Diy radiant...

raided

Reread your last comment. The plan seems to be: running at at least a 1.25 gpm deficit.

raided

Even running at 120F with a fairly warm returm temp, it seems like the unit is going to be a limiting factor if and when all zones are calling (water heater vs. Boiler argument comes to haunt). I wish I had realized that but again - showing my ignorance - thought that would have been accounted for during design. This is an area were expecting we could "accept" although in hindsight it's really black and white. We were spending our mental efforts on a great many aspects of this project not limited to hydronics.

What does this mean for system operation and, short of replacing this unit, what options does anyone see? I hate the idea of "good enough" but do recognize the busy season is coming for ya'll and local contractors. My goal: feel confident we can get through the winter comfortably if not perfect and then reset the clock come spring to make things better when mother nature takes over heating.

Zman

Your not going to get more than about 5 GPM through that heater with a 26-99.

raided

Understood Zman. What do you recommend for a pump?

But still, if i had a pump that could do 11.25gpm - anything over 10 - that's still going through a unit that can't handle more than 10gpm? If I am understanding correctly. I think that is potentially the main issue, even if cutting in the secondary loop.

hot_rod

So if that 11 gpm is correct and the system was designed around a 20 delta T, you have a load of 110,000 BTU/ hr. If you are heating 2600 square feet, that is an incredibly high load number, over 40 btu/ sq foot.

Or do you have a snowmelt zone in there

Do you have a load calc sheet that shows required BTU per zone, not just gpm?

Also to supply 11 gpm to those zone circulators, when all are running you would need an 1-1/4 primary loop.

What is the circ on the very left, that is not piped, for?

If you load is more around 60- 70,000 BTU/hr I think you can salvage the system with components you have and some fairly simple repiping.

IF the load is really 100,000 or larger it may be a bit of a reach.

hot_rod

It's not all that tough to do a room by room load calc, then we could get serious about a design and workable piping arrangement. A good Saturday project for you.

Here is one freebie, US Boiler also has an easy to use one at their site.
http://www.slantfin.com/slantfin-heat-loss-calculator/

raided

Thanks hot rod. No snow melt - but I wish!

Unfortunately no, no btu reqs per zone. Just gpm @120F.

The circ on the left is a future zone, was hoping to dry run a feed to the unfinished attic. I fully expected that to require additional heating source (tandem water heater, boiler, etc.) But would be a project when eeeverything else in the house was perfect. Ceilings are down, it seemed to make sense to at least provide the option. For the sake of this discussion - it can be safely ignored.

Would it make sense to ask the contractor to rip and replace with 1-1/4" while they are here for the previously stated loop? I'm concerned that we'll find out 1" is too small while wearing winter coats at the dinner table.

raided

Just saw your second post. Let me pull that up and see about getting that moving. Looks like tomorrow is dedicated to the house. As opposed to any other weekend, let's be serious. ;-)

Thanks again!

hot_rod

If you are willing, start from square 1 with the load calc.

Next determine if you have enough heat emitter (tubing) to move the heat energy from the heater to the rooms.

Then you can size pipes, pumps, and correct the piping.

Just do it one more time, with a complete fix instead of wrestling with the system for the winter.

Ironman

@raided
FYI: most radiant floors in living spaces are designed around a 10 - 12* delta T. That requires twice the gpm than what a 20* delta T would: btus = (gpm X 500) X DT.

The reason we use a 10 - 12* delta T is to prevent "striping" of the floor (too warm at the supply end and too cool at the return). If the floor is completely covered with carpet, then it's less of and issue, but if it's bare wood, tile, etc., then it will definitely be.

raided

@hot rod we can do that. 30s in the forecast so we are against the wall. Not great but we'll have to keep it moving. Going strong since May.

@Ironman
They did a 20F delta T. For the most part, we have circuit returns next to adjacent circuit supplies. About 1/4 of the house is tile - rest is carpet with mostly continuous thin metal plates throughout all of it.

Ironman

With pumps on each loop, you’re probably gonna see closer to 10* if you get the primary loop and pump sized right and get a real boiler. 😄

hot_rod

The delta T is a result of the heat energy delivered. With a cold floor you will see a high delta, maybe 20 or more. As the floor warms and the thermostat approaches shut off, the delta T will close, maybe 5 or less depending how the circ was sized, if at all!, as Bob mentioned.

Nothing at all wrong about the delta varying in a system, it indicates the system is in or approaching thermal equilibrium.

But for now, that is the least of you problems

raided

OK, so I went ahead and went through the slantfin calculation - zone by zone (great tool!). Actual zone layouts do not quite match up with rooms but I think it's accurate, no less. All added up, the tool sized us at 79,536 BTUs with an inside temp of 70, -20 outside. I am cautiously optimistic but am wondering if that's better news.

hot rod said:

Next determine if you have enough heat emitter (tubing) to move the heat energy from the heater to the rooms.

Are there similar tools to determine whether the tubing is adequate? I saw some online but they were focused on quickly determining lineal footage required across X sq. ft. The tubing installed is 1/2" O2 barrier, +- 8" spacing, approximately 6" away from rim joists.

Thanks as always.

Ironman

-20* design temp? Are you in northern ME?

raided

@Ironman no, but we are in the mountains of eastern NY. It gets cold and windy. The wind should he emphasized. The house sits in a clearing of trees and it funnels the wind at a direct impact. Too low? We have had some lousy weeks where the lows were sustained negative teens. Perhaps im designing to the exception but was thinking we should plan for the lousy and anything higher would be gravy.

Ironman

What the closest large town or city?

raided

Bennington, vt

KC_Jones

The design temp for Burlington, VT is -7 and that is 2 hours further north from you. No way is your design temp -20, that is close to the design temps for Alaska.

You could work with -7, but the lowest for all of NY is -1, which is Albany. I would suggest something between those two.

Here is a link to a chart showing design temps.
https://www2.iccsafe.org/states/virginia/plumbing/pdfs/appendix d_degree day and design temperatures.pdf

hot_rod

That is a bit low of design, although much of ND has design temperatures that -15 - 21Flow according to the chart Bob attached.

Try it with a -5 design, I'd like to see that load below 60,000 for radiant floor heat to be a good fit.

Of course the lower the load the better, adding insulated window coverings, for example helps with that low glass r-valve, or any other sealing of leakage around weatherstripping. A blower door test will clearly show where air leaks are. sometimes the state, county, or city has programs for energy analysis.

Go to www.dsireusa.org, click on NY and see what programs are available for you area.

So radiant floor can output around 20- 25 BTU/ sq ft. realistically.

So take one of the rooms on the load calc and determine the BTU/ sq.ft required. Rooms with a lot of windows, and outside exposure will be the high load areas. Occasionally extra heat needs to be added if the floor cannot cover the loa. Usually noticed on design days.

What type of floor covering, carpet and pad really reduces radiant floor output.

Ironman

Are you in the Adirondacks?

raided

Thanks guys. At -5, we are at 66,279. Better but perhaps still concerning? It's a newer house (<20 YO) so the doors and windows are pretty decent.



@Ironman - No, not in the Adirondacks - they get beat up pretty badly in the winter. We are further south around the Albany area but closer to the NY/VT border.

Alan (California Radiant) Forbes

Great topic !!! Sorry you had to be the guinea pig, raided. But we all learn so much, especially with hot rod as our guide.

Ironman

The design temp for Albany is -1*. I wouldn't use anything more than -5*. Heat loss programs have about a 15% fudge factor built in.

Also, if it reaches that temp, it's late at night and the mass of the structure would carry it through til day light when it begins to warm up some.

raided

Thanks all. So with that type of load requirement, am I ok to proceed with next steps (pump size, pipe size, etc) or am I jumping ahead?

raided

@Alan(CaliforniaRadiant)Forbes agreed that are some *very* excellent folks on this forum and i am extremely grateful for the advice. I wouldn't term myself a guinea pig so much as another sucker. And I suspect i will not be the last.

hot_rod

Decide how you want to make the piping correction, as shown back a page, either a primary secondary, or a hydraulic separator. You should be able to use most of the parts you have if the plumber can, or will disassemble and reuse fittings.

raided

@hot rod I think the first pic is the least intrusive assuming a 1" zone manifold is sufficient. What still escapes me - and perhaps im still misunderstanding - is whether im still limited by gpm through the unit (11.25 vs. 10 max). It seems like the btus are plenty available but the flow may be restricted to actually deliver.

Thanks for the guidance and patience.

hot_rod

I'm not sure where the 11 gpm comes from? If the load is 65,000 at 20 ∆ t design that requires 6.5 gpm. I doubt you will get 11 gpm thru that heater, 6-7 is doable, on speed 1, shown on the graph below..

Is this the boiler pump you have, a 3 speed 26-99 Grundfos?

You can plot that Pressure Loss curve on top of the pump curves (downloaded at Supplyhouse.com) and determine the OP operating point to predict exactly what the pump will move in that loop.

Read below to learn how to develop system curves and how they interact with pump curves.

The boiler circulator (26-99) moves heat energy from the heater into the loop via the closely spaced tees, period.

The zone pumps are sized to the loops they are connected to, not sure if or how those were calculated? The various circulators are working independently form one another in a true primary secondary piping arrangement, that is what is missing in your current piping.


https://www.caleffi.com/sites/default/files/coll_attach_file/idronics_16_na_0.pdf

raided

@hot rod yes, 26-99. By "true" p/s do you mean that each pump has it's own set of closely spaced tees, vs. The boilerpump being hydraulically separated with closely spaced tees with the zone pumps being on the other side of them?

The zone pumps were sized by the radiant company based on the anticipated number of feet of tubing. Might be some recalculation on my part. Originally I did the head loss calcs and applied them to the charts to come up with the same pumps as the company had. They were a bit larger but the next step down was going to be undersized. I was also told that pumps need to be about the same size because larger pumps would steal from smaller. In hind sight that sound more like a design flaw from oversimplification than anything else.

hot_rod

P/S is a confusing concept. You only have P/S if you have closely spaced tees somewhere in the piping.

However you want to get from the upper drawing, current system, to the lower is up to your plumber.

The boiler wants to have it's own loop and circulator, flow comes out goes around the tees and back to the boiler.

When zone circs turn on they take a portion of the flow to their zone. Unusual, sometimes two directional flow is going on in that set of tees.

If you want to keep all the current piping on the wall, close that ball valve, remove the handle and toss it, connect the other end near the boiler together so you have a horseshoe looking loop. The air separator is fine where it is.

So visualize two loops, the boiler loop in and out via the 26-99, the zone loop(s) from the "horseshoe out to the zone, return to the lower leg of the horseshoe.

Put a set of tees into the horseshoe, the 26-99 pumps INTO the boiler from the tees.

The horseshoe is named after Harvey Youker's Horseshoe, the first installer I saw that simplified the P/S piping concept.

Of course the 26-99 needs to run whenever any zone pump calls on, a relay board makes this wiring simple, since that water heater probably doesn't have a wiring connection for the boiler pump.

Get a copy of Primary Secondary Made Easy from the bookstore on this site if you or the plumber wants to fully understand the dynamics of primary secondary piping.

The thinking is all done, start wrenching

raided

Thank you, hot rod! This is extremely helpful and p/s is finally sinking in. As an aside, I had asked the original radiant company multiple times on the subject and they found it entirely unneccessary. And so i abandoned the thought. Shame on me.

One final question - hoping :-). Would it make sense to move the expansion tank under the air eliminator (im mot sure if that square represents the expansion tank)? Also, I *think* the radiant contractor proposed the system pumping on the supply side, away from the boiler - no good? I'm pretty sure ive read in numerous articles that you pump into it as you suggested.

Ok, that was two final questions. Let's say "final" questions.

Sincerely - thank you.

Ironman

Not to add any more confusion, but if you're not trying to make each secondary pump a separately controlled zone; and if your loop lengths are all <300', then one 15-58 pump could do all the secondaries.

You'd be back to my original drawing, with or without using zone valves.

If you're trying to zone each loop, then run the secondary pumps on low.

Ironman

raided

We are looking to have each zone independently controlled. I think your design has validity and would probably be more efficient since there would be less wattage getting eaten. That may be something to revisit but i think i have to stick a pin in too much redesign. As it is, we have rim joists to insulate, foil, and insulation across 2 floors. And thats only some of it. Thank you very much for the insight - we may revisit this design (maybe one big pump and zone valves?) During the winter to redesign come spring.

Gordy

@hot rod nailed it. In present configuration the 26-99, and 15-58 are in series when a zone calls. First zone called in the line up gets all the flow any others that call at the same time get robbed.

The whole premise of P/S is to be able to have separate boiler, and system pumps operate at differing flow rates, and not conflict with each other.

As @Ironman said, You could do away with all the zone pumps, and have zone valves with one 15-58 on the system side, and the 26-99 on the tankless side.

hot_rod

raided said:

Thank you, hot rod! This is extremely helpful and p/s is finally sinking in. As an aside, I had asked the original radiant company multiple times on the subject and they found it entirely unneccessary. And so i abandoned the thought. Shame on me.



One final question - hoping :-). Would it make sense to move the expansion tank under the air eliminator (im mot sure if that square represents the expansion tank)? Also, I *think* the radiant contractor proposed the system pumping on the supply side, away from the boiler - no good? I'm pretty sure ive read in numerous articles that you pump into it as you suggested.



Ok, that was two final questions. Let's say "final" questions.



Sincerely - thank you.

Pump into the boiler so all that head the 26-99 ads to the system shows up in the boiler. Remember we are trying to maximize a mis-applied water heated.

The best tank location would be at the closed end of the horseshoe, this has all the circulators "pumping away".

One system circulator and 5 zone valves would have been a cleaner system, but you already have all those 15-58s.

Ironman

raided said:

We are looking to have each zone independently controlled. I think your design has validity and would probably be more efficient since there would be less wattage getting eaten. That may be something to revisit but i think i have to stick a pin in too much redesign. As it is, we have rim joists to insulate, foil, and insulation across 2 floors. And thats only some of it. Thank you very much for the insight - we may revisit this design (maybe one big pump and zone valves?) During the winter to redesign come spring.

No, one small secondary pump (15-58) with zone valves could do all the zones. Be careful about too much zoning; it's gonna cause short cycling and seldom needed or benificial with radiant.

You'd still need the 26-99 to pump the water heater due to its high resistance (head loss).

Gordy

Not to add any confusion. You could find out the pressure drop of the hx by installing a port on the suction side, and discharge side of the 26-99. Connect the two ports with 1/4” copper tubing into a T with a 0-15 psi pressure gauge with a valve on each side.

By closing the valve on the discharge side and opening the valve on the inlet side take the psi reading. Then close the valve on the inlet side, and open the valve on the discharge side then take that psi reading.

Take the difference of the two readings and convert to head by multiplying 2.31.

Edit: it appears the last part of my post was lost.

You would then take that calculated feet of head, and use the 26-99 pump curve to find the gpm flow rate. If the 26-99 is a three speed you can perform this for each speed, and find the one that gives the required “total gpm” for all zones together.

The ability to adjust flow seems pertinent IF the calculations for the flow rates in each zone have that much difference. If all zones are with in +/- 10% in tube length they will all see the same flow rate. Not a huge deal, but inefficient as some zones will be overpumping for the load.

A system with out proper calculations in heat loads to determine components sizing, flow rates, and tube centers is a WAG at best.