Flushing a Hydrocarbon Line

[GBinns]

I want to flush one hydrocarbon product out of a line with another hydrocarbon product. I need to calculate how much of the second fluid is required to flush out the first fluid down to 0.01 vol% (amount of first fluid left in the pipe) in a 26" line (4 miles long).

I have found the "power velocity distribution law" equation. Has anybody had any experience using this in this type of application? Does anybody have any suggestions for solving this type of problem?

Thanks.

 

[BigInch]

You do it by introducing the second fluid fast enough to form a limited interface mixture between the two batches and then you just keep pumping the second fluid at that rate until the first is removed. The quantity of the interfacial mixture generated is a function of the two fluid properties and the flushing velocity. Four miles is not enough distance to generate much of an interface if you do this properly, it should only amount to a few barrels. Search for Austin-Palfry or similar interface generation prediction equations.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[GBinns]

Thanks for the reply.

We are, unfortunately, limited in our flow rate.

I'm interested in the Austin-Palfry equations, but I have been unable to find them. Can you help me out with that? Are they available in a document somewhere online?

Thanks.

 

[BigInch]

All you really need is to get the flow to turbulent conditions and you pretty much do all that you can to keep the interface down. It can be improved if the two products are of similar densities, or by introducing an intermediate product between two products of great density differences, but within 4 miles, you're not going to get much help one way or another by doing anything special. Actually anything you do just might make it worse. If its critical, some say batching spheres help, other's say they don't. Experimentation is the rule.

What's the products?

I haven't found Austin-Palfry equations online. Austin-Palfry method is the most well-known, but I use some Exxon interface generation equations which similar results and I have more complete documents on it. I can send those to you if you drop me an e-mail. See my web page for the address.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[abcmex]

your requirement is quite low 0.01%. If you can not create a turbulent flow (due to pump capacity or design pressure) it may be difficult to get. Have you consider to use pigs?

rgs

 

[MortenA]

as abcmex says:

Any chance of inserting a pig?

Best regards

Morten

 

[BigInch]

With 4 miles to work with, I doubt a batching pig will have much of an effect. If the flow is maintained at turbulent ranges, most of the interface is generated from wall film pickup anyway. The kicker is that pigs leave a wall film and it takes a certain length of secondary product behind them to pickup the wall film anyway. That's why I said, sometimes it works and sometimes it doesn't. Depends on the 2 products involved.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[abcmex]

Biginch,
4 miles + Turbulent flow sounds like high pressure drop. just to fill a experience gap:

Does longer pipelines (3-15 miles) operate at turbulent flow?
Are this kind of pipes typically mechanically calculated for this?

Kind Regards

 

[BigInch]

Almost all pipelines are designed to operate within the turbulent flow regime. When typical pump power cost to increase pressure is balanced against the pipeline pressure loss (plus pump and pipeline capital costs), it is usually more economical to have a pipeline diameter that yields a Reynolds number within the turbulent zone. Perhaps it is possible that some gravity flow pipelines might not fit this generalization, but all powered pipelines I know of do. Laminar flow is just too slow in relation to the required investment to make money transporting bulk comodities.

BigInch-born in the trenches.

http://virtualpipeline.spaces.msn.com

 

[abcmex]

Thanks!

 

[dcasto]

If you keep the velocity up to about 5 feet per second, you will minimize the interface, but do not stop, if there is a slight density difference in the fluids they will seperate. The interface could be as small as 50 bbls with 250 bbl on the upper end. If you put a pig in first as MortenA was asking you will minimize the mixed interface to next to nothing, 20 to 50 bbls.