Pump sizing question for heating jacket. 1

[PhilSham]

Hi All,
Hope someone can help me, i'm trying to size a pump for a heating jacket of a closed loop system (Flow 10 Meters cubed per hour).
The loop would go as follows, Hot water supply (DN50 Pipe) - Pump - Manifold to 16 lateral pipes - Return manifold - Hot water return.

My fluid dynamics is a bit rusty and i'm stumped at the manifold. Would i be correct in adding the friction losses of all 16 lateral pipes to my total friction loss?
Also how would i go about calculating the loss at the manifolds? These will most likely be DN100 pipe with 16 number hole drilled along its length with in inlet/outlet being on the bottom and closed on the top?

Any help would be appreciated.

 

[BigInch]

You'll need to size the pump head to that required by the largest head loss of any closed circuit loop and the flow capacity of the sum of all loops. If the loops are different lengths, elevations, etc. the trick will be to find the flows in each loop.

 

[georgeverghese]

As suggested, it may not be very apparent which set of laterals gives you the highest dp, so you might have to check many of them.

You say you've got the feeder and return laterals placed at the bottom of the supply and return manifolds -okay.

What about the min flow recyle tap off ? That would need to be a top tap off the supply manifold, so you get to vent trapped air in the supply manifold back to the return manifold. The return manifold should not have any high / low pockets in it that will impede the return of trapped air / gas in this circuit back to the hot water expansion drum. The return water nozzle on the expansion drum should be at a higher elevation that the return header.

Its all well to run these calcs on the basis on liquid full pipes, but the piping design should also reflect this. Else you get high pressure drop and low flow through this circuit with air trapped all over the place. Take a close look at the 3D view of this whole piping circuit and iron out all these trapped air risks before you start your calcs - the piping engineer or drafter may well have to redo the piping layout after you've cleaned up on this.

To get values for head loss for piping fittings and valves, you'll need a copy of Crane Paper 410 or the GPSA or Perry.

 

[LittleInch]

Philsham,

Your issue will be how you calculate the pressure losses between the two manifolds. Pressure/head at the manifolds will be the same for all pipes. Unless all 16 pipes are absolutely identical in length, size, ID, bends etc, then you will get different flow rates through them until the pressure drop is equal between them all. hence for a pressure drop calculation if you have different pipes, you need to look at a umber of different flows until you balance all the pipes - very difficult without flow or pressure control.

so no, you don't add all the pressure drop across 16 together, they will all be the same, the issue is the flow might not be....

A schematic of your system would be good to see to take this further with some lengths, especially of the 16 tubes. If these are much more than 3-4 times the length of the manifold, you can probably forget the manifold pressure losses in this system.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[PhilSham]

Thanks for the info guys,
I've been told that the pump will need to supply 10 Cubes, i know that the laterals will not all evenly receive this, the middle ones might get 1/16 of the flow while the bottom ones will get more and the top ones will get less. That is just the nature of the system i guess without going into flow control on each of the laterals.

Each manifold will have a 1/2" valve and an air bleeder on the top.
I have loss equivalent charts for the valves and fittings to give equivalent pipe length.

So the total dynamic head i should use would be all losses (pipes, fittings ect) up to and including the manifolds plus the lateral with the largest head loss? But to find the lateral with the highest head loss i would need to model the flow rates across each lateral.

 

[LittleInch]

Why would the laterals not get equal flow? So long as they are the same size and the same length and the two manifolds are vertical then the pressure drop across each is the same, therefore flow should be the same, +/- 5%??

All the laterals will have the same head loss, but maybe not the same flow. As a start point assume they all get the same flow and work out head loss if the length is different and then start to iterate if required.

A schematic or sketch would be good to post. If the laterals are quite short compared to the rest of the system, or average velocity a lot lower, their pressure drop might not be important to get right.

If they are different length, what is the variance.

As head loss is proportional to flow squared, you are not likely to get more than double unless you have a really short tube compared to the others.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[georgeverghese]

Oops, no temp control valves to modulate hot water supply to each user ?

Are all jackets going to be consuming design case hot water demand all the time ?

And no operating scenarios where one or more jackets is being started up or shutdown while the others are in normal operations ?