> >
>

# Calculate pump head

#1
07-06-14, 01:09 PM
Member
Join Date: Nov 2011
Posts: 88

I'd appreciate some help calculating pump head in my current hydronic heating system so that I can estimate the flow from the Bell and Gosset NRF-22 pump pump curve. Hopefully this list of the components in the longest loop feeding a 20" by 22 section cast iron radiator is enough for the calculation:
3/4" copper
1' pipe
2X 90 elbows
1" to 3/4" Tee

1" copper
10' pipe
1X check valve
2X shut off valves

1" steel (screwed fittings)
28' pipe
2X 90 elbows
2X 1 1/2" to 1" reducers
1X shutoff valve at radiator

1 1/2" steel (screwed fittings)
24' pipe
2X45 elbows

2 1/2" steel main header (screwed fittings)
48' pipe
4X 90 elbows
2X 2 1/2" to 1/2" Tee
2X 2 1/2" to 1 1/2" Tee

#2
07-06-14, 03:40 PM
Member
Join Date: Oct 2008
Location: USA
Posts: 3,016
You can Google for pressure-drop calculations, and search for a computer program. But for residential heating systems, it is almost never done. Installers wouldn't know how and mechanical engineers wouldn't do it on a bet. But you can do it yourself, manually, with a copy of the Cameron Pipe data or the Crane Flow of Fluids in pipe. If you do it yourself, you will quickly realize why few people here would want to do it for you on a lark.

Anyway, the calculation can't be done from your data unless there is a schematic of the total system piping system. And you can't just go by the longest loop - all parallel routes come into play, and then once you have the total system flow and pressure drop, you have to go back and model each series and parallel path.

This question comes up repeatedly, and so far I have heard of no posters crack the nut. I think I could do it, even though very rusty, but it would take me several hours and, of course the missing data. Are you having a problem that you need to solve?

Last edited by gilmorrie; 07-06-14 at 05:07 PM.
#3
07-07-14, 05:46 AM
Member
Join Date: Oct 2006
Location: USA
Posts: 23,539
you can't just go by the longest loop
Actually, the way pumps are usually selected is based on the longest (highest head) loop. At least for a zoned system anyway... if the selected pump can pump the highest head loop, it can pump the others.

There is a 'rule of thumb' put forth by B&G many decades ago that serves the purpose of estimating 'system head' at a 'correct' flow rate.

Measure the total length of the loop, multiply by 1.5, multiply by 0.04

The reason this works is sorta simple... the x 1.5 accounts for the 'average' number of fittings that might be used in a typical system. The 0.04 is the head per foot of pipe at the 'correct' flow rate.

Example: a 100' run of straight 3/4" pipe would be appx 4' of head at appx 4 GPM.

When piped into system, you would be looking at about 6' of head.

YOUR issue is that you've got a whole bunch of different sized pipes... but I'm not sure if this will affect the above... because the rule of thumb accounts for the 'correct flow' with the 0.04' of head per foot. In other words, 3/4" pipe is about 0.04/ft at 4 GPM; 1" pipe is about 0.04/ft at appx 7.5 GPM; similar/likewise for larger sizes... but accuracy of this method falls off as the pipes get much larger. It's most accurate, but in general accurate enough, for sizes from 1/2" up to 1-1/2".

If you use that ROT, you come up with around 7' of head for that loop and that sounds about right to me.

You don't really need to know the exact flow, just that the pump can provide the needed flow.

If however, for some reason you have a need to know the actual flow, you are back where Gil said you will be... trapped in a mire of calculations.

I presume that your system is NOT zoned, and that there are many more rads in parallel on the circuit.

So, in this case, you will have more flow in the larger main brances, and this will be divided among all the parallel branches, and that is a royal PITA to calculate!

Are you changing the pump? What's the reason for asking this?