New addition to existing radiant heat system

[Mike P]

Hi all,
I'm new here. I am looking for a little advice. We just tore down an old addition in the back of the house and replaced it with a slightly larger one adding a full basement. In the new addition I installed 2 loops of radiant tubing in the slab, 2 loops in the family room, 1 loop in the family room bath. 2 loops in the master bed and 1 loop in the master bath. All connected to a new Caleffi manifold. A new manifold is connected via 3/4" copper to the existing Uponor manifold. That was feeding 4 loops Mud room 2 loops, Kitchen 2 loops. I also upgraded to a Gurundfos Alpha2 1/16hp pump from a Grundfos 1/25hp pump. Now the boiler won't fire due to low flow(Viessmanns Vitodens 100w). I figure I have aprox. 3000' of 1/2" tubing installed. I have tried everything I know. Does anybody have any suggestions?

 

[Diehard]

The problem seems pretty obvious, wouldn't you say?
You changed pumps and resulted in a lower flow.
Each pump has different "flow characteristics". That is what they can pump in flow(gpm) vs head(ft/psi).

One of the Grundfos pumps is depicted below.
This particular model shows it has 3-speeds and is represented by 3 performance curves.

All I want to point out is that at different head conditions(pressure loss through piping) the flow rate(gpm) changes.The higher the head(vertical column) the less the flow rate(gpm)(horizontal scale).

So you have to compare pump characteristics curves to try to determine your approx. flow rates.

What is the minimum flow rate requirement of your tankless heater?



EDIT: BTW...a pump can only operate at a point on the pump curve.
For example: Looking at top curve...if head(pressure loss) was 12 feet the corresponding flow would have to be 6 gpm. But up at 15 feet the flow drops down to about 3 to 4 gpm.

 

[Mike P]

Diehard,

Thank you for the reply. I called Viesmann yesterday they said min flow required was .35gpm. I guess I did not mention that I was having issues before I upgraded the pump. I'm puzzled because pump says its flowing at 2 gpm. The flow gauges on the actuators vary from .8 to .5 gpm. I have even shut off the new manifold leaving only the existing 4 loops. But the flow issue is still there. I did not mention that I also purged the system and filled with RhoGard heat transfer fluid. I know the flow is less with Glycol but with 4 loops maintaining a .35gpm flow should be no problem. I'm scratching my head with this.

Thanks,

Mike

 

[Diehard]

Is it possible, with that maze of piping going every which way, that your flow gauges are seeing flow that may be doing some recirculating without actual using that flow?

OH...wait a second!!!!

That is a closed loop(s) heating system and the only flow going INTO that system would be for the purpose to fill and/or add make-up to the system.
Your flow meters are not telling you gpm coming from your tankless heater. RIGHT?

What am I missing here?

 

[Mike P]

You are correct, it is closed. The flow meters are on the return side of each loop, so that is coming from the boiler . Trying to calculate the head, but im a little confused with the friction loss charts for the Uponor pex. I'm trying to figure out what i'm missing also. And, I did replace the flow switch just in case. I just can't seem to grasp this one.

 

[Diehard]

I'm not familiar with tankless units but looking up yours, it appears that it may have it's own pump controlling recirculation of water through the unit.

Talk to mfr. ask if the above is correct and if so ask for performance curves of their pump. And/or any comments or suggestions.

What flow switch did you replace, just in case?

 

[Diehard]

What is you total length of the supply and return pipes between the tankless heater and where it ties into the manifold( or where it ties into your recirc loop.)?
Type and size of lines?
That would allow us to approximate pressure loss at a particular flow rate. I believe the friction loss tables are available to show friction loss for your piping. Typically given in head loss per foot(or 100 feet), at various flow rates.

 

[Mike P]

No, there is no pump in the unit. There is about 6-7' max of 3/4" copper between the boiler and the first(original) manifold with 4 ports being used. From there I tee'd off with 3/4" copper about 12-15' to the new manifold with 8 ports.

 

[Diehard]

Well if you say there is no pump inside the cabinet, I have to believe you.

It appeared that all their combi boilers have a small recirc pump behind the lower, pull-down panel.

So you're saying that your recirc pump, for your heating loops, is at the same time forcing water around through the tankless unit? Just doesn't sound right to me.

Which specific model do you have???

 

[Mike P]

its not a combi, plumbed at straight heat only. 3/4" in and out

Viessmann Vitodens 100-W, Electric water heater for the hot water

 

[Diehard]

So you have an expansion tank in you system somewhere?

I'd like to see a schematic showing how your single pump unit, for your heating loops, is arranged to also force a recirc flow back and forth to the tankless unit.

 

[Diehard]

The reason I mention the expansion tank is because typically when these tankless heaters are used for a closed loop heating, they come equipped with an internal expansion tank.

 

[Mike P]

No expansion tank built in. its a plain Jane heater. Attached is my lovely drawing of the layout. Sorry my drawing skills kind of stink

 

[Diehard]

A picture is worth a lot.

Okay, so I see you added an expansion tank in your closed loop.

Well I guess we're back to that one pump and whether it's capable of providing enough head through all that piping to have enough left to provide the required gpm.

Do you have one of those 3 speed pumps? If so, what speed is it set to? I would assume HIGH.

Do you have, or can you find, a performance curve for the actual pump you have?

Only a single zone of control? ie -single pump and no zone valves.

That discharge piping is a bit tricky for me to calculate head loss since it can flow different amounts through 2 different areas. I'm pretty rusty.

 

[Mitchell-DIY-Guy]

I don't know if my experience will be of any assistance here, but FWIW, here's my experience on "radiant heat" systems used in de-icing and snowmelt capacity, which is almost identical to what you are doing except the heated area is outside. But we had all kinds of flow rates and pumps to take into consideration.

In my area, the design criteria generally called for about 250-300' of tubing per zone or loop. The system I designed had 7 loops, held ~28 gallons of glycol. However, I couldn't afford the space or anything for a proper sized "boiler" so went with a 200,000 BTU Takagi on-demand heater, which was rated for closed-loop systems. I could have used 300,000 but that would have meant two units or a boiler...

Design criteria was you needed to move all 28 gallons IN ONE MINUTE through the system. However, the Takagi could only handle about 5 GPM. So we employed a "primary loop/secondary loop" concept.

https://www.radiantcompany.com/heatsources/ondemand/primarysecondary-plumbing/

We had one pump moving 5 GPM through the loop containing the heater, and a secondary pump moving 31 GPM through the other loop with all the zones.

I had full flow manifolds and valves, all custom made; the off the shelf units had too many flow restrictions for use in a snow melt system.

I designed my system for my driveway with a friend, and we looked at a LOT of schematics and came up with our own. It all worked well--no issues in ten years of operation, and then I sold the home.

The schematic was an early one, and two things changed as we built it: we upsized from the Takagi TK-Jr to a TK3, and we changed the manifolds from those listed (Everhot) to custom made copper units with full flow 1/2" valves.

 

[Mike P]

Mitchell,
Hi, the two things I see in your diagram that I do not have is the 4" apart Tee's and a second circ pump. I have 3g gallons of Glycol.

 

[Mitchell-DIY-Guy]

If you have just three gallons (3 g, right?) of glycol, you DO NOT need the secondary pump.

In my case that second pump is there to basically reconcile the fact that you cannot push any more than 5 GPM through the Takagi, but you NEED to move 28 GPM through the entire system. Hence the concept of the primary/secondary loop and the Tee's, and secondary pump.

I surely didn't invent primary/secondary, but it was invented just for this purpose. Honestly, if I could have had a 300,000+ boiler I could have gotten away without primary/secondary (and melted the snow and ice a LOT faster...)!

 

[Mike P]

Diehard,

The exact pump number is Grundfos Alpha2 15-55F. I have it set at auto adapt, well its all shut down now, no sense having the pump run 24/7 with no boiler kicking in. I tried speed 2 and 3 but it didn't work that well also. Not sure if this makes a difference, I had it working a couple of days ago. all thermostats on and it lasted a couple of hours. Seems like when the system hit a certain temp, and the boiler shut down it would not come back on again, I know Gylcol resistance increases at different temperatures.

 

[Mike P]

Mitchell,

Sorry, 35 gallons glycol

 

[Mike P]

I'm thinking I should re-pipe a little and create a primary secondary loop system and add a circ pump on the secondary side. Not sure if I should use a mixing valve