My SnowMelt System Troubleshooting, Low Flow Rates.

CPUjohn

My application is a snow-melt system for a 17' x 43' concrete driveway and a 4' x 16' walkway (800Sq Ft using a 200K BTU Heat source or approximately 250 BTU per square foot). Driveway is 4-6" thick concrete over an R10 2" rigid polystyrene foam foundation, over compacted gravel. I am using two Taco 0013 recirculation pumps, into a Rifeng 12 branch Manifold that feeds 900 ft. of 5/8" and 1100 ft. of 1/2" Oxy-Barrier PEX tubing (each loop averages 150ft.) back to a US Craftmaster (Takagi) AT-D2-IN indoor tank-less water heater, then out to a Taco Air Separator (VRTX100-3)/Amtrol RX-15 Expansion tank back into the two pumps. According to my LoopCad2013 computer program, I should have about 11.25 feet of Head Pressure for all of my 12 loops including the manifold. I used 1" copper pipe to interconnect all of the components. I filled my system with a mixture of 50% Propylene Glycol (Hercules Cryotek-100) 21 gallons total.

The performance last year was very slow, so we added a second pump this year, but there is NO DIFFERENCE ! We did not get the deep-freezing immediately after a snow fall, that we have this year, so that has also become a problem.

I am experiencing several issues which I believe means I may not have calculated enough head-feet for the boiler itself.
What level of flow should I be expecting if my two pumps are capable of about 23-24 PSI (52 Head-feet) at 10GPM ?

I am only seeing a 1/2 to one-gallon per minute flow rate even though my two pumps in series are capable of at least 10 GPM with 52 head-feet of resistance (or 50 GPM with 12 head-feet).

The boiler sometimes is producing a louder than normal sound when I run my system after a few minutes, when it's cold, but the noise goes away if I shut down and restart, and all lines are warm.

It doesn't appear that I am getting delivery of heat to the driveway, and the tubes are freezing, even though I should be able to pump down to -20F according to the Hercules documentation (air temperature 0 to 15F.) I am thinking I may not have enough glycol. I placed a mixture of glycol in a clear vinyl tube and sealed it into a 12" diameter loop and leave it outside to see how it reacts at what temperature it starts freezing, so far it's okay down to about 10 degrees.

I have left the system on since Sunday as the air temperature went from 32F on Sunday, down to single digits tonight (Tuesday), just so our lines don't freeze while we troubleshoot what is going on. It was set at 167F output, and is registering 79F input, at a flow-rate of 0.5 GPM, so I reduced it to 131F out, 73F input (still at 0.5 GPM flow).
If I drop below 131F, ignition will go out, and unit will not re-fire. I am afraid to turn off the system as it completely froze up last week with subzero temperatures, and then it snowed, then froze again, so I had to use a hose connected to my other tankless water heater and wash down the driveway with superheated water (took 5 hours on Sunday).

What is going on inside the this heater, that flow is restricted so dramatically ? Do you have a suggestion as to how I can eliminate the flow restriction, as I don't need a specific fixed temperature output, but maximum flow rate ?

Any ideas would be most appreciated.

Sincerely,
John
Fox Lake, Illinois

Tinman

Have you checked gas pressure while the heat source is firing? Did you account for the additional 200k BTUH in terms of your existing gas meter?

Tinman

1" mains for the water side is too small.

Tinman

I've never used a tankless for a snowmelt job, not something I would ever do. What kind of flow is required for that? Do you have pictures of the near boiler piping?

hot_rod

most tankless heaters are designed for low flow rates, high delta T operation. You may not be able to pump enough flow to get that unit to give you full BTU/hr output.

Is there a pressure drop chart for the tankless? Boilers usually show the pressure drop at various flow rates for pump sizing info.

Snowmelt

Do you have it set up primery secounday, the good thing about the takagi is you can see your flow. I have the same set up on my moms house. Most likely your water heater is flashes off some of the glycol. And your filter screen is getting clogged. Do you have any valves to isolate the water heater.
All you should need is 120 degree water.
What temp is the water going in?
What temp is comming out.
What is your gas pressure.
What is your pressure drop on system side?
Your pressure drop at the takagi is 15 or 25, you need 1-1/4 pipe
But can get away with 1 inch. I have one inch for my primery and 1-1/2 for my system side.
How many square feet are you trying to melt? For every square foot that's 125 btu's
Snowmelt .........

Rich_49

1. A question about the CryoTek you used , what was the dilution or parts per water to parts per solution ? http://www.oatey.com/doc/cryoteck.pdf

2. If you study the charts for the unit , T-D2-IN you will see that head loss at 10 GPM is clearly around 90' with 131* / higher fluid and will not go above 8 GPm if fluid is 122* / lower . Bad news is that running the numbers and coming up with like lengths you'll need app. 12 GPM to keep the slab at 37* or 11.5 GPM to keep the slab at 35* with head losses for the loops at 11.1 & 10.4 with a 50% solution . Please note the requirement for a 40PSI or above relief valve for full flow .

3. Looks as if this system could use a complete redesign and selection of components to do what you intended . 10.4 ft hd + 60 ft hd = 70.4 ft hd . I don't believe your machine has the balls to do this without raising pressure well above what you have now .

4. Might I suggest purchasing a 20 gallon indirect and seperating the unit from the solution and using the 20 gallons as the media for the snowmelt side of this system . By the way you'll also need to send 140* water out at 12.1 GPM to maintain a 37* slab or 132.5* at 11.5 GPM to maintain at 35* . I used Rockford for design and this is at design . These kinda match what you have , somewhat . But when we use 15* & 25* the story changes a bit . At 15* and slab temp of 35* you have 5.8 gpm req. at a head loss of 5.1' for the loops , S&R @ 20' at a temp of 84* , so 97 going to the panel . I would look at the solution concentrations and seriously think about seperation of the circuits .
5. Hope this helped .
If done this way you can also use Delta T , ECM circulator for the slab side of this system and use a wnole lot less electricity while always insuring the proper Delta T and flow rates .

hot_rod

A condensing boiler, piper directly to the system would be nice in that application, very high efficiencies, fits in a small space.

i think the heater you have is non condensing and will also need return temperature protection.

800 square feet of 4" slab is about 32,000 lbs of thermal flywheel.

37° fluid temperature coming into a properly sized condensing boiler will drive that operating temperature down to within a few degrees of the fluid coming into it, sucking every single BTU it can produce, and producing buckets of condensate along the way.

I see no advantage to trying to drive that at 20°∆T, and induce boiler cycling? The 32,000 flywheel slab is clearly in command of the thermal equilibrium, and the boilers operating conditions.

Snowmelt

Hot rod. The part about it needing a boiler protection. It really just needs a condensation tee / damper. They don't vent into chimney. They vent directly outside. I have the noncondensing 199,000 btu unit doing a snow melt application at my moms house.

hot_rod

What about the condensing going on in the HX when it runs for hours in condensing mode?

Can that model run for extended periods in condensing mode?

In my example the fluid flows directly to the boiler no mixing or P/S connection.

Rich_49

Rod this unit is in fact a water heater . The specs clearly show what the head through the hx would be with 100% water flowing at x-y gpm . Start adding glycol into thew mix and all the numbers change for the worse . That water heater has to taken out of contact with the glycol mixture if it is going to have a chance .

Mark Eatherton

If you didn't dilute the glycol (which you should have) its like trying to pump frozen maple syruple. It doesn't work well. You (as others have noted) MUST pipe it primary/secondary. An outside observation, this appears to be a DIY job. Didn't your supplier help you with the design and commissioning of the system?

It's not rocket science… It's much more than that!

Got drawings/schematics of what you have installed?

ME

Snowmelt

Hot rod the simple answer is yes. My unit ran for 3 years no problem. I'm trying to think outloud. Take a tankless non condensing, run 50 degree water thru it (mid winter ) say you have it for a washing machine place or a car wash, now say you have the water heater set at 110 degrees. ( again I'm only having a discussion outloud) the water heat can be running for 3 hours strait. Now the water heater is under the condensing stage. But here is what I think happening, the flue gases leave the water heater above condensing temp. As it moves3,6,9 feet down the flue pipe it cools down and starts returning back down to the water heater, that is why you need a condensation tee, preferable with a damper which I install on all my tk-3, tk-jr, & tm-50. (Non condensing tankless water heaters)

hot_rod

If the temperature coming out of the supply connection of the heater/ boiler is below the dew point of the fuel, it's probably condensing.

All cold start boilers condense, the key if for how long they can safely operate below that critical temperature, usually 130- 140F

Most agree within 10 minutes you should be above condensing mode, in a non-condensing appliance.

If the appliance was not designed to operate for extended periods in condensing mode, the flue piping really doesn't matter.

Typical gas fired water heaters may operate as you describe, but they are designed with "sloppy" ( a term I learned from a water heater manufacturer in-efficient heat exchangers, notice the flue temperature, and can live, sort of, with those operating conditions.

Even so a conventional tank water heater operating at 100F for years will probably show signs of corrosion and possibly burner damage, flue pipe corrosion, etc. Same with copper fin tube, non condensing boilers.

Snowmelt

Ok so what's your point, I just told you how the tankless water heater works.theres nothing to indicate that there made for a short burn cycle, sometimes to many short cycles is no good, the company rather see the machine on a loop and have longer run times.

hot_rod

My point, I guess is tankless DHW heaters were designed for heating DHW, high delta T low flow. Hence the name tankless water heaters. Period.

Boilers were designed for heating conditions.

Condensing boilers were designed for higher efficiency, low temperature heating application.

The ONLY reason I see for forcing a tankless DHW heater into trying to behave like the properly designed, listed, boiler counter parts is $$$.

Is there another good reason you have for using tankless DHW for heating applications?

The clique about lipstick on a pig comes to mind.

Snowmelt

I'll admit your right , it's cheaper not the correct application, there out there and they do work. It makes good for service calls. I just had a call yesterday on a Navien combo with no primery / secoundary. The sad thing is that it's right in the directions. After a while It seems like I will make anything work. I'm going to make my self look bad now. For my moms home I took a tankless water heater that was shot. I put a heat exchanger in for under 300 dollars, added some venting from stock along with isolation valves. There I have a price of equipment for a low temp application doing a fine job so dad doesn't have to go out and snow blow. Now that's pretty cheap. Although the piping around it is expensive. I think that's enough babbling about this lol.

hot_rod

Be interesting to look back and see how many times this tankless WH used as boiler question comes across HH, and other discussion lists.

Problem is homeowners think it's the radiant or hydronic system that is crap, gives this small struggling industry a bad name any time equipment is mis-applied.

Sure for your own, or family and friends you can get creative, as long as it is done safely. I've got some pretty unusual stuff in my own systems.

Snowmelt

Hot rod, I see it all the time with engineering blue prints in $400,000.00 homes / Condo's. Nothing safe going on there.

Tinman

I struggle with engineers all the time who aren't dedicated to our industry. In fact, I have a bit of that going on right now with a job. It's very frustrating. The title alone doesn't make them infallible.

The ones who are dedicated to our field like Siggy, Brad White, and others; are the very best and I lean on what they have to offer all the time.

BobC

When I was designing circuits for a living I would sometimes come up with an interesting way of solving problems. Before inflicting it on the customer I would build a prototype to be sure it worked and was bullet proof. Usually I was right but sometimes the law of unintended consequences won out.

To do otherwise is a risk I was not willing to take for both the customers and my sake.

Bob

4Johnpipe

I do not see what the target temperature is set at on the tankless. Tankless water heaters work different than boilers. They will (via solenoid valves) slow the flow through the heat exchanger to keep the desired set point. You can have control over this using a buffer tank. Or you can lower the set point and let it run all winter.

CPUjohn

I want to thank all of you for the information, this is the best site I have found with the most educated people !

I found the problem with my system was a clogged water filter (even though I had cleaned it, it immediately had fouled again).

I am writing to AO Smith (U.S. Craftmaster) this morning to verify that this inline filter (http://www.grainger.com/product/BANJO-Y-Strainer-3DUA2) is of the same particle size removal capability as what is installed in their heater, so as not to have this issue again.

Once I removed all the restriction (really had to scrub that little filter inside & out), my numbers shot right up to where you would expect them to be (between 2 and 4 GPM) depending on temperature setting (167F to 100F) where the lower the temperature, the more the flow-valve would open (I was not getting that before).

Secondly, I also did dilute the Cryotec-100 Glycol, which is why my system was freezing, so I will need to remove several gallons and install the AG product (concentrated) to bring those numbers down to where the system won't freeze for our climate. Right now it is only liquid down to about 8F degrees.

Third issue, was that the "Air-Fuel Ratio Rod" had soot on it, causing the system to shut down, when the concrete was cold at first start up.

Last year, here in Chicago, we had a ton of snow, and with only a single Taco 0013 pump, my system was able to adequately melt the snow, and we had a lot ! Prior to this season beginning, I cleaned the filter, and found a dead defective pump, got it replaced, and then decided to add the second pump for better performance, but had these several issues all season.

Our total cost of installation of this snow-melt system was under $3500 as we did all the work ourselves, 2" rigid foam board $700, pex tubing- manifold-expansion tank-air eliminator-pump $1400, Water-Heater $500, piping-mounting hardware-gas line-valves-venting-glycol $600. and second pump this season $300. We also have a much smaller driveway, and I could do everything myself. And for what I spent on the heater, a handful of seasons should more than adequately justify replacement when it wears out. It is not the most elegant or elaborate configuration, but it does work !

Thank you all for your help.

Sincerely,

John - Fox Lake, Illinois

Snowmelt

Your welcome, since your not adding anything to the system, close circuit I took the filter out. I would use one inch piping to get more btu's to get out to the driveway. What's your bill for a nice 10 inch snow fall? It also looks like you use 1/2 inch tubing. May I ask how long were the loop leanth?
When you start the takagi unit I will think the temp, is low like 40 degrees, how long does it take for the return water to get up to 140 degrees?

Snowmelt

Three pics of same heater with a p/s

CPUjohn

All my lengths are all around 150 feet, except for the farthest loop which is 5/8" tubing, and like 175. The cost all last winter was less than $100 for the season. I haven't fully tested the full run time yet, as I ordered a new much larger 80 mesh strainer (200 micron) to go before the unit to capture any particles, and it has a clear bowl 5" high filter as opposed to that tiny 1" one built-in (250 micron particle) that keeps getting fouled. I have cleaned 3 times and still getting particles of unknown origin, so I am putting a much larger filter that should solve this issue, then I can take out the baby one (or just leave it in if it remains unclogged).
It took about two hours to warm up the driveway from 38F to 65F today, on a trial run (with outside temperatures around 25F). We haven't had a good snowfall to test it with, so I'll have to let you know !

We activate our system manually from our cell phones (Insteon/UDI ISY994i) and my wife works for DOT in our county, so she has access to extensive weather forecasting, and is one of the key people in charge of initiating the snow plow trucks for the county, so we activate a couple of hours prior to a storm.

Thanks,

John
Fox Lake, Illinois

hot_rod

consider a dirt separators. These will filter down to a 5 micron size but have wide passage ways and collect the debris and magnetic particles in a bowl, the ball valve allows you to easily flush and inspect. This model can be in a vertical or horizontal application.

The trapped debris is never in the flow path, so flow rate stays low and consistent.

A Y strainer has pressure drop even when new and perfectly clean, as it traps dirt, pressure drop goes up quickly and considerably, causing the starved flow condition, low heat transfer, lockouts, etc.

Snowmelt

I rounded your number off to 15 degree rise in 120 minutes, that's a degree every 8 minutes, that's what mine does. What's your gpm thrue the machine? I think mine was 5.2 ish
We estimated about $20.00 - $25.00 a snow fall, I'm happy with that, it's a 80 foot driveway.