Lakeville, MN: Takagi 540H3-DV-P, Taco F009F5, 3/4" tubing, loses GPM, start 2.4 goes to zero in 30m

roninlakeville

I
I bought a panel online... added a manifold and takagi. Started system and it loses GPM, starts at 2.4GPM and drops to zero.

I've bypasses the manifolds to eliminate the pex... put the air eliminator in front of the pump and hung the pressure tank off it it (honeywell air elim).

So I have just the panel running... also bypassed the water flow control on the Takagi by removing the plunger in the flow meter/control.

Still starts at 2.4GPM and drops.. TACO says the pump should not do this but has no answers. Panel designer is blaming the Takagi. Takagi says all units tested before leaving the factory. I have water column right on the LP supply per manometer reading by LP coop.

Any help would be greatly appreciated as this has been a long, long haul to get this to work.

Ironman

I hate to be the barer of bad tidings, but that design is fatally flawed. An on demand water heater is just that: a water heater, not a boiler. It is NOT designed, controlled or approved for space heating.

Beside that, the two biggest problems that appear are:
1. You will need to pipe it primary/secondary due to the enormous head loss of the tankless heater.
2. It should also have the circulator "pumping away" from the point of no pressure change, which is where the expansion tank connects.

Was any kind of heat loss calculation and design work done prior to installation? What is the size, spacing and length of the loops in the floor? Is it slab or staple up?

It's sad to see that you're another victim of the Internet peddler who appeals to the desire to get something for less than its true value, but it's a weekly occurrence around here.

We're glad to try and help you as much as possible. Please provide the info requested.

roninlakeville

Thanks in advance for any help!

Circulator now pumps away from the air elim/pressure tank as it's moved to the top by the wood support.

Added primary/secondary plumbing.. tried Taco 009 primary and Taco 0011 secondary but I don't know what I'm doing as it doesn't work correctly.

Here's a pic of current; I have the pressure tank hanging off the air eliminator at the top now. All soldered up and no leaks, holds pressure. Very noisy with T-open (almost closely spaced T on the vertical).

2" insulation in the floor, 1/2" pex stapled to the ins. before pouring 4" concrete (slab) over rebar spaced 24" apart. Approx 12" spacing between tubes with tighter by doors, more space in middle of building.

235ft per loop, 12 loops.


(pic is before final soldering, installed pressure tank on air elim)

roninlakeville

Oh, forgot heat-loss info. Thought designer did the work as he asked for insulation/doors/windows... but here's the info:
50x50 pole barn, R19 walls, R60 attic, 16ft floor-to-ceiling, sliding door 12x16 with R15, overhead door 14x16 with R12+. One service door, 3 slider windows 2x3ft.

hot_rod

Where are you reading the gpm, on the boiler control? With just the boiler pump running, what type of flow rate do you get now after the piping changes.

Is it piped with all 3/4 copper?

kcopp

Is it possible that you are still getting flow but not getting any change in temp?
Tankless WH are designed to have cold water entering the unit.
You have warmer water returning...that would get the unit to ramp down the temp.

Ironman

The 3/4" piping, along with the high head of the heat exchanger, is gonna really hinder getting enough flow.

I don't know what your building's heat loss is, but if I ASSUME 25 btus per sq. ft., you would need 6.25 gpm in the secondary to get a 20* delta Tee on the floor. That's a pretty high delta T for a slab, but may be acceptable for a garage or shop.

If you wanted a more even floor temp, then a 10 - 12* delta T would be needed, but that would require about 12.5 gpm. You can't force that through 3/4" piping without far exceeding the maximum recommended velocity of 4 ft. per second.

Do you have a head loss chart for the Takagi?

I would put the 009 on the floor loops and the 0011 on the Takagi, and increase the common piping to at least 1" between the Tees because your probably gonna be getting some reverse flow between them.

Another issue: how are you going to control the water temp to the floor? A slab is high mass and it will continue to give off its heat for hours after the pump stops (the flywheel effect). If the water temp is not varied by some form of outdoor reset, the building will overheat. The on-demand can't do that, but a mod/con boiler can.

psb75

The good news: seems like you did the radiant floor install right (the stuff buried in concrete). ALL the stuff on the board and the heat producing unit--not so good. But hey, this stuff is all accessible and easy to change out. Get a boiler. Install it w/ the right sized pipe in a primary secondary config. and you can salvage this set-up.

roninlakeville

Well, last evening I found the source of the head loss. I put a Wayne 1/2HP transfer pump on the Takagi directly. Got 1.4GPM and that translates into over 100ft of head loss per the curve chart for the Wayne. On a hunch, I opened up the Takagi, took out the inlet water filter screen (it was clean), and I got 3.9GPM and it stayed there!! The inlet water screen on a takagi is very very very restrictive even when clean.

The system's been running since last evening with just the Taco 009F5 in place at 3.8 to 3.9GPM. Slab is over 60F (-15F outside).

I'm planning to replace the water filter screen with either a Y-strainer OR dirt eliminator.

When the Takagi dies (hopefully not for months/years) I'll replace with a boiler and replumb only after getting a good panel design done.