Thermostat that can remember hydronic slab temperature swings and adjust.
I am building a house at 8500ft in Fremont County, Colorado, Zone 5A. The installed design has 11 zones, but I may combine some of them where it doesn’t make sense to have them separate.
The question is: Is there a thermostat (probably PPID-based) out there that is smart enough to notice a south section of the house with a lot of windows, and stop calls for heat hours before the sun comes up, to avoid temperature spikes in that zone?
Some rooms/zones are fairly simple and should be at the same temperature all day and night.
To be clear, I am not trying to create an outdoor reset like the boiler. I am trying to prevent regular large temperature swings in specific zones with an intelligent thermostat that can “remember” those swings, and notice a more rapid than usual floor temp swing and adjust for it, or at least significantly minimize the swing if it isn’t a regular event.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Great question and a very insightful discussion. Managing temperature swings in hydronic slab systems can be tricky, especially when comfort and energy efficiency are both priorities. A thermostat that can adapt and learn from past cycles would be a real game-changer for such setups.
Alongside energy efficiency indoors, architectural elements like Tensile Structures also contribute to overall building sustainability. They provide smart shading, reduce heat gain, and enhance exterior aesthetics—all while using minimal materials and offering long-term durability.
Really appreciate the depth of knowledge shared here. Looking forward to more such informative threads!
Great question and a very insightful discussion. Managing temperature swings in hydronic slab systems can be tricky, especially when comfort and energy efficiency are both priorities. A thermostat that can adapt and learn from past cycles would be a real game-changer for such setups.
Alongside energy efficiency indoors, architectural elements like Tensile Structure also contribute to overall building sustainability. They provide smart shading, reduce heat gain, and enhance exterior aesthetics—all while using minimal materials and offering long-term durability.
Really appreciate the depth of knowledge shared here. Looking forward to more such informative threads!
Looks like AI likes your question
I think the issue is not so much the rise in temps, since many people don't mind a 72 degree room rather than a 68 degree one, but the drop in the afternoon when the slab has been off for 8 hours. Both of these ills can be solved by not overglazing.
If your room is uncomfortably hot there is waaaay too much glass. If it drops too much while the slab catches up, there is not enough insulation[which is probably because there is too much glass]
Another answer is well designed overhangs, which can eliminate shoulder season overheating
I am sure there are some smart thermostats that allow infinite temperature adjustments, but the question is more how you manage it
If the system is not slab but underfloor, it may not be so bad because of the faster ramp up time
The problem is it's not so much remembering as trying to predict.
I have heated floors in my house and lots of solar, on sunny winter days my heat often shuts off in the afternoon. When the sun goes down it takes an hour or more for the floors to get up to temperature. I have a two-stage thermostat, the second stage is an air handler which basically can deliver heat instantly. So if the temperature dips too far below the setpoint the second stage kicks in.
The same thing can be done with panel radiators or any heat source that responds quickly.
I'm not a fan of high heat-capacity emitters, you get better responsiveness and comfort with low-capacity emitters. In particular I feel that concrete floors are way over-used and are often not the best choice.
Here is what I used.
Set the thermostat to maintain a min slab temperature that is about the room temp you want. This does require a thermostat that can do both floor and air temp sensing.
In the mornings a bit before your solar gain hits, set back the air temp setpoint. Keep this setback till you normally would get solar heat.
The idea is the fixed slab min temp will only deliver heat into the room if the room temperature drops, this will let the room coast for a fair bit of time and add a bit of heat if needed. If you have a nice sunny morning, the room will be up above the thermostat setpoint by the time the setback ends, so the heat won't run. If a cloudy day, the heat turns back on after setback and runs as normal but since the slab is semi warm, recovery is quick.
Ran this for a while until luckily my south facing IGUs failed under warranty and I had the window supplier replace them with low SHGC ones. This removed the need for any of this monkeying and made the temperature much more even without the overshoot issues.
Thank you, Akos. Your solution is very practical and will be my plan if there is too much heat in a given zone. Unless there is a smarter thermostat out there that can work even better.
I am looking for this thermostat as an additional measure of control. The house is designed with long overhangs, no excessive glazing, and the walls will have 3.5 inches of exterior Polyiso and 5.5 inches of dense-pack cellulose in the walls (as soon as I get done boxing the innie-windows). The vented attic will have R-60 of blown-in cellulose.
However, I still have to purchase thermostats, and since I am buying them, it seemed reasonable to try and find a smart one that can specifically adjust to my chosen heating system and address one of the primary complaints I have read about. I have made mistakes on this project and it has taken far longer than I ever thought it would to build it. Mostly because I am doing all the detail work myself, with one hired helper who has done a lot of construction.
I am currently installing the ERV ductwork and working very hard toward zero duct leaks.
The Westinghouse T6 and T10 Pro+ with floor sensor thermostats are good options if programmed well. Is there something better than these?
As an FYI, the house (when complete) will have R40 walls and R60 ceiling. It is very tightly sealed, as I am doing all the detail work myself. There are only 29 windows in the house, and most of them face NE because that is our view of the back of Pikes Peak. Because this is new construction, I have to buy thermostats for the house and want to buy the best there currently is for my chosen heating system. I have made a couple of costly mistakes in this process, and while I don't expect solar heat gain, it doesn't mean it won't happen. Does anyone know of an in-slab hydronic aware thermostat that is better than a Westinghouse T6 or T10 Pro+ with heat slab sensor?
If you have 11 zones that sounds like floorheating or radiators. With such an insulated shell and proper air sealing, the heat load will be low. Did you consider to go overkill with the emitters to enable lowest temperature profiles? (Overkill is subjective here..) If you can accept some room temperature swing: specially a floor heating system will have automatically less output if the room goes "up" at a given water temperature. A low mass floor system will help here.
What also helps is if the main control loop (that with the outdoor reset) does not look at the supply but return temp. If you have these external gains then the return will shoot up first so that the heating system will respond faster and throttle back.
A good hydraulic balancing will always be your friend.
This is really the point.
System design will allow you to do things that a thermostat cannot
outdoor reset
variable loop temperature
constant flow
Hi Wastl,
The house is dried in, and the heat load is low. It is 5400 sq ft all on one level. My problem is that the Manual J's done so far by myself using two software applications and one paid website, and two "experts" have it anywhere between 47K btu's and 74K btu's per hour. The boiler has been purchased and is a US Boiler/Burnham Alta 180. It will be supplying an indirect WH as well as the floor. Because I cannot get a reliable load calculation, I am installing a 40-gallon buffer tank in the hydronic system to avoid short-cycling the boiler. The reason I have chosen to heat the house in-slab is that the house is completely off-grid. In the future I have engineered a fairly large evacuated tube array that will supply heat to a large phase-change heat storage tank (using an inexpensive wax/graphite storage medium). The maximum storage temp will be around 160 degrees F. The slab will be the best way to efficiently distribute that heat. Many have told me this is a dead-end technology that won't work, but my numbers say it will, so this house is designed around my potential failure in the solar heating. But I think this will work and be less expensive and more efficient than nearly doubling my electric solar array, adding heat pumps, etc. Don't get me wrong, I like heat pumps; they make a lot of sense in on-grid applications. But the numbers work both financially and physically. I will add the hot water solar array after I have hard numbers on heat loss and can model the house/system exactly.
To echo the other replies.
-Use an OAT reset to limit SWT from the boiler. Ensure the high limit setpoint isn't any higher than necessary.
-Use a thermostat that has a slab sensor and ensure it's setup to high limit and possible idle (low limit).
-Minimize thermal mass in the slab. IMHO thermal mass is good when the mass is at the desired room temp. But when the thermal mass is elevated temperature it leads to issues.
-Some hydronic stats have control logic that will PWM (pulse width modulate) based on calculated demand so it will pulse at lower loads or close to setpoint rather than just stay on constantly until it overshoots. This can tighten up the control and reduce overshoots.
-run constant circulation if possible and use the coarsest zoning strategy that works. Many rooms have a similar load profile if they have similar exposures, so if they have a similar use over zoning may not have a benefit.
Thank you Josh, I have designed it with many zones to give me options, but will be combining several of those zones as it makes sense. Can you expand on "constant circulation"? As I currently understand it, each zone will call for heat as the thermostat determines it is needed. How would constant circulation work? Do you mean the pump would be on 24/7, and different zones would be open even if the boiler/buffer tank is not supplying heat? What is the benefit?
I am a fan of tekmar thermostats and control systems for this purpose.
I will look at the Tekmar thermostats. To be clear, the house shouldn't have a solar gain problem, but many have experienced this, AND I have (with unwarranted hubris) designed and engineered everything. The only professional engineers involved were for the redlines on the stem wall and the leach field. Because I am 100% confident of my ability to make mistakes, it makes sense to have inexpensive secondary and tertiary solutions to problems that may never exist, including straightforward ones involving thermostats, which must be purchased anyway.
I would like someone to say that one or more thermostats out there can be used for all applications, but were designed with AI and from the ground up for in-slab heat. If not, then I will go with the best technical solution currently available, which Akos and joshdurston have laid out very well.
If this house is as successful as I hope, much of, if not most of, the credit will be attributed to the brilliant and generous people who have helped me here on GBA!
mhenson,
for that size at one level most of the heat loss will be through the ceiling. Offgrid heat load matters more so maybe R60 should be bumped up if possible. (ongrid it would not make sense). Your idea about heat storage - I hear you - you know the winter weather better than I - if you have often enough sun to make that a feasible system.
With that size you will have a cold north side of the house and a warmer south one. One approach would be to have the primary loop with boiler and buffer tank, and then two secondary loops for the north and south side each. Each loop has its own controller and mixing valve to control flow temperature - dont stop a loop just reduce the temperature and maintain the flow. If the south side overheats you can reduce the supply temp without having an issue on the north side. The boiler would maintain the buffer at the north side outdoor reset temperature and the south can be run colder.
Alternative: (circ pump permitting): the boiler will feed the north side directly with the buffer tank in the return loop and the south side will use a secondary loop with mixing valve - that simplifies the hydraulics and safes pump energy (offgrid!).
what josh means with "constant circulation" is just that - dont throttle flow - control temperature. That way you have even heat delivery.
Whether a simple off the shelf controller can do that I do not know: If the south side overheats, keep the valves open and move the heat from the "hot sun soaked slab" to the cold south side. Shut the boiler off, keep the pumps running and all mixing valve "open".
Beside that - KISS - simplicity is a king. Don't microzone your system to death.
Thank you, Wastl. I will reduce the number of zones to six or fewer. It is nice to have the zone options to reconfigure if needed. I will see which controllers give me different options and work from there. I am clearly overthinking the thermostats and will just make sure they have a slab sensor so that I can follow Akos and joshdurstan's recommended programming.
Now, if I can finish getting the ERV ducting in!
I took some time to sketch out a model because this is a complicated question. I put a spreadsheet at: https://docs.google.com/spreadsheets/d/15zv1vdcSiVaplR8B-hvvLGt_-a-jQjx8xpdnrXma9QQ/edit?usp=sharing
It's read-only but you can copy it and play around with it.
Some of the parts of this:
You've got a house. Outside the house it's cold, the house loses heat to the outside. The amount of heat the house loses is proportional to the difference between the inside temperature and the outside temperature. The house has a certain amount of heat capacity, which governs how much the temperature of the interior of the house changes when heat flows in and out.
The house has a heating system, which replaces the heat that is lost. In this case it's a heated floor. The amount of heat the floor delivers into the house is determined by the difference between the house interior and the heated floor. There is also a heat delivery system: the tubing in the floor, the circulator, boiler, etc. The amount of heat that the heat delivery system delivers to the floor is determined by the layout of the tubing, the temperature of the water, and the flow rate.
The floor also has heat capacity. If the amount of heat flowing out of floor into the house is not the same as the amount of heat flowing into the floor from the heat delivery system the temperature of the floor will change; the amount it changes is determined by the heat capacity of the floor.
Throughout the day the heating load varies as the outdoor temperature varies, and the heat delivery varies depending on whatever control mechanism you're using, and the temperature of the floor varies because the heat delivery rarely matches the heating load.
On top of everything else, for some hours of the day there is solar gain, which is more heat dumped into the house.
For my model, I calculated heat flows and capacities per square foot. Here are the assumptions I used:
The house itself has a heat capacity of 25 BTU/F/SF. This is just a guess based on what I know about construction what I've observed.
The floor has a heat capacity of 15 BTU/F/SF. This is assuming a concrete slab 4" thick.
The outdoor temperature has a daily high of 30F and a daily low of 10F. This is just for the sake of an example. To make it easy to calculate temperature for every period of the day, I assume the daily high is at 3pm and the low is at 3am, and the temperature follows a sine-shaped curve.
At the daily low of 10F, the house has a heating load of 10 BTU/HR/SF.
The desired interior temperature is 70F.
The heating delivery system is capable of delivering 10 BTU/HR/SF to the heated floor.
Solar gain begins at 9AM and ends at 4PM. It ranges from 5 BTU/HR/SF to 20 BTU/HR/SF (at noon). These are just numbers I made up that seem reasonable.
For every time period, the temperature change of the house is equal to the net heat flow divided by the heat capacity. The flow out is the heat lost to the exterior, the flow in is the heat gained from the floor plus the solar gain.
For every time period, the temperature change of the floor is equal to the net heat flow divided by the heat capacity. The flow out is the heat contributed to the house interior, the flow in is the heat delivered by the heating delivery system. The heat delivered is determined by the interior temperature and the thermostat setting (column N). If the interior temperature is below the thermostat setting the heating delivery system is assumed to deliver 100% of its capacity, otherwise it delivers zero.
That's the description. Next up are some results.
I fiddled around with different thermostat settings. I found I could keep the interior temperature between 69.2F and 71.5F with a simple schedule:
9AM to 12 noon the thermostat is set at 69F.
5PM to 8PM the thermostat is set at 72F.
The other 18 hours of the day it's set for 70F.
Basically, in the morning when you're expecting the solar gain to kick in you ease back slightly on the thermostat to cool the floor off slightly and reduce the heat output. Then in the evening you get the boiler going when the sun goes down so that the floor will be up to temperature.
If you take out the solar gain completely, with the same model and the same thermostat settings you get a temperature between 69.1F and 70F for the day. So this seems to work whether or not there is solar gain.
Note that my control mechanism relies only on interior air temperature.
I've never understood why it was desirable or necessary to measure slab temperature. Perhaps someone can explain the thinking on that.
Thank you, DCcontrarian, you have made some very good points. After analyzing your data, I can only think of two reasons I might like the thermostat to take into account the slab temp, and they may be erroneous or irrelevant. One is that solar gain (if any), would tend to heat the floor and/or walls rather than the air directly. If in the morning someone is taking a shower (two bedrooms on the south side of the house with attached bathrooms), the ERV is in boost mode, fresh air is being pumped into the room at a given temperature, which may be lower than the thermostat air set point. It might cause the thermostat to call for heat when the floor temp is already relatively high (total speculation, with absolutely no data to reference or even day-to-day living experience with an ERV, yet to know if it is relevant).
The other reason is not related to solar gain, but is likely governed by the mixing valve. Is there a case, say a child left a window open in winter (yes I am a grandpa), where my system might continuously send heat until the slab temp is 95-100 degrees (wherever the mixing valve sets maximum fluid temp, and cause concrete expansion, tile popping off, or other problems because that zone's slab temp would be much higher? A slab sensor with a set point of 75-80 degrees should stop it before potential damage occurs.
If both are not an issue, I will get some reasonably smart programmable thermostats and call it good.
It's a reasonable use case to use a floor sensor as an upper limit. Although I'm not sure what that accomplishes that limiting the water temperature doesn't do.
What I've seen suggested is using a floor temperature sensor to modulate floor output, I just don't see how that would work.
It’s a smart thermostat that learns how the heated floor behaves and adjusts the temperature on its own for better comfort and energy savings.
As a decision, I am going to purchase some thermostats with slab sensors. I agree with DC that they are not needed. However, the technology is there, I am a geek, and there may be some benefit not yet identified. The price differential is insignificant, and I always love more data points. I will make sure the thermostats I purchase display the slab temperature so I can geek out on the differential between the slab temperature and the air temperature. That said, I will probably use smart programmable thermostats in the north end of the house, maybe one with a slab sensor so I can compare north and south ends of the house for SCIENCE.