Yes, that would be my recommendation. No, larger tube is more for achieving larger flow rates with less head and less required circulator capacity when you need to move a lot of heat. 1/2" PEX should be more than adequate for your project. Your HR emitter supply temperature will solely be a function of how well your building and your HR emitter perform. Your maximum required HR emitter surface temperature is already locked down and is a direct consequence of how well your building performs given your ACCA Manual J8 heat loss analysis. The only question now is how efficient your HR emitter will be in getting the heat from the PEX to the surface of your floor.

Accomplishing HR via a thin gypcrete slab is usually way cheaper than Warmboard or any other above-floor plate system. Of course, material and labor costs can vary greatly based on locale. I would estimate that Warmboard and a thin gypcrete slab perform about equally well. I don’t have any first-hand experience with Dan’s sandwich system performance, but several folks have reported on this forum that they were happy with it.

We likely do about 90% concrete slab, 5% gypcrete slab and 5% Warmboard. One can usually have a concrete slab HR heating system for less cost than any other heating system option. While HR heating is unquestionably the most comfortable heating system approach bar none (heat at right location, simple, quiet, no drafts), we nevertheless tend to steer folks to mini splits whenever a HR concrete slab isn't an option. However, that's just our mindset/recommendation given that mini splits will significantly out perform any HR heating system that doesn't use a heat pump based heat source. Heat pump based heat sources can get very expensive and can have expensive maintenance issues too. So the lower acquisition cost and lower operational cost of mini splits tend to win that ROI decision for us.

Posted By sailawayrb on 10 Nov 2017 09:48 PM
Accomplishing HR via a thin gypcrete slab is usually way cheaper than Warmboard or any other above-floor plate system.

hmmm maybe I should look into that again...? Does that add 10 psf to dead load for the floor?

Figure about 15 psf of additional floor dead load for a 1 1/2” thick gypsum slab.

Roth Panels are cheap. A bit over $3 sqf. more responsive than a slab.

You don’t hear tube expansion noise in gypsum job and it heats good. Any dry system has more noise than gypsum job and doesn't heat as good. Warmboard is best dry system for new construction where it is the subfloor too. It holds tube tight for better heat transfer and less noise than other dry systems. It responds faster too. Roth is good for over concrete but hard to install wood over it. It doesn’t heat as good as Warmboard either. Look at Raupanel too if you don’t care about higher water temps and noise. Sometimes you get what you pay for. Sometimes you don't.

What about the efficiency drop of staple up with aluminum plates? Are we talking 5% drop? Or like 25%?

According to Siegenthaler, the heat transfer coefficient of a concrete slab HR emitter is 0.882 and the heat transfer coefficient of a below-floor HR emitter is 0.36. That's a difference factor of nearly 2.5. These values are for a 12" PEX tube spacing and a floor finish R-value of 0. As you increase the floor finish R-value, the heat transfer coefficient decreases significantly. This is why we like concrete slab HR emitters and recommend not covering them with wood or carpet...

Yes I know, but I'm trying to compare a staple up system v. A floor sandwich system. House is half built and I can't replace the first floor with slab on grade

According to Siegenthaler, the heat transfer coefficient of an above-floor HR emitter is 0.55. That's a difference factor of 1.53. I suspect, but I don't have actual data, that Warmboard (and perhaps even Dan's sandwich system) performs nearly as good as a thin-slab HR emitter which, according to Sigenthaler, has a heat transfer coefficient of 0.767 for a difference factor of 2.1.

OK, I think I understand now. Thanks.

Good deal and happy to help.