TWO PHASE PRESSURE DROP

[40410]

I have to perform the two phase pressure drop for the oil flowing from the wellhead to the inlet manifold at the station via the choke valve and manifold.

known data are flow rate, wellhead flowline presssure and temperature and water cut.
Seeking help to the do the above

 

[malancha]

If it is from well head, there must be some gas and you should have the GOR (gas-oil ratio) data as well. Further do you have the full composition of the crude oil, gas and water? If yes, use compositional models otherwise use black oil model (which requires, oil sp.gr, GOR, gas MW). Obviously the lesser the properties known, the lesser will be the accuracy of the results.
There are numerous 2-phase pr. drop correlations available. Commonly used is Beggs& Brill. However depending on the level of accuracy required one can use the semi-empirical correlations (like the one indicated above) or may use more rigorous and sophisticated models like mechanistic models. Refer to Pipephase/Pipesim software manuals for guidance on the applicability of the available correlations depending on field conditions and pipeline configurations.
Incidentally water phase also plays an important part if emulsions are likely to be formed. One should consider if over the field life, how the water cut will vary and if there is any chance of phase inversion phenomenon i.e. changing from oil-continuous phase to water-continuous phase (which increases the emulsion viscosity tremendously. This in turn will affect the pressure drop in the pipeline. However if it is a simple pressure drop calculation for an existing field & existing pipeline with known properties, these considerations may not be required.
You should have an access to a proper software package.Most of the applicable correlations are iterative types and usually very tedious to do manually.

 

[CHD01]

Malancha:

I'm curious if you have ever tried the old Lockhart Martinelli relationship for 2-phase flow pressure drop; and whether this method gives a result anywhere near the semi-empirical correlation you referenced?

Thanks

The more you learn, the less you are certain of.

 

[malancha]

CHD01,
Yes, I did try the good old Lockhart Martinelli correlation. But you must recognize the fact that the LM correlation was developed in 1940s (if my memory did not betray me). This was developed purely based on lab based results, totally empirical in nature and which allowed extrapolation of the same to a small range only. Modern correlations are semi-empirical in nature which means that they use advanced concepts of hydrodynamics and phase transfer rates. They use lab or field testing data(over a much broader range than LM) to fit some constants (mostly). Definitely in terms of accuracy, they are far better. Some of the modern correlations can even handle complicated two-phase hydronamic effects in transient scenarios. So essentially the choice depends on the need-if it is solely for trivial purposes, correlations like LM are okay. But when capital investment decisions are required to be taken based on the results or where vital operating decisions are to be made, use of sophisticated models are definitely called for. Hope this makes the issue clear.
Thanks.