Co-current liquid/vapor Flow

[ME1994]

Does anyone know of any articles that discuss co-current liquid/vapor flow.

I am looking for any information that would be useful in determining liquid/vapor pressure drops.

Problem:
Determine the liquid and vapor pressure drop across a flat plate with (2) slots in it. For this example, lets assume that the plate is placed horizontaly in a vertical 48" i.d. tube. The plate is the same diameter as the tube and is sealed at the tube i.d. (2) slots have been cut-out of the plate to meter the liquid/vapor flow.

 

[zdas04]

Measuring a vapor with entrianed liquid is a really hot topic in the Gas Measurement literature. Problem is that after several million words on the subject, the conclusion is the same as it was 25 years ago "any correlation is going to be wrong by over 40% in a unpredictable direction". When you dig through the papers and theses' that claim good results, you'll either find that they calibrated the correlation for exactly one set of conditions (one pipe size, one slot size, one pressure, one temperature, one multi-phase flow regime, etc.) and then demonstrated that they could measure those same conditions multiple times or they never got a good correlation.

SWrI in their San Antonio facility has had a dozen projects funded in the last ten years and none of the projects has resulted in a commercial product (and this is the area clammering the loudest for a product that I know of--it would let you get rid of offshore gas/liquid separation equipment).

You don't say if you are trying to evaluate liquid in gas or gas in liquid. The former has thus far elluded solution. The later can be approached as a pretty straightforward incompressible flow problem as long as the gas cut is small enough to allow an incompressible assumption to be ok.

 

[KernOily]

Best place to start is with a text on two-phase flow. You can start with the Campbell Petroleum series for a good brief practical treatment. There are beau coup other texts and papers out there on two-phase flow.

In estimating pressure drop there are gobs of different correlations out there but, as David said, each has applicability only for a particular set of fluid and hardware parameters. FOr example, the Lockhart-Martinelli correlation is used all the time, for all kinds of systems, but that one was developed using an air-water mixture in 1/2" pipe, I think it was, for a certain set of flowrates and liquid-vapor fraction. You would be going out on a limb extrapolating that correlation for any other application, although people do it all the time.

In heavy oil steamflooding, which uses 70% quality steam at about 550 psig, we have gotten reasonable, defensible answers for dP and quality prediction using BBM and then applying a fudge factor based on field data. That usually gets us within about 5-10% wich is pretty darn good in this two-phase flow business, certainly good enough for design purposes.

Hope this helps. Good luck. I have spent years studying this and sometimes I think I'm no further ahead than when I started.

I don't know how critical your application is, or how closely you have to nail the dP prediciton, but you might want to make a model, do some testing, and then try to fit your data to one of the correlations that are already out there: BBM, Xiao, BB, Dukler, etc., ad nauseum. Thanks!
Pete

 

[KernOily]

Best place to start is with a text on two-phase flow. You can start with the Campbell Petroleum series for a good brief practical treatment. There are beau coup other texts and papers out there on two-phase flow.

In estimating pressure drop there are gobs of different correlations out there but, as David said, each has applicability only for a particular set of fluid and hardware parameters. FOr example, the Lockhart-Martinelli correlation is used all the time, for all kinds of systems, but that one was developed using an air-water mixture in 1/2" pipe, I think it was, for a certain set of flowrates and liquid-vapor fraction. You would be going out on a limb extrapolating that correlation for any other application, although people do it all the time.

In heavy oil steamflooding, which uses 70% quality steam at about 550 psig, we have gotten reasonable, defensible answers for dP and quality prediction using BBM and then applying a fudge factor based on field data. That usually gets us within about 5-10% wich is pretty darn good in this two-phase flow business, certainly good enough for design purposes.

I don't know how critical your application is, or how closely you have to nail the dP prediciton, but you might want to make a lab model, do some testing, and then try to fit your data to one of the correlations that are already out there: BBM, Xiao, BB, Dukler, etc., ad nauseum. One problem there is that gravity effects may negate your accuracy for scale-up/similitude of the lab model, assuming you haven't built a full-scale model.

Hope this helps. Good luck. I have spent years studying this and sometimes I think I'm no further ahead than when I started.
Thanks!
Pete