circulator for radiant heat

bauer

Hello -



Thanks for all the help so far -- based on the recommendations here I have picked up several of dan's books and learned a lot.



I have a question regarding circulator sizing however. I am adding a small radiant floor section to a boiler currently heating 4 radiators on a direct return system (small 1 floor, 800-900 sq ft apartment).



Based on advice from 'pumping away', I am going with the small, simple, inexpensive option of tying into this system with a mixing valve and 2nd circulator pump -- making the bathroom floor a slave to the main thermostat. The reason I decided this is based on Dan's rule of thumb that the controls shouldn't cost more than the piping.



Anyway, I am curious as to what circulator pump I need for this loop. I calculated a rudimentary heat loss on the bathroom of between 700-1,000 btu/hr. The length of 5/16 quick track tubing is less than 75'. These are given me very small gpm rates -- how small of a circulator can I really get away with?



Any help is greatly appreciated



Thanks!

bauer

gpm & head

ok so after looking at some equations - I've come to these calculations:



75' of tubing @ 120 degree water with a delta T of 10 leaves me with a GPM of .2.  Based on the head loss of 5/16 pex at 120 degrees, my pressure loss is 0.0223/ft, or 1.67 overall.



All the taco pump curves I look at have a much higher head than that at .2gpm.  What am I missing?

Gordy

Pressure drop

 Not missing anything other than other circulator alternatives. Google LAING they make some cute little circs for your situation.

bauer

still too powerful?

thanks...



Looking at the pump curves for the LAING E1, they still seem high... they are variable speed -- at about .2gpm, the head would be close to 6.... if I adjust to 2.5gpm the head comes down to 1.5.... are either of those situations doable or am I just looking at the wrong pumps still...?

EricAune

Valve head loss

You also have to consider the Cv rating of the mixing valve.  You have a calculated head loss through the mixing valve and the circulator has to overcome that as well.  This may not bump you up too far on the pump curve but it is part of the selection process.

CMadatMe

Delta

When designing with QuikTrak should be a 20 not a 10. You never mentioned a heat loss so I don't know how you came up with .2gpm



.2 x 5000 = 1,000 btus



.2 x 10,000 = 2,000 btus



75'of 5/16 pex flowing .2gpm of 120 degree water has a head of 1.35 plus fittings so your right on. Head is minimal even with the head of the mixing vlv. I'd use a Taco VDT pump or Grundfoss Alpha. The VDT will make sure your Delta is constant, the Alpha gives you the benefit of some electrical savings.

bauer

Heat loss

Thanks - my estimated heat loss is bet 700-1,000 btu/hr, that's what I got my .2gpm from - although that was with a delta t of 10.... It's more like .1 with a delta t of 20.



What's the reason for the higher delta t?



I'll look into those pumps - thanks.

CMadatMe

Size of Tubing

I was always told due to size of the tubing. Need to get btu's out. All there sizing charts as well as their software uses a 20. See page 24.



http://www.uponorpro.com/~/media/Extranet/Files/manuals/QuikTrak_InsG_9-00.aspx?sc_lang=en

bauer

delta T

gotcha - thanks.



What is the real difference that the delta T creates in a heating system? If the delta T is small, it would seem to me that the water (and therefore pipes) are remaining hotter farther on down the loop.... whereas a larger delta T means that the water is giving up the heat to.... whatever it is that is radiating the heat?



I guess I'm a little confused on how delta T affects the system...

CMadatMe

Lazy River

Think of it this way. Water is lazy has no fight. Send it out quick the btus just go for a ride. Send it slow btus get tranferred to the emitter. You make btus to use them not for them to go on a joy ride.

Gordy

Delta T hangup

 Chris is right, but I will add that Delta t is a design tool. In real life delta t is going to be variable up to what ever number you used to design around. It just depends on the load at the time.



 Quick track uses 20* because as Chris stated the btus need to soak into the panel, and not go through it. Think smaller dia. tubing with less surface area for conduction.



Gordy

bauer

uncle

ok, I call uncle...



I'm following all the theory and equations here...and appreciate the help, but for the life of me I have searched the high and low and cannot find a pump to buy --- am I over thinking this? Nothing seems to come close on a curve to providing such a low flow/low head output.



I see taco makes an 003-vdt that comes close... but I can't find a vendor that ships it

CMadatMe

You Are

The VDT is looking at delta-t could care less about anything else. The 008-VDT will work fine. Its job is to make sure the delta you want happens.

bauer

thanks

i often suffer from paralysis by analysis.... 



thanks again for everyone's help.... some of the most helpful forum posters I've come across.... i love reading threads on this site because there is never  the nasty, rude undertone you get on some forums.



thanks again!