Transportation of two fluids with different viscosity

[cica58]

Could somebody help me about the hydraulic analisis for transportation in a pipeline of two liquids with differente viscosity. That liquids are transported in batches separated by pig. What viscosity must I consider? Pipeline is 200 miles lenght.

 

[JStephen]

Wouldn't you just assume viscosity A for the length of pipe that had fluid A in it, and viscosity B for the length of pipe that had fluid B in it? The lengths would change, and so the pressure drop or flow rate would vary as a function of pig position.

 

[Latexman]

Most probably the highest viscosity is your "design case". I'm curious though. PD or centrifugal pumps? Whats the two viscosities and and corresponding s.g.'s?

Good luck,
Latexman

 

[DesignerMike]

I agree that you would have to design around the high viscosity.

Other considerations would be number and locations of additional pumps along the route (if any).

As Latexman noted, the pump type may have an effect. If you were using centrifugals, the flow would most likely be substantially different. (and power requirements)

 

[sailoday28]

Are the fluids homogenized or flowing in layers?

 

[cica58]

Fluid A is 150 cSt, 0.94 s.g.
Fluid B is 300 cSt, 1.015 s.g.
Fluids are homogenized.

 

[DrillerNic]

Pipeflo, a hydraulic simulator produced by NMS of Calgary, has a module that will calculate interface lengths between different fluid batches in a multiproduct pipeline- it's not always necessary to separate each fluid with a pig.

 

[TrevorP]

Your question is a bit confusing...
You have a very long pipeline (200 miles). So you send product A down this pipeline. After a while, you stop sending product A, and you want to send product B. To stop cross contamination, you send a pig down in between the products. When you have sent product B, you then send another pig and product A after it. And so on...
You want to know the viscosity, because you want to estimate the pressure drop.
Fluid A is homogeneous.
Fluid B is homogeneous.
Fluid A and fluid B are not mixed with one another.

Is this correct?

Some questions:
Can you have more than 1 pig in the line at the same time?
What is the maximum length of product B (the more viscous) at any one time?
How do you get the pigs back again???

 

[Latexman]

cica58,

Just do your design calcs for fluid B. Having the higher viscosity AND s.g. makes it your worst case. Fluid A will make the pumps pull less power.


Good luck,
Latexman

 

[sailoday28]

"The properties of gases and liquids''-Reid and Sherwood -McGraw-Hill --Chapt 9 of the 2nd Ed--Section 9-13'Estimation of Liquid-mixture viscosity'
Some insight to liq mix viscosity is given in the above reference.
I haven't looked, but Bird, Stewart and Lighthill might provide additional help.

 

[TrevorP]

Latexman,
Your post is not quite true. If the pig is pumped down the line with the products it will create a significant drag on the vessel wall. The worst case may therefore be the most pigs in the line at any one time, hence the reason for my questioning.
The last question, I must admit was simply curiosity. 200 miles is a long way, especially if it's a straight length of pipe - do the pigs get taken out the line and sent back by airmail I wonder???
Still waiting for answers from cica58...

 

[DrillerNic]

Trevor- our pigs go hundreds of km from the platform offshore to shore (arriving a couple of days later) and then new ones arrive at the platform on a supply boat every now and then.

Once a year an intelligent pig, together with an Intelligent Pig Engineer (what a great job title!) arrive (him on the chopper, the pig on a boat), and we send the pig down the line, back to shore where the survey is downloaded.

I guess Cica's pigs could go back to the start by lorry if he's onshore.

Maximum numbers of pigs in a line: we'll have up to 2 in our line at once; during pipeline commissioning pig trains can be much larger but I don't know exact numbers-pose the question on the Pipeline Board!

 

[cica58]

TrevorP

Fluid B batche is 107000 bbl; Fluid A batche is 125000 bbl; the capacity of pipeline is 117500 bbl. Fluid B is pumped by screw pumps and fluid A by centrifugal pumps.

 

[TrevorP]

OK, I am still not sure if my assumptions are correct. However, if they are, then you need to assume your worst case, which appears to be 107,000 bbl of B, 10,500 bbl of A and two pigs.
You will need to contact the manufacturer of the pigs to determine the pressure drop v flowrate characteristics.

 

[Latexman]

TrevorP,

You are correct on the design of Fluid A system, however it may be different on the design of Fluid B system. Market demand could dictate that they need two batches of Fluid B back-to-back. We don't know this for sure, but, it may be possible. And, who knows what will happen in the future? If they qualify product in lots on the basis of batches, they may want to put pigs between the two batches of Fluid B. The absolute worst case of *all* Fluid B and two pigs is probably only a few percentage points of conservatism. IMO, the complication of adding the 10,500 bbl of Fluid A to the design is not worth the effort for the design of either system.


Good luck,
Latexman

 

[TrevorP]

Latexman,
What you say could be possible depending on the system. I was kind of assuming batches of the same product would be mixed in a storage tank at the far end, but you are correct in that this isn't necessarily true. As you say, the difference is probably not that great in % terms.

Cica58
One other point to make. If you are starting a new batch of A with the line full of B, it will effectively mean that the centrifugal pumps used for A will need to pump the line full of B (But with A in the pump heads). Both sets of pumps should be sized for pumping mostly (or all) fluid B (+ 2 pigs).

 

[jmw]

Are you not using "snot" (drag reducers)?

 

[stanier]

AFT's Fathom can model such pipelines. However you could do a series of manual calculations with sections of line with different quantities of products. Compare the heads with the capability of the pumps. Derate the centrifugal pump if necessary for viscosity.

This isnt an unusual pipeline by any means. Many refinery facilities pump to offshore loading terminals and change the slate. They do not even use pigs to separate the fluids as the interface quantities involved are diluted at the other end in tankage.

Also the mining industry has tailings lines that can be carrying solids in a slurry or being purged with water.

 

[DrillerNic]

I built a long distance multi product pipeline carrying 6 different refined products 350km, without pigs. We determined the interface lengths (the algorithms were incorporated into NMS's Pipeflo hydraulic simulator at our request), put a sophisticated switching loop at the end of the line and diverted the interfaces into different storage tanks.