I think I am with the right group of guys here to get this figured out. Working on a new building design that will have my light commercial and residential in the same building. The residential side will have a two stage climate master but the whole building is getting radiant tubing. floor will be 6" but may reduce to 4" in the residential area. I have read some of the 'spirited' debates over tube placement. I am getting ready to warm up my calcs to run numbers for tube size, length, etc. However, I would like to gather some good general engineering points here before I do. I need to quantify the performance by putting the tube at the bottom of my 6" slab rather than lift it. I know technically there is efficiency to be gained but you really have to consider everything. I need to be able to drill in my concrete, among other things. I also recently visited a shop with radiant and could easily tell where the tubes were with my hand. I did not like that. They also had some breaks in the concrete too! Not sure if that is related yet. Building is 60x100 and heat source is still being considered but likely to be wood fired system (country). I know it will take some lumber though and I have plans to later consider a geo heat pump system with alt supply power but that is down the road. The design is changing a bit but max heat loss is about 150K btu. I have not yet firmly decided how to crunch the numbers for radiant but I think what I was doing was assuming U/R values for an insulated slab for the heat loss, but I never really crunched the higher losses once I apply heat to the slab. That will start dumping btus into the soil through the insulation. My plans were 1.5-2.0" XPS over a vapor barrier. Attach tubing network to foam, then sand bed all the tubing lightly to help protect it, then I have to decide on my floor reinforcement system. The floor is going to see some weight from machinery and other equipment. I don't want problems and want to minimize cracks. I am still working numbers for concrete stiffness and deflections into the foam. I know subgrade will be a big deal. For my radiant tube, I am open to 1/2" or 5/8" but I am trying to stay economical here! I am also trying to look at where max efficiency will be in respect to supply temps, flowrate, etc. One thing I like about radiant is I can tune a lot of that on the fly. I was hoping with my tube at the bottom, I might have a slower response but my floor temp would be even, and I could possibly use a more thermally conductive media but that could cause more thermal losses through the insulation too. In all, i realize I might see 'more' loss with the tubing at the bottom but I am trying to quantify that loss. From what I gather around here, tubing damage during pour is common. I don't want damage. I am also thinking about things like rodent and bug damage to the pipe but I guess PE has been in the ground doing other things for decades without much issue. EDIT: I should point out that the entire concrete floor will be exposed. The shop area will be epoxy coated, and the residential area will be stained. There will be a few rugs thrown, but no floor coverings.

I am doing a radiant project now (my own house) and we are going with all 1/2" pex. That worked out well since it seems like all the accessories readily available (example: foam board staples) are for 1/2" pex. I used EPS foam, much lower cost than XPS, and poured concrete right over it. The concrete has fiber reinforcement so no rebar. You'll need to get a good analysis of your floor strength if you are concerned about it and having machinery. This is not something you want to do over. About the heat loss and BTU output: you can always raise your radiant water temperature, but when you do that with geothermal you are losing efficiency. For wood fired heat it may not matter if you want to run 95F water or 150F water. I think the pex is good for 180F. Good luck.

There shouldn’t be any guess-work or crunching of numbers to get the hydronic radiant design perfect. You just need to accomplish a good heat loss analysis and a good hydronic radiant design analysis. All of this can be easily accomplished by using software yourself or by hiring someone to do it for you. We have free DIY software on our website if you are so inclined:

Borst Heat Loss Analysis Software

Borst Hydronic Radiant Floor Heating Design Software

You will have to take the time to read the associated detailed software instructions and fully digest what is required to properly complete the analyses, but many DIYer have successfully accomplished this. The maximum recommended PEX spacing for residential floors is 12” and the reason is precisely for what you experienced first-hand. Placing the PEX on the bottom of a 6” slab will cost you a couple percent of hydronic emitter effectiveness and this is fully explained in the instructions too. You should consider getting some professional engineering assistance to sort out the structural loading aspects.

You need a comprehensive design. There are many ways to address your various concerns without going to extreme or questions practices.

First, a proper ACCA Manual "J" heat load performed on purpose-built software.

A proper design will provide.

Total heat load, your estimate is high unless you reside in Fairbanks.

PEX layout including the diameter, length, spacing. This information will determine the design water temperature.

CAD drawings to facilitate installation of the PEX tubing and future floor-fastened equipment.

Design water temperature, pump size and Delta T.

Slab insulation, R-value, compressive strength and specific project suitability.

Specification of controls and mixing valves for high to low temperature heat source such as wood-fired boilers.

Back up or duel-fuel designs.

Sand has no place in a radiant heating project.

"Feeling the tubing" is known as striping and is related to PEX depth along with floor covering, operating temperature, heat loads, PEX spacing, and heat transfer fluid velocity.

A six inch slab is extreme, and generally reserved for commercial, heavy equipment garages. Our snow melted driveways are done with 5000 psi fiber reinforced concrete with wire over PEX to single the relief cut saw operator, should he get too close to my PEX.

Response time is a factor of the load, local climate and the heat source. You can effect two of the three with the proper tube spacing and water temperature control, with out door reset. PEX placement in the slab is the least of these but can effect the response time depending on the other two factors.

Thanks for the responses guys! I will certainly review the recommended software. I am in the engineering business so I tend to want to know how and why things are approached and work the numbers myself. I have run complete heat loss/gain calcs for the building but the heat loss I am estimating would be max heat loss on an extreme day. Normal losses would be closer to 50k btu. I think I am accounting for most things, including thermal bridging effects within my framing.   I assume I need to account for the concrete mass in this though to determine how many btus of stored energy I have.

I will share some plan details to see how you guys might approach this. building is 60x100x17ft sidewall, 3:12 pitch. R-19 walls, R25 ceiling, extreme details on air seal, 1 12x14 OH door, 3 walk doors. The floor will be 1.5-2.0 foam so I am figuring R3.9/in or 5.8-7.8. My worst case heat loss is a 60* delta T, but would normally be closer to a 30* delta T.

One thing I noticed is just based on pressure losses in 1/2" PEX and some recommendations of approx .25gpm/100', and a quick estimate of 10* dT across the tubing, that would take a LOT of tubing. Much more than tubing just on 12" centers! I hope I am missing something here. My hopes in moving the tubing down is to reduce striping effects which would not make tubing centers as sensitive, and focus more on just getting the BTUs to the slab.

Having thought about this, I think I will have to abandon the sand idea. The nice thing with tubing in the concrete is full pipe support and protection so over pressure/temp should show up outside the floor before failing in the floor. I also like the fact that no varmint can get to it.


This brings me to another, and probably most important part and least understood, even by pros. The soil interactions and capacity. I was shocked to read that a pipe and boiler mfg is recommending only side wall insulation down to a certain depth and forgo under slab. The idea there is to increase the sink capacity thus allowing things like solar and wood fire to be pumped in when practical, but have plenty of carry over. That seems like a very costly mistake if something is missed!

Insulation and tube spacing are dictated by the design temperatures.

It is difficult, and unnecessarily time consuming, to try to figure all this out long-hand. You can fool around with a free trial version of LoopCad or Wrigthsoft--the one I have been using since 2006--but know how to do something and what to do are not the same as you well know.

12" O.C. is used for occupied residential building. I have a single loop running around the garage area of my shop and it meets the heat load during design conditions that represent the coldest week or so of an average winter in your area. Not the extreme mean as you suggest.

http://www2.iccsafe.org/states/virginia/plumbing/PDFs/Appendix%20D_Degree%20Day%20and%20Design%20Temperatures.pdf

As for not using perimeter insulation and neglecting the field of the slab entirely. This is done in mild areas with high ground temperatures and low inside design temperatures. Again, a good dedicated radiant heating software will allow you to model practically any scenario you please.

And don't forget to review Modern Hydronic Heating...

OK, I apologize for how my brain works but I have to have a firm foundation so I can understand. I tried the Borst heat loss calcs and none to my surprise, my calcs are inline and totally dependent on two factors.... The heat loss in the slab(huge), and whatever delta T I decide to put in there for air temps. I know there is more interaction there and I need to run more numbers on the slab because you cannot use a blanket delta T for air temp when the slab is not seeing any of that. Even when we get to a 0*F day here, that ground usually will not drop below about 20* on the coldest winter and usually will stay above 30*. Determining the actual temp gradients in the soil is the tricky part.

I am anywhere between 80-300K btu heat loss depending on how I want to approach it. However, I still need to consider the mass effects of the slab.


Now, I did look at the Borst radiant calcs and cannot even tell what the units are in so I just cannot rely on numbers when I don't even know how they are applied. I don't think there is any reason to start applying tube size, spacing, etc until we know certain, basic variables such as the dT across the tubing, the conduction of the materials to determine what dT is even possible, and then how we might get those btus to the slab.

When I am looking at basic numbers, I said something generally like, "ok, lets say I need 150k btu/hr pumped inot that floor, that is 2500btu/min, there is 1 btu/lb h2o/*F or 8.34 btu/gal/*F. If I target around 10*F dT (assuming conductivity works out) that works out to 2500/8.34/10 or 30gpm". You can then determine many more variables such as determine pump/tube sizing, pumping losses, etc, etc. That is all easy jazz to figure out after the fact but the basics are determining btu input to the slab and how the slab, insulation, and soil react to this.

I see guys pile up insulation under slabs around here but I think some things are being missed. They are assuming just because it gets cold here a few days, that the soil is that temp. Even before I pump heat to the slab and assuming a rounded R4/in for the foam and 1" of foam, that soil may only be 40* because the top is not exposed to the cold. That means it's heat is trying to leave outward towards the colder surface areas around the slab. However, that won't happen in an hour, or even a day. So just for general discussion and generalization, that temp under the slab insulation is maybe 40* and my target floor temp is 60*, so now we have only a 20*F dT. However, if I want btus to flow into the floor, I have to have a dT across the tubing/slab interface. At the lower part of the slab where the tubes are, that slab temp will come up, lets say to 80*F. Now I have a 40* dT at the insulation but, those btus are not being drug away like air would do. That heat is further heating the soil under the insulation. ------I am still trying to determine and understand this geological interaction------ But, I suppose I can use general data such as specific heat and conductivity of soils to better predict this, but it is my my half informed opinion that the insulation interactions under the slab must be CAREFULLY calculated, NOT just applied uniformly to the heat loss calcs such as mine and Borst are doing.