Pipe/tech help: new build panels, radiant, ODR step down

BBros

Hello all, it's been awhile since I last visited. You were of tremendous help converting my old CI rad system to home-run and ODR/TRV. System is great.

I am a GC so have some resources and knowledge, but my contractor does 99% fintube installs, so need to have some "outside" assistance.

I have a new project, new construction (parents retirement home)... have done all the heat loss calcs, sized 7 new panel rads for 140F and floor loops (7, 1/2" 220-280ft loops) building load is about 40kbtu, split evenly between floor and rads. Boiler is oil CI low mass, as that's really all we do up here. No NG and propane is not economical. Domestic is a stand alone unit.

The main floor zone (living/dining/kit) design is for 120F @ delta 10, .6gph (total pipe 1600ft, about 4gph with 6ft head). Single thermostat location.

The rads are home run, TRVs and currently planning on a Taco i-series for constant circ ODR being fed by a buffer tank/low loss header P/S piping, boiler protection, the same way I did it in my personal home... very comfortable with that system and if it were only the rads, I would be golden.

What I don't know... is best piping and control for the radiant. Obviously could just slam it with a dumb mixing valve, set temp, and pump on demand. That just doesn't sound right and would rather have it on a curve to keep floor temps stable, much like the rads.

From what I've read, step-down mixing makes some sense, given the rads should run at delta 20, it would follow that with near balanced load the return should *almost* directly feed supply for the floor. I haven't read up enough on it or how to accomplish it in practice. Or should I just eat it and get a separate smart valve. Floor sensors? School me, please.

Lastly, there is a single 280ft loop in the m.bath on it's own tstat that is designed for 110F supply. Annoying little bugger that I anticipate needing it's own pump.

Thoughts? Suggestions? I would think this is a fairly common scenario. I just need the color-by-numbers book.

Thanks for any guidance.

EdTheHeaterMan

This is a job for Bob! @hot_rod will set you in the right direction!

BBros

Thanks. Yes, he has helped in the past, I wish there were some local Bobs! I'm re-reading idronics 7 and getting some ideas... still a bit blurry to me, and the whole using the return water from panels to feed floor is totally foreign, but was mentioned in another discussion regarding efficiency.

For reference (diagram 4-11) they suggest proportional mixing valve for the radiant, fed from a hydro-link, but my "hydro-link" will be the buffer tank and set at 170ish w/o ODR... so that's not going to get me where I need to go. The boiler won't modulate, so it's going to be charging the "battery" that is buffer, with all emitters coming off that. None will have direct access to boiler temp, thus I can get long burns from the boiler, and it's easy to protect.

Secondary loop fed by 4-way smart valve (or maybe 3-way), then take off for the radiant there? As though the entire floor manifold is a sub-set of the panel manifold? Maybe... since the panels are constant circulation... and then a third time for the bath, but it would only run if the "main" floor radiant pump was running (which is quite likely, but seems like a hiccup). This becomes like the movie Inception. How to does the water ever get back!!

hot_rod

Maybe another I-valve is the best option. One concern with return from a higher temperature loop is that temperature will constantly change. On cold start you will see a wide delta cold return. As you approach set point that delta will shrink. So SWT to the radiant will change.  Also for radiant to heat, the higher temperature loop needs to be flowing. I think a mixing valve, smart or dumb is the best option,

BBros

Thanks, that's what I thought but was dreaming up more elegant/efficient ways. KISS and make them independent.

I can't see a reason to use mixing blocks, anyone have a compelling reason (other than labor costs) to use the all-in-one?

Also, utilizing 2-pipe buffer, (image from J. Siegenthaler):



Multiple distribution systems off single header tees (A&B) or should they again be independent. My worry would be some unwanted interaction if they are off a single riser (in series), but then extra tees (in parallel) makes for longer headers where trying to keep them "short & fat" ... probably doesn't make much difference. Just thinking reverse flow somehow into the radiant... I guess a check valve in the pump(s) would prevent that.