clarify my pipe size plz 2

[koshyeng]

Existing design in my facility is a 310'of 1" pipe runs from the discharge of a pump with a control valve to a tank(inlet at 10' from the liquid surface). At 100% open flow rate is 18gpm. Now with approval of plant expansion we need 25gpm thru this line.
With 1.5" line I get 65GPM but to be more economical I decided to have 91' of 1.5" pipe and the rest 218' of 1" pipe. I have calculated K using Crane TP 410 and came up with 524( I have changed K from smaller to larger pipe). Now

Q = 19.65*d^2* Sq.rt(Hl/K)

I have d and k. I am couldnt interpret wat would be Hl for this situation. I am assuming it as pump system head which is 321'(from specification sheet). Is this right? Please help me out. These are my first calculations on pipe sizing as an engineer(entry level)

Thank you
Koshy

 

[Latexman]

Existing design in my facility is a 310'of 1" pipe runs from the discharge of a pump with a control valve to a tank(inlet at 10' from the liquid surface).

Is the inlet 10’ above or 10’ below the liquid surface? It matters.

At 100% open flow rate is 18gpm. Now with approval of plant expansion we need 25gpm thru this line. With 1.5" line I get 65GPM but to be more economical I decided to have 91' of 1.5" pipe and the rest 218' of 1" pipe. I have calculated K using Crane TP 410 and came up with 524 (I have changed K from smaller to larger pipe).

I interpret this to mean you calculated K for 1” line, then used K_a = K_b x (d_a/d_b)⁴ to get K = 524 for the 1.5” line, right? If you did, that’s wrong. You do not want a 1.5” line with the equivalent resistance of the existing 1” line. If you do that, you’ll only get 18 gpm flow out of the 1.5” pipe.

Now

Q = 19.65*d^2* Sq.rt(Hl/K)

I have d and k. I am couldnt interpret wat would be Hl for this situation. I am assuming it as pump system head which is 321'(from specification sheet). Is this right? Please help me out. These are my first calculations on pipe sizing as an engineer(entry level)

If your pump generates 321’ of head at 18 gpm, that is right for 18 gpm. But, what head does the pump generate at 25 gpm? If it’s a centrifugal pump and 18 – 25 gpm is in the flat part of the curve, then it may be okay for both. It depends on the pump curve for your pump.

Some other thoughts:

Can you get new trim (larger Cv) for the control valve and get 25 gpm and have adequate contol without installing any 1.5" pipe?

If that isn’t enough, can you also install a bigger impeller in the pump and get 25 gpm? It may need a larger motor. Check the suction piping for NPSH.

If that’s not enough, I’d change the pump’s suction piping to 1.5” and then see if that does it?


Good luck,
Latexman

 

[koshyeng]

Latexman, thanks for your quick response.

Inlet is 10' below the liquid surface in the tank( tank pressure is 1psig)

I interpret this to mean you calculated K for 1” line, then used Ka = Kb x (da/db)4 to get K = 524 for the 1.5” line, right? If you did, that’s wrong. You do not want a 1.5” line with the equivalent resistance of the existing 1” line. If you do that, you’ll only get 18 gpm flow out of the 1.5” pipe.

This is what I found in CRANE book. I ain't sure if its right. I will check pump curves tomorrow as I just headed back my home.

Latexman, existing 1" control valve(level control) max Cv is 5.4. I calculated pressure drop using Cv equation with 18gpm which is ~ 8psi. I also back calculated by removing pipe and friction losses between two pressure guages in the field. it came out to ~ 9psi. control valve inlet pressure is 65psig. How can i size a new trim to get 25-28GPM. Plz advice me. Meanwhile I will also read Fisher's book on sizing.

Thank You
Koshy

 

[koshyeng]

I skipped few points in my earlier post. When I sized(excel spreadsheet) the existing pipe by taking all friction losses in to account, I got 18.5 gpm output. Does it help if go to new control valve without increasing pipe size to 1.5"

Thank You
koshy

 

[ash9144]

From a practical stand point, 2" pipe is about the same cost and more readily available than 1 1/2" (unless some expensive alloy or lined). Might be easier to just go with larger size and eliminate questions altogether.

 

[Latexman]

This is what I found in CRANE book. I ain't sure if its right.

It is wrong. I repeat, you do not want a 1.5” line with the same resistance as the existing 1” line. If you do that, you’ll only get 18 gpm flow out of the new 1.5” pipe. Do this - calculate the equivalent length of the 1.5" line with K = fL/D = 524. 3,348 feet, right? Why would you put 3,348 feet of pipe in if 310 feet will do the job?

Slow down and figure out the technology before you do any engineering. Read, and I mean really read, not skim, Crane TP 410 and your college fluid flow text. Work the examples. Then, systematically analyze your situation.

I usually break these type problems into sections as follows:

suction vessel, if any
suction line and fittings
pump
control valve
discharge line and fittings
discharge vessel, if any

I recommend you use Bernoulli's theorem for this problem and generate the "system curve". Plot the "system curve" onto the pump curve to get your operating point. Adjust the control valve Cv, pump impeller, length of 1.5" line and fittings to meet your objective. Don't forget to check the NPSH for the pump. Most pump problems are suction problems!

Good luck,
Latexman

 

[bimr]

Koshy,

The first thing that you need to start are the physical properties of the pumped fluid. It makes a difference whether you are pumping molasses, sludge, or water.