I'm heating a 1,000 square foot area (heat loss is 31,000btu/hr) with 5 x 300 foot 1/2 pex loops under floor attached to thick aluminum plates. Finish floor is not installed yet.

The 5 x 1/2 inch pex loops are supplied from a radiant manifold that's about 20 feet from the boiler manifold. 3/4 inch pex supplies/returns water to/from the radiant manifold. I'm using a Grundfos UPS015-58frc pump (3 speed in the high setting) to circulate water from the boiler manifold to the radiant 5 loop manifold. Over 24 hours, while only operating the circulation pump (not turning on the furnace with a thermostat) water temp to the radiant manifold is 110 degrees (set via mixing valve) and return temp at the radiant manifold is 100 degrees. Room temp is 74 at 30 degrees outside.

Here's the question: I'm not able to read any flow rate off the 5 loop radiant manifold guages...the red markers in the guages are so near the top that they are nearly reading zero flow. My guess is that they are reading 0.1 or so of flow. When I close the valve on the return side of the radiant manifold, the guages barely move. On a different 2 loop manifold with the same pump and, in this case, only two 300 foot 1/2 inch loop lengths, I'm getting a 0.5gpm reading on each loop at the radiant manifold. Does this mean I'm getting 1gpm total flow to the 2 loop manifold? If so, the 5 loop manifold receiving the same 1gpm divided by 5 would yield about 0.2gpm at each loop...which is practically imperceptible to read...is this flow rate too low?

When I use a 3/4" pex friction loss rate at 10gpm, the head loss is very large (21.1 foot head loss per hundred foot)...could my pump be too powerful and creating too much friction in the 3/4 inch supply/return lines, therefore, very little flow is reaching the manifold?

I've tried to calculate the head for the 5 loops and get a very low number (5.43) while using the .2gpm flow rate. 1gpm out of the pump is minimal with the pump I'm using. Even if I use 10 feet of head with the pump chart, it shows I should be getting 10gpm out of the pump. Can the pump really be putting out 10gpm and each loop only read 0.2gpm, with only 20 feet from the pump to the radiant manifold? Why is it so low? What am I missing?

Also, to calculate friction loss and flow rates, I will use:

Flow Rate Required = Heat loss Btu/Hour/ ((YY * XX * delta T) .... what are variables XX and YY?

Ideas and feedback are greatly appreciated.

Thanks, John.

This message has been edited. Last edited by: diyjohn,

so I've done some additional calcs to see if this is a pump issue...

flow rate required = 31,000/(500*20) = 3.1GPM

3/4 inch supply/return tubing is 44 ft actual pipe plus 4 x 90's (equiv. to 38 feet of pipe) = 82 feet pipe. This yields 4.264 feet of head at 3.1GPM (.052 head loss factor x 82)

One loop of 300 foot 1/2 inch pex (manifold is 5 - 300 foot loops) yields 3 feet of head at 0.62GPM (.014 head loss factor x 300). Are all 5 loops included in head loss calcs or just one?

Total head required = 7.264 feet.

It appears the circulation pump is adequate ( http://s3.pexsupply.com/manual...8fc_Product_Data.pdf) as it delivers 17 feet of head at 3gpm...

Can someone tell me if my numbers are close? In particular, are my head loss factors accurate?

thanks, john.

I think your 5 loop manifold is plumbed in backwards, or you have an air problem and you need to purge.

thanks Rob -- I thought the manifold could possibly be plumbed backwards and checked for this...but it's not, and if it is, then the 2 loop is also - the supply goes to the manifold half with the guages and the return line comes from the manifold half with the balancing valves.

Is my math correct, if the 2 loop is showing 0.4gpm thru each loop, does logic follow that the 5 loop is getting .8gpm total and therefore this is why I'm seeing such a low number on the gauges?

The system was purged without any change.

thanks again, John.

you can't estimate your 5 loop flow by your 2 loop flow.

however, you have flow meters here, which usually double as balancing valves. have you tried opening them? usually if you twist them, they open up.

0.2 GPM isn't really a problem except for the fact that you're claiming a 30 BTU/sq ft heat load here (which 110 will not service, by the way). you might want more than that when it's cold, along with a higher supply temp.

something is obstructing your flow here. with seven loops you should easily be over 5 GPM I would think in this case, without doing any real math.

gotcha...I've opened up the balancing valves all the way but that doesn't help...with so little flow, i have to close one or 2 down a bit so those loops don't run colder than the others. The circulation pump only serves one manifold, so the 2 loop manifold has one pump and the 5 loop manifold has its own circulation pump....could the mixing valve on the return side somehow obstruct flow?

Can you check my math to see if the 5 loop should be at 3.25 gpm or so? did i calculate the pump head correctly? I thought I could use the 2 loop to guesstimate the 5 loop flow since these are 2 independent radiant manifold loops each powered by their own circulation pumps and exact same length supply and return lines...

I hear ya on the 110... I was only testing the system and was pleased with the results even at .2gpm and only 110 in...figured i would be in hog heaven if i could figure out why the flow is so low...

John.