Replacing Slab with Radiant Heat Tubing
What happens to a radiant floor slab system when, after maybe 45 years, the PEX wears out?
Normally I’d think the easiest way to deal with it would be to put another 2″ of cement with PEX in it on the top of the floor…but there are situations where I can imagine that that might run into problems. In particular, I’d like to use the slab for thermal heat storage so I can largely heat my home with solar. If I’ve got maybe a foot thick slab+sand mass that I’m heating (which–if my poor understanding of thermodynamics got it right–I would need in order to store enough heat for a cold Zone 6B night) then I imagine that having tubes only in the top two inches of a 14″ mass might result in overheating the home as I attempt to heat the mass sufficiently during the day.
So in this case it seems like it might be better to just plan on ripping up the slab when the PEX finally fails so I can replace the whole system and heat the slab not just from the top but also…wherever I should be ideally heating it from.
However I’m unclear whether ripping out the slab in an existing home is actually safe. Would it risk damaging the foundation wall? Maybe it depends on whether the slab is “floating”?
What if I have interior mass walls, would ripping up the slab next to them risk making them unstable (because they’re so heavy)?
I’m not married to any of this, I’m just trying to understand what the options are, and what the results of any decision would be. I don’t love the idea of designing a wonderful home that’s heated by the sun….for 45 years at which point the whole plan becomes worthless. I mean…I get that many homes last maybe 45 years and then get trashed. I’m just hoping for something a little better.
Thank you!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
PEX hasn't been around long enough for anyone to know how long it really lasts.
When radiant heat first got big in the 80's and 90's embedding it in a concrete slab was considered the way to go. Now we know that low-heat-capacity floors give better responsiveness and comfort. I wouldn't recommend tubing in concrete for new construction.
Functional obsolescence tends to be a much bigger problem than materials just wearing out.
Yeah, I think the reason I'm picking at this issue is that I want to understand the different ways of storing thermal heat from solar. One option is obviously using a huge water tank, but the cost of a tank large enough to heat a home--not just domestic hot water--would be $3,000. Plus shipping, plus installation, plus insulation, plus whatever the base price is for probably 81 more square feet in the home (8 foot diameter with 6" leeway plus insulation)... I think I'd easily be paying $6000-$7000 I'd guess. And I believe that was for a tank I was heating to 180 degrees. If I'm using a heat pump then I can only raise it to 120 using the hydronic system....so maybe I'd need...what...3-4X as much?
Whereas if I heat a concrete slab and sand beneath it, 120 F water will do that fine, and at a lower price, I think.
Placing a layer of insulation between the top cement layer and the sand could--if I'm getting this right--allow me to "store" that thermal energy and release it to the thin, top of the slab when needed (there would be much more insulation beneath and on the sides of the sand). It seems like a good solution....except for the question of repairing it all whenever the PEX gives out, which as you said we don't really know when that is but I don't want to assume it'll live forever and so I'm assuming maybe 45 years.
Which brings me back to my original question: What happens when one demolishes the interior portion of a concrete slab upon which a building is sitting?
PEX is cross linked polyethylene. It's essentially a more durable variant of "regular" polyethylene. Regular polyethylene has been in use in the telecom world (as cable insulation) since the 60's, maybe even a little earlier, and doesn't really degrade unless exposed to UV light. It's a VERY stable material. All the new fiber optic cables and underground ducts are made of polyethylene too, and we have several life expectancy "ratings" for those cables:
The IRS says they are good for 24 years. We engineers think that is silly, but the bean counters like it.
The manufacturers say they are good for 30 years. We engineers like that that's on paper, so we can point at it, but we think it's too conservative.
Practical experience has shown these cables to be good for over 40 years, and probably MUCH MUCH longer if not exposed to physical abuse or ultraviolet light. We engineers expect the underground duct systems to last over a century easily, provided no one gets into the booze too heavily before sitting down in the driver's seat in a backhoe...
I would expect PEX tubing to last for 50+ years easily in a slab, it's very unlikely it will "wear out" even by then, unless you run either very high pressure, or some really nasty water chemistry. Fittings may be more suspect, but in a radiant system, those are usually not buried in the slab, so they are accessible and serviceable. I think you can expect a radiant slab system to last at least 50 years, and probably longer if you're careful with it.
All that said, if the PEX fails, why add more concrete? Just abandon the PEX in place, and use the slab as a regular ol' "cold" slab. I see no need to put more concrete on top. In the commerical world, it's common to put in cable pathways and conduits in slabs, which are almost always abandoned and unused starting with the second tenant in the space. No one every worries about the abandoned cable raceways in the slabs.
The only reason I could see for putting more concrete over the top would be if you wanted to put in a new radiant system on top of a slab with an older failed system in side. In this case, you'd just need to leave a little more ceiling height to accomodate the now-higher floor, and you'd have more thermal mass to deal with which would slow down response times for your system. You could potentially limit the effect of that extra thermal mass by putting a layer of rigid foam between the old and new slabs.
In practice, I wouldn't worry about any of this. A lot of homes are demolished before they are 50 years old, which may not be ideal, but it does seem to be reality. There isn't much point making any one system significantly outlast the rest of a structure's expected useful life.
Bill
Thanks Bill
That's interesting about what engineers think of buried PEX.
My intention is to build a home that should last at least 100 years. My understanding is that it's done in Europe a bit, and I'd like to do it here in the states too. Maybe after reading more and making a nuisance of myself on forums like this for a bit I'll understand why it's not commonly done here but that is my aim at this time.
Which brings me back to my original question: Is there a way to build a slab such that removing it doesn't damage the foundation? Perhaps if the slab were thin and floating so it wasn't directly connected to the foundation wall?
The slab isn't normally attached to the foundation.
Keep in mind that concrete is one of the most environmentally damaging materials you can use in a house. Producing it releases enormous quantities of greenhouse gases. If you want to be responsible you should be thinking about ways of minimizing the amount of concrete in your house.
That's helpful. I had read a little about a "floating slab" but was unclear whether that was the norm or a rarity.
Does that mean that you'd be comfortable taking a jackhammer to a slab that was not attached to the foundation?
You would typically sawcut the part of the slab to be removed first so that your jackhammer work won't crack and damage the part of the concrete that is being left in place. You'll see this done often for road work, where they first saw cut a section of roadway, then they jackhammer that section and remove the chunks. After the chunks are removed, there are clean edges on the remaining concrete, which are typically drilled for rerod that is used to tie the new pour into the existing concrete.
Bill
Desert area in 6B is probably one of the few places with dry enough climate and sufficient diurnal temperature swing where a house without air conditioned and "thermal mass" can be made to work.
"foot thick slab+sand mass that I’m heating (which–if my poor understanding of thermodynamics got it right–I would need in order to store enough heat for a cold Zone 6B night) then I imagine that having tubes only in the top two inches of a 14″ mass might result in overheating the home as I attempt to heat the mass sufficiently during the day"
I think before worrying about the PEX durability, you need to set up a model of the house to see what you actually need. The chances of the system you are proposing working is pretty low and working as you hope is essentially zero. You'll most likely find that you need a lot more thermal storage and a pretty large thermal collector. Also chances of using the slab as storage is pretty low without significantly overheating the place. These are not easy problems, delta T is the killer in terms of human comfort, heat storage and solar capture efficiency.
Once you get the model up and running, post your results here. These are the kinds of projects that us engineers love and would be very interested in the results.
Yeah, if you're not modeling you're just guessing.
I have modeled it as best I know how--I have a whole spreadsheet with thermal conductivity coefficients that tells me how much heat I can store in a material per kg per degree of temperature difference, et. Using that I've done my best to figure out how many kg of material I'd need assuming X temperature differential.
There are definitely holes in my understanding though, and I would employ a professional. But professionals are expensive so I'm trying to ask what may be stupid questions here before I wind up paying someone $300 per hour to answer them.
However as you said, in climate zone 6B using a thick slab to store thermal energy is not uncommon, at least for "alternative" builds. It's what the local solar hydronic expert advocates for this climate. Plus, as I said, my modeling--which may be incorrect, I'm not an engineer and have no one to double check it--suggests that 12 inches of cement or sand should be about enough to store the thermal energy needed for one day. In reality I may have overestimated what was required, since daylight lasts for...I don't know, 8 hours or something in the winter, and I was just taking the lump sum of heating I'd need for an average day in January and multiplying it by two.
But my point is, I wasn't getting a result like "you need .5 inches of cement" or "you need a 500 foot thick slab". So that, and the fact that other people do it this way, has me hoping that I'm in the right ballpark.
Obviously the issue with using a thick slab like this to store heat is 1) the inability to stop a heated slab from releasing heat in the middle of the day and 2) the exacerbation of problem #1 the more heat you need to store. Seems like some folks just lived with those issues, but I'm hoping to avoid that by using the sand as a heat bank--in other words, the sand that sits beneath the slab will be surrounded on all sides by insulation with the cement slab--or simply cement flooring--sitting on top of that insulation. Hydronic piping will then be used to transfer that heat to the top of the cement flooring as need be.
This would allow me to use a heat pump to heat the whole house, which I like. I can't do that if I use water as the storage medium--or at least with water my choice is to spend maybe $15,000-20,000 on the water storage tanks (tanks+space in home + insulation + shipping + installation) that store 115 F water or heat the water from 115 F to 180-220 F using less efficient forms of heating and still pay maybe $6,000 for water storage.
As opposed to the cost of a bit of extra insulation, excavation, and sand. Which seems both more durable and less expensive.
But if this whole plan is not repairable...well, it would make me think twice, anyhow.
My recommendation would be to get a piece of software called BeOpt, it's produced by the US Department of Energy and allows you to model the annual energy usage of a building. It's designed to allow you to optimize energy usage by seeing the effect of various changes in design choices.
It allows you to enter heat capacity (or "thermal mass" as its sometimes called). Rather than listening to opinions here, I think you'd be better served by using the program and seeing for yourself how little impact "thermal mass" has on annual energy usage.
Do you live in an area where the electric utility allows net metering of solar electricity? If that's the case, the most effective way to use solar energy is to produce electricity when the sun is shining, which you bank with the utility and then withdraw when you need heat. The most efficient way to heat with electricity is a heat pump. Net-metering and a heat pump can often lead to a house that completely meets its energy needs from solar.
I suspect that what you will find when you model your house is that the lowest annual energy use and most cost-effective construction comes from building tight with lots of insulation and otherwise conventional construction. The big wildcard will be window size and placement, that is highly site-specific and by itself should make modeling worthwhile.
You'll need something more than a static model in Excel. You can set a dynamic model up with a bit of effort but there is software out there from passive house land that is specifically built for this that would give you more accurate results. Won't be cheap to run though.
My $0.02. A reasonable sized, well sealed and insulated house can be heated and cooled very comfortably by a $3k or so cold climate ducted heat pump. The operating cost of this especially if you have PV is so small that ROI on any complicated system is pretty much never.
It is fun to think outside of the box sometimes, there is good learnings there.
There is no zone 6B when it comes to building, which makes me wonder whether you're using the USDA plant hardiness map, which does have 6B. Oddly, these zones do overlap a lot, in spite of the numbering system being inverted (higher number is warmer in the hardiness zones, higher number is colder on the building climate map).
So in either case, speaking as someone in zone 6, an 8" slab and a Passivhaus rated building, I don't think you'll ever be able to passively heat your house. On winter days when it's sunny, our heating system doesn't run during the day. It runs during the night, even after those sunny days. This is with R56 walls, R120 roof, R48 under slab, 0.2ACH air sealing, 85% efficient ERV and <U0.12 windows (whole assembly).
My suggestion would be to jettison every concept of "heat mass" you have. None of it works well enough to justify the expense, either in $ or carbon.
I also have a PEX in floor heating system, which doesn't even get used. Unless you're going to have it fed by a heat pump of some kind, don't waste your time and money on that either.
I am using the map from here https://www.greenbuildingadvisor.com/app/themes/greenbuildingadvisor/dist/img/climate-zones.jpg
My mistake. Most references don't mention the "cold" zones being subdivided, but the subdivision does go up to and include zone 6.
Perhaps I should have been more clear: I don't intend to heat this through passive solar, but through solar panels. Right now the idea is to heat it with solar panels that power a heat pump, but I haven't looked into that deeply...will that require an inverter and battery setup that will make the whole thing cost more than I save on heating? I don't know. But whether I land on using solar-electric to power a heat pump, power a dc heating element, or use solar hydronic of some kind, all of those systems require some kind of heat bank.
So I think I get what you're saying. Some people selling some products oversell thermal mass. The benefits vary a lot depending on climate and method of heating. And big windows won't really get you all the way there in climate zone 6, and anyhow they're really expensive if you want a decent R value. And insulation by definition gives diminishing returns--probably someone somewhere has a chart of when it becomes ridiculous to double the insulation on one's walls.
So I'm not trying for perfection here. But where I'm building there's a lot of sun, particularly in the winter. And it just seems silly not to make use of that if it can save me money on my heating bill. And I believe I can using solar panels of one kind or another.
Creative people have been trying since the 1970s (and long before) to find the magic mixture of glazing and mass that work without a lot of success. Most of us who have researched and experimented with different approaches have settled on building airtight, well-insulated with high-performance glazing and not too much of it. In other words, typical Passive House details. If you do that you get very comfortable interiors all year with amazingly little energy use.
Thanks Michael
The concern that, not being the smartest person to ever consider this problem, I am rather unlikely to come up with some new, genius idea, is certainly on my mind. But for the most part I'm not asking these questions because I think I'll figure something else out that no one else has, but just because I'd like to understand why things are done the way they are and not another way. As I'm sure everyone here can appreciate, if I can understand "why" then I can make more intelligent decisions going forward.
I suppose I'm being a bit insistent in this case because I wonder if the reason that homes are done the way they're usually done is because heat pumps haven't been a big thing until recently, and they put a limit on water temperature--at least water which is heated only by the heat pump--which in turn makes using water as a heat storage medium less practical. So perhaps alternate heat storage solutions haven't made sense until now?
So I'm still wondering if someone can explain to me why I wouldn't want to store heat in an insulated sandbox under my floor.
If it helps, here's an image of what I imagine I'd do. The bottom of the foundation is not included but hopefully it gives a good idea of some version of what I'm imagining.
For what it's worth I would bet money (as I'm sure yall would too) that other people have thought of it before I did, and I'm definitely open to hearing the reason why people don't do it. Much better to know now than to figure it out after the fact. If it's just that water-based heat storage solutions were marginally cheaper...well then ok. If it's because there's some huge expense I'm not seeing--well I'd like to know about that too. If it's because breaking up a concrete slab is a huge undertaking that costs many times what it cost to install...that's something else I have not much knowledge of.
So truly, if you've got the time to explain I'd love to hear why folks have deemed this impractical.
Desert_Sasquatch,
"So I'm still wondering if someone can explain to me why I wouldn't want to store heat in an insulated sandbox under my floor."
There are a few drawbacks:
- Trying to control a large, hot surface in a way that allows you to maintain consistent indoor temperatures as the ones outside fluctuate.
- In a well insulated house, storing heat only makes sense if you have an intermittent source. Otherwise why not just use the heat as it is generated?
- It's more costly and complicated than putting the same effort into reducing demand.
Storing heat in some medium - rock, sand, concrete or water - was a common strategy in the late 70s. One of the most prominent proponents of the approach was William Shurcliff, who wrote a book on three test houses using solar collectors and basement storage. https://www.amazon.com/Super-Solar-Houses-Saunderss-Low-Cost/dp/0931790476
He later repudiated the approach to focus on highly efficient envelopes, but you may find his book interesting all the same.
Martin's blog on the subject of thermal mass: https://www.greenbuildingadvisor.com/article/all-about-thermal-mass
Let me throw some numbers around that are just examples, but also somewhat representative. Let's say you have a house with a 1000 square foot footprint, and that on a cold winter day it takes 30,000 BTU/hour to heat it. You get enough sunshine during eight hours of the day to heat it for the whole day and you want to save that sunshine during the day.
During the 16 hours that the sun isn't shining, you need 16*30,000 or 480,000 BTU to be stored. You need to store that during the eight hours that the sun is up, so you need to average 60,000 BTU/hour during those hours. A rough rule of thumb is that the heat flow from floor is 2 BTU per hour per square foot degree of temperature difference. So to get that 30,000 BTU per hour at night with a 1,000 square foot floor you need 30 BTU per square foot, which means that the floor has to be 15 degrees warmer than the interior. So let's say you sleep at 65F, the floor has to be 80F, that doesn't sound unreasonable. But what about loading it during the day? To load that 60,000 BTU/hr you need 60 BTU per square foot, or a temperature difference of 30F. To get that, you either need your interior temperature in the day to be 110F, or you need a magical floor that is 80F at night and 40F during the day. And that magic doesn't exist as far as I know.
So it's not going to work with natural heat flow. OK, what about if you use a heat pump? Run hot water into the slab during the day, and then run cold water into it at night? It would work but you're going to run into the same problem that in order to get significant heat flow you need to have a significant temperature difference. And the energy use of a heat pump goes up with the temperature difference. You'll end up using around the same amount of energy as if you just had a heat pump, but with a much more expensive and complex system.
I don't have the perfect answer. But I was just researching a similar issue this morning. You might want to compare the options of doing electric resistance floor heat vs hydronic. There are pros and cons on both sides and a lot depends on your specific situation. But don't overlook the option.
I would say, based on past experience and that of others, that electric resistance floor heat is generally less reliable over time than hydronic systems. That said, electric resistance floor heating systems are likely also much easier to replace if they fail.
Bill
Thanks Bill and JollyGreenShortGuy
Yeah I share Bill's concerns. I'm inclined towards components with better longevity, and I prefer hydronic heat, which can make use of either solar hot water panels or a heat pump (I read about some pretty crazy tax incentives out there right now for heat pumps, but mostly for people who don't make much money).
But it's true that it's an option that's out there, and appropriate for some situations. It's just not what I'm looking for (but I appreciate the input).
Just to let you know, I have a home in France and electric is very common. My house had a concrete slab with ugly tile. I put the electric mats down in the living and dining rooms, right over the existing tile, and did new tile over the whole thing. I'm very happy with the results. I have only had it for 2 years so I don't know about longevity. But I can't see why it would wear out. It will certainly never spring a leak or get brittle. I did the work myself and it only took a few days.
You can get a 172 gallon water heater tank for $1700. Heat that to 180*, and you hold a large number of BTUs, nearly 160,000 relative to a 70* interior.
If you want to store energy in sand, you need to decouple it from the house in some way. Dig a pit, hold a volume of sand 6ftx6ftx6ft, heat it to 100* and you have similar heat capacity to the 172 gal heater. If you insulate that to R16 with 4" EPS on 6 sides and have it buried in 50* dirt, you'd be around 800btu/hr heat loss. Would it cost less than the 172 gal tank? Probably not, unless you own an excavator. In the end, you are storing 40ish kwh of power/heat. You could also buy a forklift battery and have similar energy storage, but more usable year round. If you had a heat pump anyway, your battery size could be much smaller, since you aren't pulling heat at 1:1 from your hot sand pit.
Keep in mind - as far north as 6B, you may have 8-9 hours of sunlight in the day, but your peak solar hours are really only 3-4 hours for south facing panels.
https://www.mbtek.com/collections/air-to-water/products/apollo-bwt650-buffer-water-tank-172gal-stainless
That 40 kWh of electricity at $0.20 per kWh is worth about $8. If you could get the same heat with a heat pump with a COP of 3 it would cost about $2.75. If you heating season is 180 days that's $495 per year. Plug that into a mortgage calculator and see how much equipment that buys.
For a typical house that might need 30,000 BTU/hr that tank provides about 5 1/2 hours of heat -- although it gets really hard to get the heat out as the water temperature gets close to room temperature. If you can heat your house for twenty hours on 160,000 BTU, congratulations, 8,000 BTU/hr is impressive, but your energy bills are already going to be low.
The problem with storing water at 180F is that the efficiency of solar collectors falls dramatically as their temperature increases, you're going to have trouble collecting much of anything at that temperature.
Let's also look at the thermodynamics of storing heat in sand. If the house is at 70F and the sand is at 110Fyou need to run your hydronic fluid at a temperature that's low enough for heat to flow into it from the sand, yet warm enough to provide heat for the house. You'd probably have the fluid exiting the sand at about 95F and returning around 85F. If you start looking at the size of the heat exchanger you'd need in the sand, and the radiators you'd need in the house, to get any kind of meaningful heat output I think you'll see that it's impractical. And as the sand cools the heat flow gets lower, there's no way you'd actually get all of the heat out of the sand.
To heat the sand up to 110F during the four productive hours of solar available to you, you'd have to run water around 140F. Then you'd run into the diminishing output problem with the solar collectors.
Yeah, I was more doing a straight BTU calculation of Sand temp at 100 - indoor temp at 70. But there would be much more involved in actually transferring that heat. Pump losses too.
The way to heat water with solar these days is to just use PV. You can do PV direct to a DC in-tank heater for a really simple/cheap option. Get some cheap panels from Santan Solar and plug in an element. But actually storing enough energy, even for a really well built house, it's a decent amount of energy you have to store from peak sunlight hours.
Essentially what we are talking about here is a way to heat with PV in an off-grid house. If you are set on electric heat - Battery+Heat pump and lots of insulation will probably be the most reasonable way. With how cheap PV panels are (and still dropping), I've had this thought as well. Even created a few excel sheets with 24 hour calculations to see what it would take.