Pipe Line Sizing 5

[mggyi2006]

Dear Professionals,

I am a fresh graduate in Chemical Engineering and started working in a petrochemical plant as a junior process engineer a few months ago.

My fist assignment is to size a pipe line and centrifugal pump from a storage facility 4 km away from our plant.

I need to do some fluid flow calculations such as pressure drop, friction loss, pump sizing, etc, pipe stress analysis as well.

I am kind of lost where to start calculations. So your kind professional advices regarding my concern would be highly appreciated.

 

[Latexman]

The best advice I can give you is buy your own personal copy of Crane Technical Paper 410. It is available online:

http://www.flowoffluids.com/publications/crane-tp-410.aspx

Good luck,
Latexman

 

[chemebabak]

Do not forget GPSA Engineering Data Book:

http://gpsa.gpaglobal.org/databook/

This is the best handbook that I have ever seen for Process Engineers.

 

[sheiko]

mggyi2006,

Don't panic! It's both very easy and a good training for you. So do it the best you can.

Begin with the pipe sizing and after, size the pump. Think about it.

See also:

http://www.chemicalprocessing.com/articles/2010/129.html

"We don't believe things because they are true, things are true because we believe them."

 

[mggyi2006]

Dear Professionals,

Really appreciate the courage, advices and information.

May I trouble you guys for further technical advices?

Best Regards

 

[MJCronin]

For a pumped system going 4 km, I suggest that you consider truck delivery of the liquid, especially if the supply needs are intermitent.

Youroverall cost of the pipeline, piperack, electric power for the pump etc. should be weighed against the truck delivery scheme

If your pipe rack already exists, the capital cost will not be as great

 

[mggyi2006]

Hello MJCronin

Appreciate your advices.

Our daily need of base oil is supplied by ISO trucks and pipe line delivery is considered as better option to cut cost in the long run.
Besides, we already have piperack structure in most parts of proposed route.

Best Regards
mggyi2006

 

[zdas04]

MJCronin,
I see this trade off all the time. Piping has high CapEX and very low OpEX. Trucking has Very high OpEX and very low CapEX. The plant operating people control the OpEX budget. The engineers control the CapEX budget. At design time some engineer says "Dang, it is only 4 km, we can save $200k by planning to truck it". He's a hero and gets a bonus.

Then a few months later, the plant manager looks at his OpEX and sees $10k/month for trucking over an expected plant life of 25 years and comes up with a project to trade that $3 millon in future OpEX for $500k in CapEX to retrofit a pipe and pump. He's now the hero and gets a bonus.

If we did realistic OpEX vs CapEX analysis this wouldn't happen so often, but I get sucked into variations on this theme about 3 times a year. The economics I run almost always come down in favor of spending the CapEX to avoid the OpEX. Sometimes it is nice to have a couple of these projects in your hip pocket for the semi-annual "how can we cut 5% out of OpEX?" memo.

David

 

[PetersPhotography]

Pipe stress analyses can be fairly intensive and often requires specialized software.

I'd start with defining your system:
1) What fluid are you pumping?
2) What is your suction pressure? What is your desired pressure at the end of your new pipeline? From these, you can define your total developed head requirement and size your pipe. Pipe sizing should be based on keeping pressure drops reasonable (typically below 1 PSID/100' in my experience) and keeping velocities reasonable. Your pipe MOCs and your pipe design pressure/corrosion allowance will play a role in determining your pipe size as well (economics).
3) Once you know your TDH and your desired flow, you'll need to calculate your NPHSa based on your storage system characteristics. Together, these will determine the type of pump you'll purchase (pump vendors will help answer these questions). You'll need information on required electrical classification (for the motor), desired seal types, flush plans, etc. as well.
4) Finally, you'll need to evaluate what instrumentation and controls are necessary. How do you plan on keeping your pump on the correct point on its pump curve? A downstream control valve - and a minimum flow recycle valve - are common. What other physical parameters need to be measured?

That's a very simplistic summary, but hopefully it helps!

 

[DubMac]

mggyi2006,

Very good advice from PetersPhotograpy above.

I might add, that on the pump side of things, you will most likely be using a horizontal splitcase, double suction pump which is typical for pipeline service. Whether it is a single stage or multistage will depend on the pressure reqired; anything over say 500' of head will most likely require a multistage.

The double suction impeller allows a much lower NPSH Requirement (in a multistage, only the first stage will be double suction).

A vertical lineshaft pump or "can pump" is usually considered when NPSH Available is extremely low, or if footprint needs to be very small. Many times the decision to go horizontal or vertical is based upon Maintenance/Operations familiarity or preference with one style or the other.

Whatever style, it will need to comply with API 610 since you are pumping hydrocarbons. This narrows down the manufacturers considerably.

Next you will need to consider what speed to run at. This is a balance between cost and reliability. The faster a pump spins, the faster it wears out; as speed increases, wear increases to the CUBED POWER!

So you would want the slowest pump you can get; BUT, as speed decreases, head decreases to the SQUARED POWER! That means the slower you run, the larger diameter impeller required to meet the head requirement, and COST GOES UP with size.

Try to stay at synchronous speed for main design point; that is, 3600, 1800, 1200, or 900 rpm (you would only consider 3600 if that is the only way to acheive the head, its just too darn fast). Not relying on a VFD to hit the design point will earn you bonus points from the electrical guys as it simplifies their world.

When you do select a pump, make sure there is some "head reserve" capability; that is, select an impeller that is at least 10-15% smaller than the maximum diameter offered for that model. If your head calcs are off, or down the road you require more head, you can accomodate by putting larger impeller in existing unit.

Also try to select a pump as close to the Best Efficiency Point for that model. Again if your calcs are a little off, you will still be at a decent efficiency and won't overload your motor.

I would highly recommend calling one of the "Big Boys" in the pump world, Flowserve, Sulzer, Goulds, CLyde-Union. They have all been making these pumps for many, many years and probably have them all over your plant. Ask each one of the salesmen to drop by and visit you and show their best selection. Don't just ask one of them. You can get painted into a corner, and believe it or not, some pump salesmen are REALLY good at pumps and can teach you a lot. That is what they get paid to do; USE THEIR EXPERIENCE.

Good luck my friend!

 

[mggyi2006]

Hello

To Petersphotography & DubMac,

Your advices are extremely useful and very much appreciated.

We are working on the project now

mggyi2006

 

[knapee]

Calculate Reynolds number Re first,

Re=3160*Q*S/(D*viscosity)
D is diameter (in),
find the table for Friction Factor
Then get hte pressure drop
deltaP = B*F*S*Q^2/(D^5)
For turbulent B=13.47

thanks