PEX in a sub-slab sand box; where?
For you folks who put a sand box under your slab and heat it, where do you put the PEX? I’ve seen it at the bottom, but it makes more sense to me to have it near the center line.
My reasoning is thus (assuming you want the sand uniformly heated.): Say you have 2′ of sand under your slab, and foam below the sand. First, assume the heat in the sand is uniformly distributed, ie, it’s all at the same temp. To heat up that sand reasonably uniformly, you’d want the PEX at 1′ below the slab and 2′ OC. However, since the sand is likely going to be cooler at the bottom, you may want the PEX a tad below 1′. Too, the top of the sand will lose heat faster than the bottom, so you may want the PEX a tad above 1′.
Then you’ve got the problem of the sun (and your slab-heating method, if you have one) heating the top of the sand more than the bottom, so you may want your PEX below 1′. Not being able to move the PEX as conditions change, where are people generally locating the PEX, and what is the reasoning?
Second question: Since sand is not going to “take” heat as well as concrete, what size PEX, and what spacing? Siegenthaler says the pipe size does not make a big difference in delivered heat, despite the linear differences in circumference. BTW: I’m not interested in sub-slab water storage right now; I’ll have that super-slab w/ my wood gasification boiler.
Thanks. john
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
John,
Here's my opinion -- subject to revision when someone gives me good monitoring data to contradict my statement: you can put the PEX wherever you want, because these systems don't really work.
To pull off useful heat from these sand beds during the cold heating months -- mid-November to mid-February -- you need the sand to get hot enough to deliver water for a hydronic heating system at 100°F or higher. That means your sand needs to be hotter than 100°F during those heating months. It's not going to happen. The sand doesn't get that hot -- or if it does, it doesn't stay that hot from early September (when it is likely to be hottest) until mid-November (when you begin to need it).
Moreover, the pumping energy is a big energy penalty -- parasitic energy that needs to be considered when analyzing possible benefits.
Finally, the capital costs of all those extra solar collectors is high -- an investment without a significant payback.
Martin: Thanks. I have read some debates about this here, and am struggling w/ it. I see how many tons of sand can store, and slowly return heat, but I worry about how I am going to keep it warm when solar is not enough. And "not enough" may be many months. My plan, if I go this way, is to heat the sand with the wood gasification boiler as well as passive solar. However, when the wood boiler is not running and solar is not enough, my propane boiler, which will heat the slab, is going to have to work its butt off. On the sand temp, I don't see any need for it to be 100 F, as it won't be doing active heating of PEX, but rather just storing heat for night time, or maybe a couple of days. I have not looked at the numbers yet, but I suspect that if I could get the sand up to 80-85 F it would carry me for a few days. That is just a hunch, and could be way off. In lieu of good monitoring data, I, too, am concerned about this method up here. Thorsten Chlupp uses it, and apparently with good success, but his houses are designed differently than what I'll have. He and I need to sit down and chat.
I think we are talking apples vs oranges again:
If you want to be able to capture and utilize passive solar gain there is in my experience no better way of doing this then with adding INSULATED mass to the foundation. A layer of sand between the slab insulation and the actual concrete slab is the cheapest and most economical way to do so. I added 180 tons of mass for $380 in material costs at the SunRise home. It provides 3 functions:
• Insulated heat sink for passive solar gain and storage
• Moves all your under slab plumbing in conditioned space, minimizes your thermal bridges and acts as a drain/waste heat recycler if the temperature differential is right
• OPTIONAL active heat storage if secondary solar heat lines are placed
Sand is a poor medium to store heat but is very cheap and maintenance free. Having this additional storage available to dump heat from a solar thermal system offers a much higher system efficiency as you often times have lower temperatures on your collectors then you can utilize for DHW - but is more than plenty for heat your home. It would make no sense to heat your store actively from the bottom up with anything besides solar heat.
Anyways, I argued these points before and probably will do so many more times as it seems an alien concept for many. I have independent data collection on my system and that might bring a bit more weight to this argument in the near future. To me at least the fact that I have not actively heated my home since 02/16 besides passive solar gain in temperatures well below freezing kind of proves that this concept works fairly well if it is implemented right. I never even had to load my sand bed actively this spring as it stores the suns energy so well that my slab is over 70F - my kids run around barefoot all day on it, nice and comfy...passive solar energy is the only free energy there is and should be the very first source we should always tap into in a heating climate.
John - pipe placement is dependent on what you are trying to accomplish. If you use it as option to store heat you want to be on the bottom - as it is a simply heat in heat out with no means of control besides your temperature differential - Delta T. If you would not place heat lines in your concrete you would place them under your actual slab on the top of your sand bed. That is most likely what I will do in the future to save costs and be able to run high temperatures from my solar heating system and not be having to mix down to 140 max for supply temps.
Martin: I send you the report and data when it becomes available...sand beds are strictly passive heat sinks, you dump heat into it and that heat moves in the direction of the Delta T. There is no control, nothing to extract....an active system you describe where you trying to pull energy back out again to make radiant heat elsewhere would not work. Heat in - heat out by temperature differentials is all there is to it and it takes some figuring out on finding your comfort zone as it functions and reacts very slowly. If you're trying to use it as an active storage - like a rock heat bank - I agree with you, it makes no sense and is not feasible. That's when you need a much better medium for storage, water as your cheap option or a form of PCM. Than you can distribute heat and produce DHW...two very different systems. Apples to Oranges ;- )
PS: solar collectors are not expensive if they replace your boiler....
Thorsten: Thanks for the clarification there, and I hope we get some other voices on this matter. As indicated, I am trying to get my brain wrapped around this, and the matter of having to keep all that sand warm for three months via propane (if necessary) is yet a concern. I just have no real idea how much solar energy my lot will be getting in those three months, especially w/ the size of windows I have. Maybe your little high-tech meter will answer that for me. Here's hoping. I still think that the lower PEX should be near the middle, and some kind of sensors to indicate upper and lower sand temps. I think that would be cheap, fairly easy to monitor, and would cut down on the T&E guessing how long to circulate heat from the boiler (or when to close the drapes) to get a proper response. It may give a little more control over the situation. Cheers. john
Thorsten,
I agree that sand is cheap.
The problems with your approach is:
1. Where do you put the sub-slab insulation?
2. How do you control heat flow from the sand to the slab?
If you have no insulation between the slab and the sand, and you are dumping solar heat into your slab in July and August, then heat will flow from your hot sand to your slab in July and August. That may work in Fairbanks, but it won't work in any climate where summer overheating is a potential problem.
Thorsten,
One more point: consider the following two scenarios:
Scenario #1: You keep the interior of your home at 68 degrees F all winter long.
Scenario #2: You keep the interior of your home plus 180 tons of sand at 68 degrees F all winter long.
Scenario #2 will always require more heat input than Scenario #1 -- unless, of course, you have a source of free heat. And heat generated by $3,000 (or more) of solar collectors plus $800 of pumps and controls isn't free -- nor is the electricity required to run the pumps.
Martin: "...unless, of course, you have a source of free heat." Which he does most of the year. That is his point. Now, as far as "cheap", that is relative. My 330 tons of sand will cost over $3K, because it won't walk from the pit to my house. By volume, sand has about 40% of the heat capacity of water. (Specific heat of sand is 0.19 btu/lb-F, whilst water is 1. Sand is roughly 2x the mass/volume.) Dry soil (no reference to type of soil was given) also has a 0.19 heat capacity. What I see as being the critical issue here is not keeping some big mass of sand warm, but taking in the btu's that are lost on a daily basis. Once the sand is 68 F, all you have to do is meet equilibrium. Doing so w/ a wood boiler is a cake walk; my question is, can I do that passively? Re-thinking the heat loss from that mass of sand, I don't think the mass is the issue. I don't see how a btu lost from 1,000 cf of air is any different than a btu lost from one cf of sand. I don't see any mass parameters in heat loss equations. Does an empty house lose heat slower than one full of furniture? I don't see why it would, but, again, I could be missing something here; happened once before in my life.... Let's keep hashing this deal out. It is quite an appealing concept to me.
John,
The cost of one year's worth of heat produced by a solar thermal system is not free. It is equal to:
[(The capital cost of the equipment) / (The expected lifetime of the equipment in years)] + (The annual cost of the electricity required to run the pumps) + (The annual maintenance cost to repair and maintain the equipment).
Bringing tons of sand up to temperature, and maintaining the sand at an elevated temperature, takes heat. If I build a house without that heating load, my annual heating load will be less than yours.
Martin: Thorsten "solar thermal system" is his slab and windows. You seem to keep missing that. Again, I don't see the mass parameters in heat loss, so I can't see the mass as being more of a heat drain. It seemed to at first blush, but when I got to thinking about it, I don't think mass is of importance. In other words, a ton of sand won't lose heat faster than an ounce of air in the same room. Q will be the same. If someone can explain the physics of why I am thinking wrong, please advise. I love learning from my mistakes. john
John,
There are two factors at work:
1. The initial heat required to charge the system -- that is, to bring tons of sand from 40 degrees F to 80 degrees F.
2. The fact that the thermal envelope has been expanded in volume, increasing the surface area of the thermal envelope, and therefore increasing the rate of heat loss.
John,
Here's another way to look at it:
House A has a basement filled with 1,152 cubic feet of sand.
House B has a small basement (12x12) with a bar, sofa, and TV.
Both houses have the same thermal envelope and the same rate of heat loss. You get a basement full of sand. I get an extra room in my house.
Thorsten, I want you to know that I am enjoying following you here at GBA. You are what is great about those of us who take a risk and work on alternatives to all the standard ways we build and heat homes.
Martin: I thought I sent this once, but apparently did not. That is an interesting look on volume, but I don't quite see how dropping your floor 20" equates to a room; maybe for my mini Dachshunds??? Too, that room will never gain you any heat. Cheers. john
Martin, come on now - you can do better than that.
"Both houses have the same thermal envelope and the same rate of heat loss. You get a basement full of sand. I get an extra room in my house." Are you serious???
I have been called ignorant before but not stupid. We are discussing insulated internal thermal mass and its ability to store passive solar gain. If you think that this is bogus and you don't want to step in such a "sandbox" environment then I surely won't try to argue. I spend a few years studying and experimenting with internal mass and its correlation to a buildings function in a cold climate...as thermal mass is nothing you can easily model and there is limited research about it to be found. There is a whole lot more going on than volume to heat ratio and your basic thermal dynamics of heat flow... decrement delay and ever changing temperature differentials.
If you don't care about being able to utilize passive solar gain by all means your best bet is to omit also a slab and build a wooden floor which you insulated very well (don't even need any foam!), add either glazing with a low SHGC or exterior shading so you can cope with overheating from sun radiation ... and then you have your fast reacting efficient house, you turn the heat up and it will be warm. You turn the heat off and it will be cold fairly soon. I walk away from my house at 40 below and come back in the spring and don't worry about it.
Sand and slabs can be a very bad idea but if designed right they can also offer great benefit from my experience...and it is my firm believe that we need to look at insulated internal mass in cold climates carefully as it can help us to make our buildings function better. I will leave it at that. TC
AJ: Glad I am at least entertaining somebody - even if I am just full of sand apparently. :-)
My take from reading this is that it's too small a mass to use for long-term heat storage, but too big to use for active heating, unless your heater needs to be buffered for a few hours or days (solar and wood might qualify).
I can't see it storing more than a few hundred btu per square foot.
Thorsten,
I'm glad to know that you aren't giving up a basement room. Trust me, however -- many people have filled their basements with sand.
However, you still haven't answered my question about controlling the heat flow from the sand to the slab, nor the question about the cost of your solar thermal equipment.
If there is no Pex in the slab, I agree that it behaves just like a very thick concrete slab. But if there is Pex in there -- and that was the topic of this thread, after all -- one has to analyze the thermal benefits and losses associated with generating heat and transferring heat to the sand bed; the costs of those operations; and the question of how to control heat flow from the sand to the slab.
Kendall: The heat capacity of sand is 0.19 btu/lb-F. If I did my arithmetic correctly, that is 380 btu/ton-F. A ton is roughly 2/3 of a cubic yard. 300 tons => 114,000 btu-F. My house should lose (and Thorsten's is far better) about 360,000 btu/day at design temp (-50 F). So: Will I get enough passive solar (minus the approx 80,000 btu internal gains) to offset that loss most of the winter? That we are going to look at. That is the crux of the matter. OTHERS, PLS TAKE NOTE: I believe from Martin's above post that I did not emphasize some strategic information in my original post. See my statement "...(and your slab-heating method, if you have one)..." I should have made it clear that I will have, as Thorsten does to the best of my knowledge, PEX in the slab as well. That would be the primary conventional heating method (to fill in the gap when passive solar is not sufficient). The PEX to which I am referring is the SECONDARY unconventional heating method, which in my case would be a wood gasification boiler. (Thorsten would use solar; never a boiler). A huge issue there is justifying the expense of the boiler and pumps, as Martin has mentioned, to inefficiently heat sand. That is something I need to look at, as well as solar panels. The bottom line is: Thorsten does heat his house w/ passive solar much of the winter, whether it works or not. That house calc'd out to less than one btu/sf/hr heat loss at design temp, which is not your normal house. This concept may or may not work for my house; that remains to be determined by my lot's radiation. The radiation calcs may show that there will be too many months where my propane boiler will be running at WOT, and the sandbox will not be an option for me. Cheers. john
I am looking into slab work in my own house and I would agree that it is difficult and probably not cost effective to store heat energy for an entire winter in an average sized home. The idea of doing this in a multifamily storage unit -- see Drake Landing in Canada, http://www.dlsc.ca/energy.htm -- seems to eliminate the drawbacks .
However, the idea of using daily solar heat to refresh a pool of heat stored underground is something that seems useful to me. It would be "charged" on sunny days, released at night or as preheating for hot water; it would warm an insulated basement enough to alleviate moisture issues.
I also like the idea of adding some form of "intelligent" control to the system. I did find a system in New Zealand where they attempt full season storage for single family new construction, see here: http://www.dshh.info for details, I don't know if this system actually works!
Rich: Limited storage is all I think one can expect to get; maybe a few days worth. If you look at the heat capacity of sand, you would need it quite hot to store heat for any length of time, and who wants a slab that is 96 F till the sand cools? If I go this route, I will sure install some sensors to tell me when (if?) the slab is getting hot enough that we should shut down the sun. I'm sure there is a learning curve with it, but what is the worst you can do? Open a few windows and try again the next day?
BTW: Concrete is very similar to sand in this regard. They have about the same heat capacity (.18 and .19) and about the same weight/cubic yard (approx 3100 lbs). So, if you are thinking about thickening your slab to store more passive solar instead of using sand, that may be an expensive way to do it. Somebody pls check my info. I'd also like to know if you can dump heat into concrete faster than sand, since the sand is not (?) in contact w/ the PEX as much as concrete is.
John, Bruce Brownell has put decades into concrete verses sand. He now is using 12" concrete slabs but just to even out temps not for storing solar panel heat like Thorsten. Bruce says that concrete is much better than sand at moving heat in and out of it.
He and his ways are not well liked here at GBA.
Search GBA and the web for lots to read.
Adirondack Alternative Energy
Bruce Brownell
http://www.aaepassivesolar.com/
John, your numbers look about right (I have more numbers experience than building experience :). I was guessing at most 10F of difference for the mass, at 200 lbs./sf. .20 is usually a good number for the specific heat of masonry, so 40 btu/F-sf .
Not being familiar with passive solar design (I'm learning as I go here), one question I thought of is how much heat storage in btu/F-sf is normally considered adequate in a building? I recall reading that normal building mass should be enough, but that doesn't seem likely here.
AJ, your source is correct, concrete is somewhat more conductive than dry sand: http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html . It's apparently best with stone aggregate.
AJ: Thanks. I'll dig around, and Bruce's opinion about moving heat in and out (I assume via PEX?) sounds reasonable. For simply storage, it appears that both are quite similar. Why Bruce is using 12" slabs will be interesting to read about.
Kendall: "...normally considered adequate..." is, of course, dependent only on your structure. My spread sheet suggests I will be losing less than 2 btu/hr-sf at design temp. Thorsten's is about a half a btu. I THINK that it is simply a matter of calc'ing your heat loss for a day or two, then figuring how to store that. I'd have to stop and think about what a "normal" building weighs, but the folks who are serious about passive solar will design in plants (black soil), dark, heavy counter tops, walls of concrete, shark tanks w/ black poly on the side opposite the sun (kidding), "something" in the floor, etc. I don't think normal sheet rock and sticks is terribly important. Thanks for the engineering link; will read. j
KW: I think that link is the site where I got the specific heat info. Whilst reading there, I had this thought: If a cat could get scrap steel cheap, why not mix it in with the sand below your slab? It has a specific heat of 0.12 and weighs about 480 lbs/cf. Seems like a great medium for heat storage to me. There is a lot of scrap iron up here; very little recycling happens, so I am confident I can get tons for the taking. Food for thought.
Wow, your economics are somewhat different from the rest of the world. Any kind of metal would also give a huge gain in heat conduction. I was going to suggest a few runs of rebar if you have any concerns about conducting heat either horizontally or vertically, but I assumed it would cost too much.
Just be careful if you walk in bare feet. Metal at 70F feels quite cold, I've found.
One has to decide if your building will be lightweight or have a lot of thermal mass. A high mass building will negate the ability to benefit from thermostat setback during the evening or nonoccupied periods. There is an energy savings to be realized by nightly setback as the Delta T will be lower for that time.
KW: I don't see heat conduction as an issue; where is it going? From sand to sand, and the sand is going to be about the same temp anyway. I don't see much "traveling", until night time when the room starts to cool. I think your point is to not have, say, a 4'x8'x1" thick sheet right under the concrete; that could generate some interesting bare footing on a sunny day! I would want mass. Massive mass. The fact that steel feels cool at 70 F is exactly why I think it would be a good medium for passive storage; density. Thanks for the thoughts, and, yes, our economics is quite different, as are many things here (for better or for worse). Doug: Setback is a concept to consider, but in a highly insulated building I don't see it as a factor. A well insulated house should not change temp stink over night, esp if you have high mass to carry it. My gut is that the concept is more for houses that lose a lot of heat quickly, and/or are not relying on passive solar. That's just my gut on the matter. Good point, though. j
John,
You are right, my comment really was not directed towards you, more as a discussion point for others considering extra mass in new construction.
This is a timely topic for me as I am actually considering this on a home to be built this summer. Please comment on this... While using sand for a heat sink in an active system may not provide all of the heat needed to make a home comfortable, is there not sufficient value in the ability to provide some heat? It seems that such a system could provide a base-line temperature of several degrees above the outside air. Couldn't such a system reliably keep a house above freezing even in mid-winter, northern US/southern canada? I'd love to be able to go on a vacation for several weeks and not have to heat my house while I'm away.
Scott,
There are too many variables to answer your question. The airtightness of the home, the insulation level, and the type and orientation of the windows all play important roles in determining whether your house will freeze.
The typical new home in Vermont will not stay above freezing in December without heat for more than 4 or 5 days, especially if the weather is cold and cloudy. If your house has an average rate of air leakage, you need more than a little sand to keep your house warm.
Scott: As I see it, the crux of the matter is in how much solar you have available to begin with. If that is insufficient, then you can have all the insulation in the world and 0 ACH@50 and you'll freeze. Thorsten has a little gadget (a few $K in it) that gives him that information at the site in question. Based on that, he can then determine what will and won't work. If you are thinking about this concept, I think it is imperative to get solar data first. Then you'll need careful planning of all the other variables involved. A sand box is a big heat sink, so one needs to figure our ahead of time how it will be heated.
I am a contractor, not a scientist, but it has always seemed to me that of any fill/ material over insulation type, sand would be the last on my list. Sand, as far as I can tell, will gain and lose it's heat faster than any other fill. Deserts get cold very fast when the sun goes down and heat up fast when it comes up.
Just a thought.
Heating the sub grade takes to much energy. You need to put a reflective blanket on top of the sand. Than the plex should be at the center of concrete slab, with 6 in loops.
David H: The heat capacity of sand and concrete are 0.18 and 0.19, if I recall. I don't recall the units, but that does not matter to compare them. Their densities are almost the same, too, roughly 3000 lbs a cubic yard. So, if concrete is good, so is sand. Google "heat capacity of sand" to check my numbers, if you want, but I think I am right.
to address the original question: my thoughts are: what about placing 2 levels of PEX, one in the middle of the sand and 1 near the surface(maybe 3" from the surface). Connect the middle-PEX to the gasification boiler(and/or any other alternative heat supply) and the other PEX to propane.
Having the shallow PEX line would allow the propane source to be more responsive in providing heat to the living area and not have to "work its butt off".
The deeper you place the middle PEX , the greater the response time of that heat becoming useful ie reaching the surface. So the tradeoff in placing the PEX deeper is that you would more completely deliver that heat into the sand bank VS having that heat being available to the surface in a relatively timely way. When i say "deeper", i mean between the surface and the midpoint depth of the sand bank. Going deeper than midpoint increases the available heat-to-the-surface time which i dont think is desired in this case(?) but doesnt "more" completely heat up the sand bed.
So whatever heat energy the middle PEX is able to bank in the sand, that would be that much less work that the surface-PEX would need to do to provide the desired floor temp.
John,
As you noted, dry sand and concrete have similar specific heat (which is the number of BTUs required to raise the temperature of one pound of material one degree F), but they have different thermal conductivity. (For comparison I've also included the specs for water)
Specific Heat:
Dry Sand -- 0.19 btu/lb.F
Concrete -- 0.18
Water -- 1.0
Thermal Conductivity:
Dry Sand -- 0.15 - 0.25
Concrete -- 1.70
Water -- 0.58
http://www.engineeringtoolbox.com/specific-heat-solids-d_154.html
http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html
The net result is that while the two materials can absorb about the same amount of heat per pound of material, the concrete will both be able to absorb and discharge the heat faster than will dry sand, and it will also be able to store more heat for a given volume of material, because concrete is more dense than sand.
Density:
Dry Sand -- 100 lbs/ft^3
Concrete -- 150 lbs/ft^3
Water -- 62 lbs/ft^3
http://www.simetric.co.uk/si_materials.htm
Kack: Ahhh, the old "rest of the story". Thanks very much. Somewhere I mentioned that I need to investigate this further; this is good. I wonder if the lower conductivity of the sand would be a benefit; bigger flywheel. As I've been thinking this over, just dreaming in an impractical manner, I wonder if some large, black EPDM bags of water under a slab would be at all feasible. The size of the bags would be limited by the practical span of a 4" concrete floor, making a large honeycomb. Bag durability would of course be another issue, so maybe TC's sand is the most reasonable.
John,
I think Thorsten Chlupp's comment #3 answers some of your questions: