Radiant Floor heating - nonliquid?

[Rob Dekker]

Leaks in hydronic (water-based) floor heating systems used to be a big problem : In the 60's, many hydronic floorheating systems were installed using copper pipes inside concrete slabs. It took 30 years but most of these systems eventually start leaking at the pipe joints inside the concrete causing very tricky leaks. Leaks in piping in concrete is hard to locate (use helium gas detection method), you need to rip up your flooring, and jackhammer into the concrete and fix the broken joint. Very expensive leaks indeed.

Luckily, life got a lot easier with PEX (cross-linked Poly Ethylene). Apart from being probably the future for all residential plumbing in general (replacing expensive copper plumbing), PEX is almost ideal for hydronic floor heating systems.

PEX is very, very stable (corrosion free; lasts longer than your home), can handle high pressure if that may occur has flexibility but is not 'floppy', so it is easy to install and it is much cheaper than copper too. Typically hydronic systems have a continuous PEX loops, so that there are NO joints in the tubing that is under the floor. This again reduces chances of leaks, since leaks typically occur at joints.

Basically the only way to get a leak in a correctly installed PEX hydronic system is if you drill into the floor (and into the PEX tubing) or jam a nail through it. PEX is sturdy, but it cannot handle nails and drillbits.

Once you have a leak, it really depends on how accessible the tubing is if it is costly to fix or not. If the tubing is inside a poured concrete slab, then fixing it is difficult and expensive. If the tubing is 'staple-up' underneith an existing wood sub floor, then there is access to the tubing, and fix is easy. If the tubing is in a 'sandwitch' on top of a sub floor, then of course you need to remove flooring to fix it, so it is difficult.

I have installed a hydronic floor heating system (with PEX tubing) in a sandwich on top of an existing concrete slab sub floor. Installed 2 years ago. It's working great, and the PEX tubing was not a problem at all to install. Risk of leaks is only there when people start drilling into the flooring.

Pictures below

Actually, the interleaved runs are called 'counterflow' loops. Counterflow loops minimize temperature differences on the floor, by placing inlet and outlet of a loop next to each other. Multiple loops in a room are needed for pretty much any reasonable sized room, since the temperature drop limits the length of the loops to about 250 feet.

I've never heard about installing loops so that one still works if the other is leaking. If anything leaks in the system, you want to fix that before continuing. Unless you want to risk wrecking your hardwood (or whatever else you have installed on top of the floor heating).

Here are pictures of my (PEX in a sandwich configuration) system under construction. Some pictures also show counterflow loops with

Regards

Rob

[Jeff]

You can get electric radiant heating mats.

[David Lesher]

Or if it does not last as long as the venders claim it will. Obviously, we don't have 50 years of experience with it.

I know it has temperature limits. If using a boiler, better have a hi-temp monitor/safety.

Interesting; I thought they always started at the outside wall side of the room, and returned from the inside. But I didn't lay the floor, just designed the GSHP controls and installed them.

The idea was you can disconnect that one loop and use the other{s}.

The house I refer to had them cast in ?3-6"? of Gunite, with random slate as the surface.

[Rob Dekker]

True. But there is 30+ years experience with it in Europe. Also extensive durability and very stringent tests show that it is extremely stable in even very harsh environments (like inside concrete) and/or using brutally hot water.

PEX comes in various ratings, but typically can handle close-to-boiling water under pressure easily.

My system is a closed-loop system driven from the water heater, and can't get any hotter than 130 F and low pressure (12 psi PT relief valve on the loop). But even if all safety measures would fail, the my (cheap) PEX can handle 100 psi at 180 F, so I feel pretty secure that something else will blow out first.

That's often used for in-slab or slab-on-slab floor heating, where you have total freedom of putting the PEX wherever you want. The idea is that the outside of the room (where the windows are) should get the warmest, since they experience the most heat loss. With that you can even use put the PEX lines closer to each other at the windows, and further away in the center.

Using a panel sandwich system, I had more restrictions because of the pre-set groves in the panels. Also, the panels are less good at distributing heat between the PEX tubes than a slab-on-slab, so I was concerned about local heat distribution also. So to avoid local hot spots and cold spots, I used counter-flow loops. I just made sure that the loop that serves the window sides is a bit shorter than the other loops, which results in average higher temperature and higher heat throughput than the loops that serve more of the center of the room.

Cool. I did not go that far. How much piping went into the ground ? How deep did you have to go ? What were the economics calculations of such a ground-heat system versus other heat sources ? What was the water temperature requirement for the floor heating system ? How did that affect your GSHP specs ? I wonder how that system looks like. Have any pics ?

I currently use my water heater as a heat source, but are now planning to hook that up to DHW solar panels to get some heat from the sun.

OK. Makes sense, especially in emergency situations. But often a leak did already do water damage to the flooring before we find it out, so I would think that it requires fixing either way.

I think that's called a insulated sandwich slab-on-slab. Great system !

Over existing slab, put 1 inch (or more) insulation, lay out rebar and PEX tubes, and poor a couple of inches of concrete over it for the new floor. Cheap and fast to build, and excellent floor heating : relatively fast response (as opposed to putting PEX in the slab itself), energy efficient (due to the insulation) and great heat distribution (no risks of hot spots). Slate (or any tile) finish is also close to perfect for floor heating : high heat conductance, so the floor feels very comfortable.

Downsides are that the system is pretty thick (4 inch minimum) and that water temps need to be relatively high (140 F and up).

I had only 1 inch space for the floor heater plus finished flooring, or else would get in trouble with all doors and some windows. That's why I chose this panel system. It's insulated (3/4 inch thick enforced polystyrene) and topped with alu strip for heat distribution. Finished flooring is hardwood and backerboard+tile. The tiled part is most comfortable. Apart from the ultra-thin design, another advantage the panel system (with alu fins) like mine (as opposed to a slab-on-slab) is that the water temp is low. Rarely higher than 110 F, or else the floor gets uncomfortably hot. Disadvantages are that the panel system is more expensive, less 'solid', and risks of hot spots (if loop layout is not carefully designed).

[David Lesher]

I'm talking degradation, not sudden pressure failures...I'll let you find out if it's true...your tubing is easy to replace.

Well, it was that or propane; and the 10 years since have proven the wells a wise choice.. but it was lots of capital. [3 wells, ~250 ft deep.]

I don't have any pictures. I think there are 12 floor zones and

6 FCU's used for AC as well.

We installed a spare 40 gal in case of a GSHP failure or lengthy power outage. My SWAG showed it would keep the house about freezing. It's never been used; when occupied & fed, the 2 wood stoves can drive you out onto the porch in a blizzard, if fueled with pine.

The desuperheater feeds the domestic HW tempering tank.

Hard to damage concrete with a little water.

This IS slab; installed at construction. Yes, it is slow response.

No way. I have to think but I recall our upper limit is 30C. With a GSHP, it costs exponentially more to raise the loop temp.