Liquid in gas pipe 1

[oillio]

Good Afternoon,

I have a question.
We have a pipe about 9km. We send gas inside this pipe.

I would like to know when we are going to have liquid. I know that we can use the Dew point.

Even if I know that we are going to have liquid is it possible to calculate the volume of liquid?

Thank you.

Best Regards.

 

[SNORGY]

You probably will want to predict liquid hold-up using software like NEOTEC Pipeflo or equivalent. You would need a fairly good pipeline profile to determine it accurately.

You might try Section 17 in GPSA for two-phase flow if you wanted to go through a lot of calculations manually.

Regards,

SNORGY.

 

[zdas04]

Measure the water vapor content at the beginning of the line. Measure the water vapor content at the end of the line. Subtract the second one from the first and multiply the difference times the volume flow rate (since the water vapor sampling gives you answers in mass per volume). Answer is mass of water "missing".

In 9 km you can be very certain that you have a large number of sags and lowpoints. Every one will accumulate liquid. That liquid will support bacteria colony's (think MIC) and will absorb any available gases (think CO2 Mesa Attack). It may only be a few kg/day, but that number adds up quickly into real problems.

If your velocity is high enough (say around 3-4 m/s) then the smaller sags will generally get dried up by the sweep effect. Bigger sags will collect liquid even in very high velocity lines.

All of the pipeline models handle condensation and water accumulation as an average of averages. None of them is going to give you an answer that you can use. The multi-phase flow correlations are even coarser and of less value.

David

 

[dcasto]

if you have a multi componet gas, then you will need to run a flash calculation on the composition at the inlet conditions and at the outlet (or continously as the temperature and pressure change along the line).

 

[oillio]

Snorgy,

I would like to do it manually. I read Section 17 in GPSA.

I found the Flanigan Liquid Hold up correlation. For the calculation of the Liquid Hold up they just use the gas velocity.

I try to do the calculatio I found about 30 m3 of liquid in the pipe. It's strange because they don't need the temperature or even the density of gas to calculate the liquid hold up.

Best regards,

 

[BigInch]

According to this, that's the worst method.

http://ghajar.ceat.okstate.edu/pub_pdfs/Woldesmayat & Ghajar IJMF 2007.pdf

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[zdas04]

I've had success using both Flannigan and Duckler correlations. The trick is understanding the assumptions that went into the development of the equations. I wrote a program for a client that required some degree of confidence in the effect of elevation changes and liquid hold-up on a flow stream. So I dug into the old literature and found the original papers on both correlations (before you ask, those papers are now in a box in another state and I'm not digging them back out) and reviewed their assumptions and their data. The assumptions worked for the program I was writing and their data looked applicable to the program so I used them. Other times they haven't fit with the goal of my calculation and I looked elsewhere.

Back in the dim dark slide-rule days, a lot of effort went into developing correlations that the average Engineer (who was no smarter, no more motivated, and no less lazy than today's crop) could successfully solve more times than not. Some of the correlations stood up to a wide range of problems and others were applied far outside their applicable range (e.g., Turner did his vertical flow analysis at over 1,200 psig and I see people today doing the "Turner Calculation" with 10 psig wellhead pressure), but all were more or less slide-rule friendly.

Flannigan is not evil, users of it just need to do their homework to determine where it has a chance of success. You have to dig pretty deep to find out why it has the parameters it has.

David

 

[dcasto]

what good is it to have all those nice papers when you have no idea what the vapor liquid split is a .005% liquid versus 10% will give 180 degree different results because you need to to know the flow regieme.

 

[zdas04]

And they all assume steady state in a single flow regime which is simple nonsense--steady state in multiphase flow lasts for milliseconds and then changes to something else.

For the program I was writing, I was trying to predict the effect of a flow-profile-modifying tool and in general terms was able to predict flow improvement in the range of plus/minus 30% which is better than flipping a coin. The client was happy and the results made their clients happy.

The starting assumption was that there was no liquid water at the head of the pipe, so the water cut was a function of the dew point and was in a range where Duckler and Flannigan were ok.

David

 

[oillio]

Thank you very much it is very interesting to read your point of view.

Actually I don't need an accurate value. That is why I wanted to do it manually.

I have never work on this kind subject. I would like to give some details about the situation.
The gas is coming from an MP separator and then go to the 9 km pipe. There is no liquid in the gas. But for me, I will be faced to liquid formation during gas transportation. I just wanted to know if the scrubber at the end is sufficient for the liquid formed. That is why I wanted to know the volume of liquid.

I really thank you. I learn a lot with your responses.

Best Regards,