Multiple pumps into common header and header sizing

[Fidens]

Hi all,

I’m hoping to get some confirmation in my thinking. My current project situation is as follows:
An ash pond will have multiple sump pumps located all around the pond for ground water management over the years. There will be a total of 30 sump pumps, roughly every 300 ft. and the calculated flow rate is very small; 10 gpm (assume continuous operation). The water pumped out by the sumps will all be collected in another pond and the furthest sump pump will be about 3600 ft away. Currently, my set up is to have the each pump (2” outlet) connected to a common header (check valve before each header connection and angled entries downstream to the header and clean outs along the route). I have a 2 part question: Sizing the header to handle the total flow and concern about the pump sizing.

1. For header sizing: I know in rough numbers I will add up the cross-sectional area of each 2” inlet into the header and size the header to handle that volume. So on the longest leg, I’ll have about 12, 2” inlets, at 10 gpm for a total of 120 gpm. Will at 4” pipe handle that? I’ve done the calcs and the answer is yes, easily. Works out to about 1.02 ft head loss per 100 ft. Is this the correct approach? Or is a better approach is to start with a header at 2” (furthest point) and increase up to 4” header when it makes sense pressure wise (about half way)? OR Just make the header 4” and be done with it. I’m aware that economics plays a role and I will run those numbers down myself, I’m more interested on feedback on the proper engineering approach/philosophy. Or maybe economics is the major factor !

2. From the perspective of the furthest sump pump: does the flow rate of the preceding pumps (closer to the discharge pond) have a head loss effect on the ones further away? I’m pretty sure the answer is no, each pump is responsible for pumping their respective 10 gpm per their respective X distance. But I can’t help to think that the volume of the preceding pumps will influence the pressure/head loss. What is wrong with this thinking?. So when sizing the pumps I only have to worry about the individual distance that each pump will “see”? The furthest sump will see X of head loss and the pumping needs will get progressively less as we get closer to the discharge pond? (Even though there is more volume to “push”). I’m aware that the distances are progressively shorter so the pumps will have to be selected to have an “operational range” and it may not be possible to have them all the same.

I would appreciate any insights the community might have to straighten out my thinking.

Thank you in advance.

 

[LittleInch]

This is a classic network problem and there are some software programs out there and tutorials.

However the first thing you need to do is draw out your system in a schematic fashion. I can't see from here as my crystal ball is a bit cloudy at the moment how your 30 wells are connected so it's difficult to be exact.

Probably best to start on the RHS of your diagram with the end point and then work backwards in sections.

The key point is that each branch at the junction point must be at the same pressure.

So the start point is the pressure / head at your end point. The common pipe (if there is one) has all the flow and hence you can calculate head loss back to the first junction. The you can start to split each branch off.

So in simple terms the pressure ( assuming the pipes are flat) will gradually go up and up as each branch and each pump is until you reach the end.

So I think your item 2 is not correct.

Item 1 - it's a commercial thing, but with 300 feet between each one I would go for the smallest size I need. Due to the big difference in head loss as size increases, my guess is that the optimal choice out of your 12 is 3 or maybe 4 wells on a 2" pipe before going to 4". all depends on the head loss numbers and flows. For such a small flow system, you might be at 8 r 9 before the head loss in your 2" pipe becomes uneconomic.

The equivalent area trick only works if the flowrate in the smaller pipes is at such a level that you're in the one foot per hundred mark. Your 2" pipes are probably much less than that so combined at the end flow your header loss is OK for a 4" pipe





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

 

[Fidens]

LittleInch, thank you for your prompt response and sorry for the late reply (I was reading further into all your comments/suggestions)

I missed this thread (thread407-303444) the first time around and my problem is very similar as described there (I'm not residential and have more pumps) but the layout is practically the same. It pretty much agrees with what you've mentioned above.

thanks again.

 

[bimr]

Not sure if this will work for your application, but have you considered the semi-positive displacement pump made by eone? The semi-positive displacement pump characteristic means that the flow (Q) from the pump is very nearly constant, no matter what the back pressure.

http://www.eone.com/sewer-systems/regions/us/product-catalog/explosionproof.html

http://www.pumpsandsystems.com/topi...lacement-grinder-pump-wastewater-applications