only run when there's no fire in the fireplace and no solar going on"
- Correct, unless as said above we may have solar in the day for DHW and see that night will be cold and the heat pump then prepares the bottom of the tank using solar power. Fireplace I am expecting we may use very little once the heat pump is active, it should only be needed for very cold periods.
"So I would connect the storage tank only to the solar and the fireplace"
- I understand your perspective on this, and I would not be surprised if this may be something I may move to. But, first I wish to make use of the bottom of the thermal store as a buffer, if this brings little value I can see this setup being the alternative.
"If it turns out that the water coming out of the storage tank is unsuitable for the heat pump"
- Pressure: I do not expect this, DHW to the house is 1.5 bar, the heating system is pressured by this and the thermal store goes to around max 2.5-3 bar when at 75c (top half). This is within the range of all components. But, I will be monitoring this to be sure that is always the case.
- Dirt: then I would look at moving the fireplace or heat pump to a coil on the thermal store to isolate them, to be monitored
"If you're not OK with having hot water only when something other than the heat pump is providing heat"
- DHW will not be heated by the heat pump alone, it will allow the water to be warmer than it would be without the heat pump but solar or fireplace should be heating the top of the store to be sure we get the hot water temperature we wish to have. Also, another benefit of having the heat pump heat the lower part of the store is that the fireplace return will be closer to its target temperature of 55c so the heat pump will be supporting the fireplace in this situation.
Thank you again for all your feedback, I will be keeping this post updated with progress so others can benefit. Having this knowledge will allow me to monitor the system and make changes more confidently so thank you again for this.
OK, let's talk about how a heat pump works.
The key to controlling a heat pump is knowing that they have a compressor with a variable speed motor, and a circulator with a variable speed motor. So they have two direct controls, the speed of the compressor and the speed of the circulator. The compressor speed dictates how much heat they produce, and the circulator speed dictates how much the temperature of the water rises and thus the output temperature. You as the user don't get to directly control either, that's up to the internal logic, you just provide a flow of cool water and accept a flow of warmer water back.
The heat pump is happiest when the compressor and the circulator are running constantly at a consistent speed. In other words, when its output exactly matches the amount of heat that is being taken away. The water that flows out of the heat pump goes into one or more emitters, devices that release heat. They could be radiators, fan coil units, underfloor heat, whatever. The heat pump is happiest when the heat flow out of the emitters is constant and equal to the heat flow out of the heat pump, which will happen when the heat flow out of the emitters matches the heat flow out of the building to the environment.
So how do you match the heat flow out of the emitters to the heat flow out of the building? For emitters that have a variable output, like a fan coil unit with a variable speed fan, you can use a thermostat to modulate the speed of the fan and thus the output. For static emitters like a radiator or UFH you can modulate the flow of water using a zone valve and a thermostat. You can also modulate the temperature of the water, this is usually done with an outdoor sensor (known as "outdoor reset") and is less precise but still useful.
So the heat pump is trying to figure out what the demand for heat is, and all it sees is how much water it sends out into the world, what temperature it sends it out at, and what temperature it comes back at. The temperature it sends it out at is determined by the set point of the heat pump, which is either fixed or varies with the outdoor temperature if you're using outdoor reset.
So if there is anything else in the circuit other than the heat pump and the emitters, the heat pump is not going to be able to measure the load and is not going to work properly.
That said, the compressor typically has a minimum speed at which it will run, and thus the heat pump has a minimum output. What happens if the actual load is more than zero but less than the minimum output? The compressor has to cycle on and off. But compressors typically have a minimum run time as well. So the system has to have enough water in it to absorb without overheating the amount of heat the compressor puts out when it runs at minimum capacity for the minimum run time. In a system the size you're talking about that's about 60 liters or 17 gallons. That's where buffer tanks come in. The circulator in the heat pump is also going to have a minimum flow, and if your system has zone valves the buffer tank is also useful to ensure minimal flow.
The thing to understand about the buffer tank is that it doesn't enhance the performance of the heat pump in any way, it's a necessary evil. If you design your system to work without a buffer tank it will work better.
So a system like the one you've drawn up where a heat pump shares a large -- 800 liter! -- tank with other heat sources is just not going to work. What will work is having the heat pump share the emitters. But they can't share the emitters at the same time, you have to have some sort of switchover mechanism that enables one to be turned on and the other turned off. That could be as simple as a pair of 3-way valves. The heat pump will see those valves closing just the way it would see a zone valve closing, it will finish its cycle and shut off. When the valves open up again the heat pump will start seeing cool water at its input and turn itself back on again.
DC,
The OP has a large PV array that is now being used to run a resistance heater to heat the tank. Doesn't matter if this setup is a bit less efficient than optimal, it will still be a heck of a lot more efficient than running that resistance heater. Heck you can even set the heat pump to target 60C to heat up the tank most of the way and would still be about 2x as efficient as a resistance heater.
With the large buffer, what will happen is the bottom of the buffer tank will eventually settle down near the outdoor reset temperature targeted by the heat pump. Once this happens, the heat pump will still pick up the load from the zones. As they come on-line the return water temp will dip which will cause the unit to ramp up. Since the zone and heat pump flows won't be matched, this return water will be a mix of bottom of tank+UFH return which although not ideal for efficiency, it should be enough to cause the heat pump to module up.
Overall, I'm pretty sure it will work.
Define "work."
What restricts the heat pump to heating only the bottom of the tank? The heat pump is going to see that tank as essentially an infinite load until the whole tank is at temperature. So it's going to run full out and then shut off. No modulation at all. The tank water will oscillate between the high and low cutoffs on the heat pump. When the water is near the low cutoff it will stay there for a long time because the floor won't be losing much heat because the water isn't very hot. And the floor won't be putting off much heat.
Isolating the emitters and having them connected either to the heat pump or the tank, never both, will work and isn't really any more complicated.
He hasn't talked at all about how this is going to be controlled. The way I've laid it out is pretty straightforward. A thermostat in the tank controls the 3-way valve and switches between heat pump and whatever's heating the tank. The only thing that gets tricky is if you're relying on outdoor reset to modulate the UFH, in that case you'll need two separate outdoor reset circuits, one for the tank and the one that comes with the heat pump.
Otherwise, how do you control the heat pump. You can't just turn it on and off with a thermostat. That will burn out the compressor. You need to have some sort of zone valve to shut off the demand for heat, and a buffer tank on the heat pump side of the valve to keep the heat pump from short cycling.
What I don't see is any detail for how cooling is going to be implemented. If hot water is going to be available during cooling season then the emitters are going to need to be isolated from the storage tank and connected directly to the heat pump anyway.
The OP has a huge buffer tank, these can be stratified, so the the bottom can be kept at a lower temp than the top. The only time the heat pump will be able to heat the top portion is if it is set to provide hot enough water, otherwise it will only heat the bottom.
If you smoosh the indirect into the buffer tank in the attached picture, it is pretty close to the setup above.
Their setup is not ideal as it is using high temp water for the floor heat, my suggestion was a way to use the low temp water from the heat pump directly though the piping or from the buffer tank if the heat pump is off.
The combination of the 3 way and solenoid isolate the heat pump from the tank for cooling.
During cooling all heat pump water needs to flow through the emitters, so this is why my earlier remark on running the UFH pumps all the time and allowing the heat pump to set the water temp to adjust cooling.
P.S. You can also use the cooling setup (set UFH pumps to on and heat pump to heat with outdoor reset) for heat and only toggle over to the buffer tank when there is excess PV.
Two comments:
1. There's no reason to believe the storage tank will stratify. There are circulators on both the heat pump and the emitters, unless those circulators are moving exactly the same volume of water there will be flow from top to bottom in the tank which will tend to mix it. The tank will tend to stabilize at either the heat pump send water temperature or return water temperature, depending on which flow is greater.
2. Note that in that diagram there is an actuator line between the heat pump and the boiler, the heat pump tells the boiler when it needs help and the boiler fires. That's the only way you can run a heat pump in a circuit with other heat sources, the heat pump is in ultimate control and draws in the other sources as needed. The problem in this case is that neither a wood stove nor a solar panel are the kind of device you can just send an on/off signal to.
Edited to add: it's a control problem, not a plumbing problem.
Hi both,
I agree that my setup is not the most optimal, 5 years ago it seemed like the idea plan for GSHP and solar. So I have the 800 ltr tank is what I have to play with unless I decide to remove it.
The tank throughout the year with resistance and fireplace back boiler heating does stratify with a consistent hot 50% at the top and cooler bottom. The top meets our needs for DHW, my aim is to up the bottom of the tanks temperature what normally sits around 25c to 40c and to maintain it for our ufh.
I understand that connecting the heatpump directly to the ufh emitters makes a lot of sense and seems to be the common approach, but I have this 800ltr tank and if I can use it for additional heat if needed and as a buffer then I'm keen to try.
As for the resistance heating, is it efficient? not at all. Is it cheap, simple to setup, cheaper than solar thermal and has no risk of freezing, yes. Remember this is our spare solar power, I'm using it before giving it back to the grid (in exchange for credit).
For controlling the system I will be using Home Assistant, this will have full control of the heat pump, servos and readings of temperature across the house, heatpump, outside and forecasted. I will write some automations to control this and be sure that the heat pump is well treated.
I feel I'm moving in the right direction, but understand this may not be the final solution. At least by doing this myself I will have earned the knowledge and have the confidence to change this later. So far my experience with installers here they just do what's cheap and quick, not what's ideal. Also, I plan to renovate a log cabin next year and now I know how to install a heat pump and improved heating system, what's going to save a few bob doing myself :)
I will report back once I have made some progress, right now I have started adjusting the pipe work in preparation for the heat pump that should be with me next week.
The advice I would give at this point is to make sure you have the control strategy worked out before you do any plumbing. There are a lot of possible operating scenarios, I would make out a chart that lists them all and what you plan to do.
Just for starters, I would say the inputs are:
* There is a call for heat
* There is a call for cooling
* There is a call for hot water
* Solar heat is available
* Wood heat is available
You possible actions include:
* Enable distribution of heat to heating
* Enable distribution of cooling
* Enable distribution of heat to hot water
* Enable reception of heat from solar
* Enable reception of heat from wood
* Enable reception of heat from heat pump
* Enable reception of cooling from heat pump
There are 24 possible combinations of the inputs: heating/cooling is really a tristate of heat/cool/off, the other three have two states. (That's assuming you treat them as binary on/off and not having intermediate values like "some solar is available.") That's a lot of cases to work out.