Rail Car Offloading Issue (air lock of pump)

[Riley15]

I was involved in designing a rail car offloading system at a petroleum terminal and now that it is operational the operations folk are not entirely ecstatic about how it turned out. There is a pump that offloads 3 rail cars at the same time (runs at 800 gpm). The issue is that the rail car that is closest to the pump emptys first, draws air into the suction and air locks the pump. I realize now that I probably should have chosen a PD pump, or at least a self-priming pump, as opposed to regular centrifugal pump for this application.
The way they operate it now is once the first rail car goes empty then they close the valve on that rail car’s inlet, re-prime and start again with each rail car going empty.
My question is if I put an air eliminator on the suction piping, at the high point, would that do any good?

Any advice would be appreciated.

Thank you.

 

[BronYrAur]

Sounds like all the pipe sizes may be the same, so the friction loss to get to the 2nd and 3rd tanks is more. I am assuming each tank has a vent on top to break the vacuum. What about having graduated vent sizes, so that the closest tank has the smallest vent and the farthest tank has the largest vent? Just a thought ... not sure how much of an effect it will make.

 

[cvg]

I would not suggest reducing the size of any air vents, that could be a recipe for disaster

https://binged.it/2NIOhru

suggest reading this recent thread
thread407-443748

 

[BronYrAur]

yea, copy that! The last thing you want is someone closing a vent and imploding your tanker car. It will probably be a balancing act on the fluid side.

 

[LittleInch]

Have a look at those posts and replies from posts by fooooks who has similar issues.

If the three tankers are always the same then some sort of flow restriction in tankers 1 & 2 with a bit of trial and error should allow draining of all tankers at the same time.

Best bet is to drain tankers into a low level tank and then pump from there, but maybe not enough space.

air eliminator won't work because your inlet (suction0 piping is almost certainly at a pressure below atmospheric. Therefore air will flow in not out...

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

 

[Riley15]

Thanks a lot guys that is very helpful.

 

[itsmoked]

If you can take a little longer to get the cars empty then just reduce the extraction rate enough to wash out the dynamic issues caused by the existing flow rate. Throttling the pump output to some point will stop the cars from competing with each other. In your case it will be trial and error but there is a point where the problem will go completely away.

If speed is of the essence then you should automate the system you have now.
Put valves on the lines from each car.

Open only the far car and drain it to empty.
Close the valve and open the second car valve - drain empty.
Step and repeat.

You could probably sense a car going empty with flow, sound, or current to the pump dropping, whatever is easier.

Then the operators bring in the cars. Hook them up. Press a button. And at some point a signal tells them, 'all cars empty'.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[bimr]

By adding a pressure tranducer to the suction of the pump, pump controllers like the Goulds PumpSmart may be used to monitor the fluid level in the rail car and can automatically reduce the pump flow and pump speed to avoid the air intake and cavitation problems.

Pump Smart

Do you have a budget that you are working to?

 

[SandCounter]

My solution to this problem can be found in this previous, similar thread: [URL unfurl="true"]https://www.eng-tips.com/viewthread.cfm?qid=443748[/url]


(Has anyone noticed types of questions tend to come in waves? For example, last week was "Orifice Week" a couple of weeks earlier was "Rail Car Week." Even the similarities in specifics are uncanny. Is this phenomenon some corollary to the concept of multiple discoveries?)

I used to count sand. Now I don't count at all.

 

[itsmoked]

Sand; I see those waves here all the time! I too scratch my head. No doubt it has something to do with google's search garbage.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[georgeverghese]

Another solution would to set up the feed manifold as a ring main to balance out the flows - from the tie in to this common header from the railcar furthest way, route another equal size header directly to tiein in at the pump suction. The length of this second header should be approx the same as the current header length.