Primary-Secondary Decoupler Piping

[bkallen]

This may be difficult to describe without a diagram, but I'll try.
Generally, the "text book" diagrams for a P-S pumping system (chilled water) look as follows: The decoupler tee's into the CHWS prior to the secondary pump header. i.e. The decoupler connection to the CHWS is "upstream" of any tee's to the secondary pump suction.
(reference fig. 7 at

http://www.bellgossett.com/Press/VVFund.htm)

Question is, would the following arrangement be hydraulically similar:
CHWS main comes from chiller(s), tees off to sec. pump #1, then tees off to sec. pump #2, then continues on, effectively becoming the decoupler. i.e. The decoupler connection to the CHWS is "downstream" of the tees to the sec. pumps.
Looking at the same fig. 7 referenced above, this would mean extending the horizontal header pipe past the sec. pump on the right and connecting the common pipe over the "b" in the word "Distribution".

Sorry for the complicated post.

 

[KenRad]

bkallen,

I believe that I understand your question, and that what you are proposing will work just fine. In fact, I designed a system like this a while back, and it works just like any other P-S.

The only difference that I can see would occur if your secondary flow was to overtake your primary flow -- i.e., if you had reverse flow in your common pipe. In this case, the secondary pump closest to the secondary return connection would receive all of the excess flow. That is, one of your secondary pumps (the first one tee'd off of the primary) would be receiving all of its water from the chiller discharge. The other pump would receive a mixture of this cold chiller water and warmer secondary return water coming from the common pipe. The net result will still be the same -- as long as your secondary pump discharges combine into one header, all of the water will mix.

Hopefully my response was not any more complicated than your post.

---KenRad

 

[ChasBean1]

I'm glad KenRad gets it because I took pen to paper trying the draw the system as you describe it & have the secondary pumps not connected to anything... definitely tough w/o a schematic. Good luck...

 

[zakh]

You may also want to limit the pressure drop on the common pipe. Based on your proposed config., i would believe that the pressure drop will include all piping (and fittings) from the last tee (right-most sec. pump based on the ref. drawing) to the point of coupling at the return pipe. For discussion sake, imagine if we extend this piping to infinity...the pressure drop would also be infinity and the system would revert back to a series pumping system.

 

[ANILKR]

CONTINUING ON THE DISCUSSIONS ON THE DECOUPLER LINE PLS ADVICE HOW THIS LINE IS TO BE SIZED.CAN IT BE ARBITARILY TAKEN AS THE SIZE OF THE HEADER LINE OR IS IT THE CASE WHEN IN CASE OF A MULTIPLE CHILLER INSTALLATION OF SAY OF THE SAME CAPACITY THE LINE SIZE TO AN INDIVIDUAL CHILLER TO BE PROVIDED OR SAY IN CASE OF MULTIPLE CHILLERS OF VARYING CAPACITY THE LINE SIZE TO THE SMALLEST CAPACITY CHILLER IS TO BE TAKEN.PLS ADVICE

 

[KenRad]

Size the decoupler line for design flow – if your system is designed to run all chillers at the same time, use this flowrate. Then design your decoupler to give you a negligible pressure drop at this flow.

The idea behind P-S systems is to have two interconnected loops that are hydraulically isolated. That is, flow in one loop will not induce flow in the other. For this to be true, there must be virtually no pressure drop in the common pipe. Obviously, you need to look at both pipe diameter and length, and should not have any fittings in this decoupler section.

Bell & Gossett says to make the decoupler a maximum length of 4 pipe diameters, but this doesn’t take into account pipe size. If you keep the pressure drop down to a fraction of a foot (say, 0.1 ft, or .04 psi), you will be in good shape.

Look at the B&G website for more information.

http://www.bellgossett.com/Press/2ndpump.html

---KenRad