UFH controlled by return temperature

[LukasV]

Hi guys,

 

I got a lot of info here to the possible system. I especially appreciate the infos about systems some of you which hold system perfectly tight with just one thermostat with tight hysteresis. But I got another game-changing idea.

 

Unfortunately I am struggling because I cannot find a room/place, where to place that thermostat. Every single possible place, that I have thought about, seems that it can be unpredictably under strong influence of heat sources.

 

I obviously must make the system somehow different. 

I will use a buffer tank, this will help me to disconnect the ASHP from UFH. I can charge the tank anytime I want, how much I want in advance (which will help me to ameliorate the electricity price spikes or outside temp drops..

And the UFH will be supplied with hot water from that buffer tank based on the needs of the house.

Charging the tank is easy task with temperature measuring in the tank, the rest will do the ASHP self.

 

But the house... ouch. 

1) theoretically I could use weather compensated heating with a lot of thermometers (one in every room?) and apply a logic on them, that would work with mean/median temperature for example. Then I can have a chance to get something like real temperature/real house needs and regulate with low hysteresis like you do. At least, I hope so. But that would be also really really expensive solution and possibly unstable.

 

2) But I got another idea - I could set a flow temperature from the tank to UFH to a fixed value. I can calculate the highest UFH flow temp needed in cold winter and set it accordingly. With that flow temp I can heat all year around. What the problem is, is the fact that I muss find a correct timepoint when to stop heating from tank and just circulate the water that´s left in UFH to prevent overheating. I hoped I could use a return temperature - this represents the ability of the house to accept more heat (with the flow temperature fixed and aimed air temperature). ie. when the house is fully saturated for the aimed room temperature, the return will rise - and that can be the point when I could disconnected buffer from UHF, thus no more heat from buffer, just circulating UHF water.

But WHAT is that correct return temperature that represents the full saturation??

 

Let`s assume I want a room temp 22, the mean water temp (mean of flow and return) must be 2 degrees higher to satisfy the house need in -12 degrees (heat loss 2600W, UFH with 200m2 heating area, ie with usual 6W /m2K must be mean temp at 24 to have a 22 room temp) and I would set the flow temp to 26. This means that when the flow is 26 and return is 22, then I am heating just enough to cover the actual heat loss. When I start the circulation, I could get something like flow 26, return 18, which in time gets to 26/22 and would go further until having 26/26 (then the home accepts no more heat and would be reaching to a room temp 26). I would prefer to get to 26/22 and then to make a little booster, then close the valve a let the UHF circulate until next heating period will come. So in that case I could stop heating at for example 26/24 and then wait some time (let say until next hour or until return drops enough). But that can be a solution for this situation, but I am looking for something universal. When the outside temperature is +10, I would need mean temperature only 0,8 above room, thus with 26 flow, I would need the return to be only 19,6. ie ((26+19,6)/2)=22,8 ie 0,8 above room. At that moment stopping the heating when the return temp reaches 24 or more could cause dramatical overheating.

So what is the solution? How to set the expected return temp at which I need to close the heating from buffer?

Ofc, I know the heat loss et every moment of outside temp thus I could stop it accordingly, but what if the calculation is not enough correct? 

Can I do it successfully with setting a fixed return temp too?

 

Do I have a solution?.... An example:

Flow temp is set to fixed 26, return temp to close the heating valve is set to 24, mean temp when the heating valve closes is 25. Requested room temp is 22. When outside is -12, the heat loss is 2400W, so this setting would need to heat 40 minutes in a hour to meat the heat loss (thus far enough).

But it is strong winter. The morning temperature is really cold and hits -5 degrees but when the sunshine starts the outside temp goes quickly to 10 degrees.

- It is 8:00 , outside is -5. Heat loss is 2100, I would need delta mean temp 1,6, instead of heating of 60 minutes I would need only 32 minutes to meet the house needs. At that moment, the return would be 20,6, but I will heat more until the return reaches 24 (which can happen anywhere between almost immediately and indefinitely according to house heat capacity and looses). But with simplicity I have boosted the house with cca 2000W more than needed. Because the diference was 25-20,6 = 4,4K - this was the maximal difference, the mean difference is 2,2K. UHF hat an output 1200W/K (over 200m2 area), thus 2640W total. The heat valve closes and waits until next command to open.

- It is 9:00, system again opens the valve. Outside temperature is 0 degree. But between 8 and 9 the loss was only 1900W but I boosted 2640W, the return is still above 24, so the valve immediately closes and I am not heating anything, just circulating a hot water in UFH.

- It is 10:00, system again opens the valve. Outside temperature is 5 degrees. For that temp I would need only 1400W heat output to meet needs which would need delta mean 1,1 thus heating only 21 minutes. But between 9:00 and 10:00 I lost "on average" another 1600W. Total between 8 and 10 is 3500W and at 8 I boosted only 2640W, thus the return temperature is bit under 24, the heating starts. The necessary heating to reach missing 860W is supplemented quickly and also quickly is reached return 24 because there is not a lot of capacity where to store the energy when the house again reaches equilibrium at 22 with outside temperature 5. Also the sun is up and starts shining inside, that shortens the next booster time again, same as do the family and friend coming inside to celebrate at birthday party or so. In next few hour the suns is making its job heating the house and until sunset no more heating is needed so every opening a valve will find out that the return is at least 24 and close the valve again immediately. With the sunset the return falls and the process comes again.

 

---- I understand that the numbers are not prefectly correct, the heating/loosing is continuous, not relying just on one temperature at the moment. I am just talking about the principle. That sounds reasonable and really good to me. 

For that system, I would need just one mixing valve that will keep the flow temp fixed at given temp, one sensor sensing the return temp, one valve disconnecting the UFH and buffer tank and one relay which every hour opens the valve and closes it again when the return temp reaches the preset temperature. The system is self-solving, self-healing, self-regulating. No matter what happens outside or in the house would complicate it. When the house is full of heat, it just closes heating quicker, when it is cold, it heats longer. It supply just enough to keep the temperature totally stabilized without needing anything other like thermostats, specialized weather compensing systems, outside sensors,.... just nothing.

 

It seems so studiply easy that I must be missing something bc otherwise this system would be the best at the world and totaly universal at every house. So what I am missing, what is the problem?

 

Of course the acting of the system is strongly dependent on the house type and its heat capacity - in a wood house the return will rise steadily, in accumulating house it will take its time. But at the end, during the time, it does not matter, the accumulating house will release the stored heat thus the return will have not a change to chill/fall quickly, but in wood house it will fall quickly because no heat is stored. Thus the difference would be just between the length or frequence of heating periods, but would work good in both of the extremes.

Also the flow speed of water must be set at some fixed rate... probably somewhere in the middle of possible values. 

But still seems to me like an ultimate system. What am I missing?

 

Please what do you think about it? Can I use it? How to make it better? 

[JohnMo]

27 minutes ago, LukasV said:

will use a buffer tank, this will help me to disconnect the ASHP from UFH.

Ditch the idea you are over thinking it. Ditch the buffer, ditch the thermostat(s). You have taken zero account of floor inertia. Which can be huge. You can simply buffer excess energy in the floor, way more capacity than a water container buffer.

 

Simple - your heat pump already manages the return temp. But it needs to be directly connected. No mixers or additional pumps needed.

 

You set a WC curve to match your heat loss.  The compressor within you heat pump is started and stopped based on target flow temp, a delta T and therefore by default the return temperature is managed. The compressor starts based on return temp, stops based on flow temp control hysterisis.

 

A simple self regulating system is ASHP, WC, UFH manifold. You really don't need anything else. Your room gets solar gain, the floor stops giving out heat, continuous circulation moves the energy elsewhere in the house.

[Alan Ambrose]

Yeah agree - our present place takes a couple of days to get from cold to normal room temperature.

 

What’s your floor build up?

 

Interesting ideas though.