Multiple Pumps into One Header 3

[nornrich]

All,

I am trying to pump water from 5 different storage tanks back to one large common tank. I would like the header to be gravity flow back to the one common tank, but I am unsure if this is even possible. I would like to pump 60 gpm, through a 2" Sch 80 pipe, from each of the five tanks into a 6" sch 80 PVC header back to the main storage tank. I don't know if this is possible if all five pumps are fighting one another to get back into the header. What will the last leg coming into the header need to be fighting not only the static pressure head, but the pressure head from the previous four pumps?

Regards,

Rich...

Richard Nornhold, PE
nornrich@redrose.net

 

[Montemayor]

Rich:

Your query, as stated, raises questions:

1) It is presumed you are pumping water from 5 identical pumps at a common flowrate of 60 gpm into a common, overhead, 6” header that is sloped or flowing by gravity into a main storage tank. The header’s contents are “dumped” into the main storage tank through its roof via free-fall. The header does not feed the main tank through the bottom of the liquid contents in the tank. Each pump is located at a different location and each pump is assumed to operate simultaneously, yielding a total of 300 gpm total flow into the overhead, 6” header. Is this description correct? If not, please correct my assumptions.

2) It is assumed the main storage tank is open to atmospheric pressure. Is this correct?

3) If the main storage tank is at atmospheric pressure, why the employment of Schedule 80 PVC pipe? Is it due to piping support spans? Is it due to expected water pressure requirements? Heavier pipe schedules reduce the pipe carrying capacity.

4) All 5 pumps do not have to “fight” each other to get their discharge into the 6” header, if piped correctly. Why makes you think they would? I have assumed the 5 discharges into the 6” header are located at different locations along the header – at least as apart as the 5 source tanks are. Is this correct?

5) What do you mean by the “last leg coming into the header”? Is this the 2” discharge into the 6” header that is located the closest to the main storage tank? I expect this point to have the least pressure drop in the system.

If my assumptions are basically correct, then what you have to do is to calculate the pressure profile of the 6” header as the incremental flows are commingled in it and progressively approach the main storage tank. Basically, pipe sizes increase in proportion to their capacity – as viewed from a velocity point of view. In other words, I know 2” pipe will easily transport 60 gpm through a 25 psi drop; therefore, since 2” pipe has a cross-section area of 0.0205 ft2, and 6” has 0.180 ft2, it figures that the 6” header has a chance to handle the total flow of 5*60 = 300 gpm since 5*0.0205 = 0.1025 ft2 – which is less than the 6” flow area.

You should make the necessary pressure drop calculations, using the relative distances and heights of each of the 5 pumps’ discharge into the main header. You start the calculations at the main storage tank and work backwards, obtaining each of the 5 individual discharge pressure required by each of the pumps. Since you are using PVC, your friction factor should be very low and if you use a minimum of fittings and only use long radius elbows, you should get a reasonably low pressure drop in the piping – depending on the trajectory length of each pump discharge and the main header. The hydraulic calculations done with the Darcy-Weisbach equation should be pretty straight-forward and yield a reasonably accurate pressure profile of the system.

I hope this helps.

 

[nornrich]

Montemayor,

Your assumtions were spot on. I was actually approaching it from the other end and figuring the total pressure the first pump added to the system, then the second pump and so on until it got to the main storage tank.

1. I am using identical pumps each pumping 60 gpm into the header. Total output at the main storage tank is 300 gpm.

2. Main storage tank is open to atmospheric pressure and the header is dumping into the top

3. Schedule 80 due to the spans between the hangers and the existing system we are tying into is sch 80.

4. I was worried about the flow from the upstream header being pumped into the drops downstream.

5. This is the last leg coming into the header prior to the main storage tank. Using your method for calculation of the pressure profile gives this leg as having the least amount of pressure drop. Using the way I was approaching it gave this leg the most pressure drop and hence the assumption that the pumps have to fight each other.

Thanks for your help. Any other suggestions please let me know.

Regards,

Rich....

Richard Nornhold, PE
nornrich@redrose.net

 

[dicksewerrat]

Put check valves near the header on each 2" line.

 

[SAK9]

You mentioned that the pumps are identical.They may be identical as far as flow rate is concerned.I can not see why they need to be identical in terms of pressure head.The pump farthest from the Main tank will have maximum head and the one closest to the tank the least head.

The pressure drop and therefore the flow rate through the system will vary widely depending on how many pumps are working.If the farthest pump alone is working,it will experience less than design pressure drop.This may result in overflow and and possibly motor tripping on overload.I would recommned an analysis of system pressure drop under various scenarios ie all 5 pumps working,4 working, 3 working etc.Then examine the range of head each pump needs to work against.

I would recommned the use of flow limiting valves at each pump outlet so that flow variations can be minimised.

Finally please try and define your problems clearly in order to get good responses from the forum members

good luck!

 

[vtpipeme]

Nornrich,

I think you've probably gotten all the feedback you need, but I'll just add my two cents since I've done this quite often:

1) Don't do it this way if you're trying to keep solids suspended. Use individual lines if you are. At the very least, make sure you always have the required minimum velocity.

2) Place a check valve on each line just before it feeds the header.

3) The question of whether or not you'll have 'fighting' in the pipes is really answered by this. Can each pump deliver the appropriate flow rate at the highest possible pressure loss through the feed line and header. Look at your pump curve to determine this. If the pump can overcome the calculated pressure at the required flow rate in each pipe, you're in tall cotton. Remember, when studying your pump curves, each pump is only 'responsible' for the flow it is sending, but the pressure in the header will obviously go up with the amount of pumps feeding it simultaneously.

In other words, add the pressure loss of the longest feed line to the pressure loss in the header with all pumps on. Can each pump over come that pressure at 60 GPM?

 

[brainstorming]

The following should be considered in your analysis:
- Since the common header will collect all flows 60 gpm per each pump, it should designed to handle the total flow Q.
- The minimum back pressure on each connected piping to the header should be considered for the flow to get in to the common header.
- The total pressure drop of the common header should be calculated based on all fittings (tee, bends, ..etc), total straight length and total flow with piping characteristics.

Regards

 

[katmar]

Art Montemayor's answer is so full of good advice that I am afraid that his most important comment is at risk of being glossed over.

Art said "You start the calculations at the main storage tank and work backwards".

This is so important, and I have seen people go wrong with this regularly. You must start at the point where you know the conditions, and then work away from that point. You can save yourself so much trouble by just tackling the job from the correct point.