Chevk Valve Losses

[sjohns4]

I got handed what should have been a cut and dry duplex lift station (4 HP pumps), however it does not work. There is about 5' of head I can not account for in the 4" forcemain.

In my calculation I used a friction loss co-efficient of 2 or 3 (notes are at the office) for a 4" swing check valve (Americal Flow Control was installed). Is this enough? I have not been able to locate a flow vs. loss diagram yet.

The station is pumping from 56 to 75 gpm depending on WW water level. I was wondering if this was enough to even open a 4" CV all the way, which I would think would add more friction loss to the system?

If I can be satisfied there isnt more loss in the CV than I am accounting for, my next guess is a local high point with a bit of air causing the losses.

Thanks,

Mike

 

[cvg]

your velocity is less than 2 fps in the 4inch pipe. Unless the velocity through your cv is considerably higher, it will not lose 5 feet of head.

 

[bimr]

As CVG states, your flow is so low that you will have little dynamic headloss in your system.

If your are having a problem, then you should investigate the static head as represented by the difference in elevation.

Note that force mains should have a minimum flow of 2 fps. That is approximately 90 gpm.

 

[gerhardl]

sjohns4,

apart from that I agree with previous comments, I was wondering about your 'duplex' pump station. Is it a possibillity that pipeline layout is forcing one pump to pump 'against' the other by connecting pipelines higher in the system? (If your loss is connected with two pumps running together?)

 

[sjohns4]

Static head is definately the culpert. The land development guys didnt see a problem in moving the station a few feet - which translated to an additional 10' or so of static head. Since my company gets to buy new impellers for the pumps, i figured I had better beat this one to death, and make double sure we only buy one set of impellers, not two.

Using as-built data, I'm still comming up a few feet of head short on what the gauge reads vs. what it should read though. I'm still thinking a local high point and velocities not enough to move the bubble down stream.

Mike

 

[FiltrationEngineer]

I have been in a similar situation, but I designed the system which came up mysteriously short on head. I would suggest checking the pumps. It is not uncommon for off the shelf pumps to come up short on TDH.

I suggest you turn on the pumps and check the dead head pressure on the pumps by closing isolation valves and looking at pressure. Sometimes the impeller is not seated correctly at the factory, etc. I always check pumps now and lots of times they don't meet their dead head pressure, which is one reason why you oversize.

 

[cvg]

also a good idea to perform the pump test shortly after installation so that pump supplier can then be asked to repair / adjust / replace as necessary so the pump meets the specs - before acceptance and payment

 

[rcooper]

I can't think clearly in imperial units, so please ignore my comment if it is not relevant, but if you are pumping at a low flow rate, the pipe may have fouled, resulting in a narrower bore than the installed pipe. You say you are measuring additional head compared with your calculation. The roughness of the pipe varies with time, typically from 0.1mm up to 1.5mm.

With an existing system, get some good, calibrated instruments. Measure your static head and your head at various flow rates, using a valve to throttle the pumps as necessary (being careful not to damage the pumps). What you then measure is the system curve. As pipes get used, there are too many variables for you to be able to calculate the system curve reliably. The calculation will only give you an estimation of what the pumps will see.