Best Of

Existing boiler set up is a 1/25 hp Grundfos pump on primary loops and the boiler (80k btu, hangs on wall) is on a secondary loop. A T-branch runs a few feet to boiler and back to primary loop. The boiler has a small circulator inside the cabinet. The question is: if I replace this boiler with a boiler that does NOT have a circulator on this secondary loop, should I expect any problems getting heated water back to the primary loop? I'm planning on getting one with more btu per hour, approx 100k btu. Thank you so much!

Our new home has a polished concrete floor with radiant tubing. The walls are ICF so the foam provides a thermal break and expansion joint at the perimeter. The slab was poured on 3" of 25psi EPS foam for a thermal break to the ground below since we are in a northern climate. There is a 10 or 15 mil vapor barrier under the foam that was glued to the ICF perimeter wall and taped around all penetrations.

We debated placing saw-cut control joints and placed the PEX as low as possible to allow this option. The tooled control joints filled with polyurethane at our local WalMart look pretty bad so we took a chance and didn't cut any control joints. The floor is about 80' x 35' with 9 embedded column bases. It was in the late fall so we poured concrete mixed with hot water, I think it was 3500psi, no fiber mesh. I can't remember if they added any plasticizer. It stayed warm and very humid inside as the concrete cured. We had some minor cracking but only where the slab was restrained. For example, at the doors there was rebar from the footings folded back into the slab along with rebar out to the thermally broken threshold so when the concrete shrank we got some small cracks at the corner of a couple of the doors. There are also two doors where the concrete continues into the garage and into a mechanical room. We have some cracks there where the slab was continuous. It would have been better to place a control joint in those doorways that would have been covered by the door threshold. Around the perimeter there are gaps up to 1/4" or more in some places due to concrete shrinkage but these will be covered by drywall and baseboards.

Our initial plan was to use a mix with a larger black aggregate. This stone was crushed so they wanted a pump truck with a 3" hose. At the last minute we had to work with a smaller pump and they were afraid the larger crushed stone would jam it so we went with standard washed round stone, probably 3/4" if I recall. There is also a fair amount of smaller stone in the mix.

They hand floated pads at the screed depth and then used a vibrating power screed. We had hoped so see more of the large rock exposed after polishing but the vibra screed pushed the rock too far down. It would have required a lot more polishing to expose it. As a result, there are areas with good rock exposure and other areas that are more of a uniform grey. We also have some 'interesting' bonus features caused by small twigs and other items that were blown into the sand and rock used to mix the concrete.

The PEX was tied to a 2' grid of rebar. The rebar was placed on custom made 1" dobies to set it as low on the EPS foam as possible. The PEX was buried in the under slab foam from the manifold to the desired room in some cases. A 2" deep groove was cut with a hot knife, PEX inserted, the removed foam glued back in the groove above the PEX with spray foam and then trimmed flush. This same method was used to run 3/4" PEX to a central manifold. A piece of 3/4" PEX was slit and snapped over the 1/2" PEX in areas subject to damage. Buried PEX was brought up out of the foam in a piece of blue flexible conduit (smurf tubing). The entire system was pressurized to around 60psi before and during the concrete pour to make sure any damage was found and fixed.

Overall we are pleased with the floor even though it is not perfect.

Since the house is very thermally efficient it will not need much heat. The water to the floor will only need to be about 95F on our coldest design day of around 5F. The water temp is set by an LK outdoor reset on a mixing valve which seems to be working very well. In order to increase perceived comfort, the PEX was not placed on a continuous grid in most rooms. Instead, less PEX was placed in locations where people would most likely stand or be seated. This allows a higher water temperature in those localized areas so the concrete will hopefully be warm enough to feel the heat. I found a couple of articles online describing this approach. Here's a FLIR image taken as the system was initially warming up.

We polished the floor before any of the interior framing was built. This lowered the cost of the polishing but requires extra care during the rest of the build. The densified concrete seems quite durable but as @DCContrarian pointed out, you can damage it and any repairs will be difficult and filled in cracks will show.