radiant floor heat in garage

tylerkeck

i have a 4 loop 1/2 pex that goes to a manifold, it has a alpha 2 pump and a takagi T-H3M-DV-P Indoor Tankless High Efficiency Condensing Water Heater. my room temp is 53 my inlet temp is 140 return temp of 80. i used 100 gallons of propane in 2 weeks, the pump on high shows a gpm of 2 and pressure of 12 psi..not sure why im using so much fuel. if i shut down 2 zones i can get 1gpm per zone but with all 4 open they show like .4per zone.. any help would be appreciated. building is 1400 sq ft with 12 foot ceilings

Ironman

That's the wrong appliance for space heating! You need a boiler, not a tankless water heater.

The water heater is not designed, controlled or approved for space heating. Its heat exchanger has way to much resistance to flow for a normal hydronic circulator to overcome.

Did an Internet peddler convince you to buy this setup?

tylerkeck

no i had a oil boiler and it was using 250 gallons of oil a month to heat the garage so the local service company told me to get this and it would be a lot more efficient.

Ironman

Sorry for your situation, but this scenario gets posted on here almost weekly.

The willful lack of knowledge in this trade is amazing.

You could try piping it p/s and add a large circulator like a ups26-99 on the heater side, but it still won't be right. It would improve it to some degree, but not cure it. And, it probably won't last more than 3-5 years.

You need a boiler.

Is there any insulation under the slab?

tylerkeck

yes 2 inch foam board around the edges also. 6 inchs of concrete. i ordered the pump you suggested hopefully it helps. sorry for the ignorance im so tired of trying to make this work and everyone i call to look at it has another idea for me to spend money. what is p/s

Brewbeer

P/S is primary / secondary, which is a near boiler piping method designed to completely separate flow through the boiler from flow through the radiant tubing.

Ironman

Primary/secondary:



The tankless would connect to the side (bull) of the Tees with the ups26-99 pumping towards the boiler.

The floor would pump from left to right through the run of the Tees. It would be a good idea to increase the Tees one pipe size larger than the largest pipe..



In this diagram, the flow through the primary (main loop) is from right to left.

kcopp

How tight is the envelope?

tylerkeck

Ok while searching for what p/s stood for i cam across this info for takagi tankless heaters. Figured i would ask if its correct before i do it
"Takagi on-demand water heaters, for example, do not require Primary/Secondary plumbing. Nevertheless, it would be a huge mistake to run even a modest radiant floor system directly off the water heater.

Again, it comes back to flow. The inlet and outlet ports of on-demand units are tiny and that restriction can “choke” the flow rate needed by the radiant floor. For this reason, a properly sized three-way mixing valve is installed between the on-demand heater and the radiant floor Zone Manifold. With this configuration, the radiant circulator pump(s) draw from the “mix” port, a combination of the “hot” port of the mixing valve (the line from the heater) and the “cold” port of the mixing valve (a tee off the line returning from the floor to the heater) solving the volume/flow problem. Tiny ports no longer matter because only a small amount of heated water is required from the on-demand heater — for the simple reason that the floor water returns to the heater only ten degrees cooler than when it left.

In other words: The water going to the radiant pumps is mostly the very same water that just returned from the floor — with a small injection of hot water from the on-demand heater"

tylerkeck

Its a pole barn r38 in the ceiling and 19 in the walls with 5 mill vapor barrier plywood walls, insulated garage doors. If i heat it up with a torpedo heater to 70 it will not drop a degree for 2 hours.

Ironman

Thats hogwash! On a slab, the pump would be trying to pull everything from the heater until it reached temp. Sometimes, that could be 12 hours or more.
They'll come up with anything to try and sell their product.

Gordy

For that model the only drawing piped direct for a space heating application is for a fan coil..

tylerkeck

So it needs set up as a p/s ?

Ironman

Yes.

tylerkeck

This is my current set up..would it just be 1 zone..one pumping into the heater and 1 pumping water to the manifold?

tylerkeck

Manifold

Ironman

Like this:



Keep the components in that order and follow the spacing dimensions for the Tees in the previous diagram.

tylerkeck

Thanks!

tylerkeck

so i built this system and heating for 7 months in the winter in pa it cost me $1500 roughly 657 gallons to heat a 1400 sqft garage , if i got a boiler would it cut my costs from a tankless water heater?

Solid_Fuel_Man

That question can only be answered accurately if you have a combustion test done on your current setup. That will tell the real story on how efficient it is. Then the collective wisdom can guide you in what direction you should go.

You did say there is 2" of rigid foam under the entire slab including the edges correct?

hot_rod

657 gallons to heat 210 days? A bit over 3 gallons a day?

Zman

What supply temps are you running? Outdoor reset? You could be wasting energy overheating your slab.

newagedawn

1400sq ft 12 foot ceiling, did someone do a heatload before the install? sounds like someone didnt do their homework, please do a heatload, because if you just pick a boiler and install it the same thing is going to happen, and you'll blame the good folks here for your lack of diligence

tylerkeck

Yes 2 inches rigid foam under and around slab
Usualy pushed 140 and returns 80..if its running for a while it returns 110 but it usualy only runs for 10 mins at a time. No outdoor reset the winter was very cold..many weeks at 0 and garage was 65

tylerkeck

Set up

STEVEusaPA

What are your loop lengths?

tylerkeck said:

...it usualy only runs for 10 mins at a time...

Sure sign of design/equipment sizing problems.
As mentioned, you really need to figure out the design on paper, and adjust the flow (circ selection) and temps based on what you have for loop/length. 1400 sqft and. 4 loops sounds like some pretty long loop lengths.

tylerkeck

They are 200ft. It was designed a 5loop but the one in the middle failed ..the loops were all designed on a program and laid out

GroundUp

Why are you supplying 140 to a slab? Turn that thing down to 100 to tighten the delta T and let it run for hours instead of minutes. Constant short cycling isn't doing your LP usage any favors. Design is poor regardless, either with too long of loops or not enough tubing. That same unit is rebadged and sold by HydroSmart in Menards stores as a "boiler". I've installed a few customer supplied units against better judgement and they do work just fine without P/S piping when set up properly otherwise- until the inspector finds out. No, a tankless is not ever a viable substitute for a boiler BUT it will work if you turn the temps down. Is your Alpha on AutoAdapt or what setting? The head loss through the unit is minimal, and that pump should easily push .7 GPM through four 300ft loops assuming you don't have a mile of 1/2" for the system piping

tylerkeck

Tubing is 1/2 inch spaced 12 inches..it was all laid out on a heating program and followed. I had a failure of the middle run but with the p/s piping it pushes 140 but inlet temp is around 110 at manifold once it runs for a little bit..but onces it gets 110 out and 80 back it turns off and has to do it all again

hot_rod

So you ended up with 800' of tube in a 1400' slab? Averages to about a 20" on center spacing.

What is the heat load number? With that and the actual tube amount you can come up with some better answers.

With only 4 loops you may not have enough heat emitter for the boiler output, hence the short cycling.

You mentioned 140supply, 80return and then 110 supply 80 return? The delta T right at the manifold is the key number.

GroundUp

So if spacing is correct, which it appears to be in the photo, loop lengths would be ~280ft on a 1400ft slab, but one of the loops has been abandoned. Are you sure it's 200? I wasn't aware you had changed to P/S. What kind of thermostat are you running to control the system? A PWM stat will cause some short cycling when the temps are too high, as yours are. As hot rod says, the numbers at the manifolds are what are important. That unit will modulate down to 19,500 BTU, so it's unlikely that the water heater is overpowering the load causing the short cycling unless modulation is locked out. What size pump do you have on the boiler loop? 1.6 GPM (if you have .4 per loop) and a 30 degree delta t (110-80) is 24,000 BTU, which should allow the unit to run much longer than 10 minutes especially with a wall stat. Either way, your temps are still too high. Turn it down to maybe 110 on the water heater and see what it does.

hot_rod

If the boiler cycles off and the room is not maintaining, you either need more flow or higher temperature to cover the load condition.

Could be you are not getting adequate flow through the boiler, did you check if any screens are plugged? Or the piping and pumping is not moving the flow.

IF you had 5- loops covering 1400 square feet 12" on center there would be at least 280' loops, plus some leader length. I'd guess you have 300 foot loops?

So now you are left with 1200' of tube in 1400 square feet of space, correct.

The accurate load number would tell if you have enough heat emitter to carry a design day load.

.6 gpm per loop is about right for 300'- 1/2 pex loops. So in essence you have about 15" on center with one loop abandon.

.6 gpm, 300' loops, 15" OC should get you about 15 BTU/ sq. ft output from the slab. Just some quick number crunching with the RPA RadPad.

Your design should have more detail like total load at design condition, supply temperature required, tube spacing and loop length, pump sizing, etc.

So two issues, high fuel and inadequate heat?

The heating design you mentioned if what we need to see, then possibly the answers will be clearer.

tylerkeck

No im not sure its 200..i bought 300 foot loops and had some left over from each run but cant remember how much
I bought the uponor heat only thermostat for hydronic floors.
The pump is Grundfos UPS15-58FC
Im at .6 gpm through each run
I will turn the heat down thanks

hot_rod

Ideally with everything sized and installed properly, on a design day the boiler should run non stop, inputing just what the space needs to maintain the desired and designed for temperature. if some of the tube is missing you may still be able to get adequate output by increasing flow and or supply temperature.

I'm not sure 567 gallons of fuel for 7 months, 210 days of heating is all that unusual. 567 divided by 210 days is 2.7 gallons per day?
What temperature does the system maintain on a 0 degree day, for example?
If it was designed for 65° indoor 0° outdoor and you can only maintain 55 or 60, it could be related to the abandoned loop, you are running on 4/5 of the designed system distribution.

If you paid $1500 for that 567 gallons that is $2.64 per gallon X 2.7 gallons per day= 7 bucks a day to heat the shop?
Granted they were probably not all design condition days. And it depends on how the space is used, overhead doors opening and closing etc.

tylerkeck

No the tubes are spaced 12 inches..its close to 300 per run i dont remeber the exact number. But i bought 300 foot coils and had a small amount left

When the building was built they did a heat loss calculation and said it needed 77k btu

Its 110 and 80 at manifold bit as it runs longer the numbers get closer

hot_rod

That load number seems pretty high? 77,000 divided by 1400 square feet would be 55 btu/ square foot required from that slab.

High 20- low 30 but/sq. ft is usually about what you can expect from an indoor slab while maintaining 70° indoor temperature..

Expect about 2 btu/ sq ft for each degree difference between slab surface and ambient air. In a shop you can run a bit higher slab surface temperature as you are probably not bare-footin?

Assume 85° slab surface temperature. At 58° ambient room temperature the slab could out put 85-58= 27 X2= 54 btu/ sq foot.

However that assumes your entire floor surface is at 85. With one missing loop, maybe 300 square feet of unheated slab? the deck is stacked against you.

Fairly easy to run your own heat load, especially on a finished building, knowing your r-value, window size, etc. You need to pin that number down a bit more.

http://www.usboiler.net/heat-loss-calculator.html

tylerkeck

> @hot rod said:
> Ideally with everything sized and installed properly, on a design day the boiler should run non stop, inputing just what the space needs to maintain the desired and designed for temperature. if some of the tube is missing you may still be able to get adequate output by increasing flow and or supply temperature.
>
> I'm not sure 567 gallons of fuel for 7 months, 210 days of heating is all that unusual. 567 divided by 210 days is 2.7 gallons per day?
> What temperature does the system maintain on a 0 degree day, for example?
> If it was designed for 65° indoor 0° outdoor and you can only maintain 55 or 60, it could be related to the abandoned loop, you are running on 4/5 of the designed system distribution.
>
> If you paid $1500 for that 567 gallons that is $2.64 per gallon X 2.7 gallons per day= 7 bucks a day to heat the shop?
> Granted they were probably not all design condition days. And it depends on how the space is used, overhead doors opening and closing etc.

Thats where im at..i really dont know if its good or bad..it is a body shop with a paint booth that sucks the heat out and 2 10x10 doors. If i could buy a boiler and cut my heat cost down then it would be worth it. But im feeling that with my use its not gonna be a great cost ever..thanks

hot_rod

Ahh a paint booth!! Does it have a heated makeup air unit for the booth exhaust? The EPA and OSHA have some pretty strict and specific regs for ventilating a paint booth.

Whenever and wherever the booth sucks air from, that could be a huge load. That $1500 fuel bill is sounding better all the time.

Snowmelt

First thing I noticed is the pen at the machine is too small being being pex, I liked to see a minimum of 1 inch I have a takagi tk-3 doing a snow melt system on a 80 foot by two car driveway. Your primary looop should be a 1 inch with a 0011 or a 1-1/4 piping, you want to get as much gpm out of the machine as you can. The air scoop and pex is garbage, I like to see oversized copper and a air-eliminator like the taco 4900, you need two of them one on space side one on water heater side. After you chance the pump and copper you may see 4 gpm coming out of the water heater.

GroundUp

Snowmelt said:

First thing I noticed is the pen at the machine is too small being being pex, I liked to see a minimum of 1 inch I have a takagi tk-3 doing a snow melt system on a 80 foot by two car driveway. Your primary looop should be a 1 inch with a 0011 or a 1-1/4 piping, you want to get as much gpm out of the machine as you can. The air scoop and pex is garbage, I like to see oversized copper and a air-eliminator like the taco 4900, you need two of them one on space side one on water heater side. After you chance the pump and copper you may see 4 gpm coming out of the water heater.

Uhh.... That is nothing but a waste of electricity on these things. The orifices are much too small to make that work. Agreed on the air scoop though

STEVEusaPA

Why was one of the loops abandoned? Was it oxygen corrosion gumming it up? If so, probably more to come.
If it's a leak, can be easily found with a thermal imaging camera, and repaired.