Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Viscosity of crude-water emulsion?

Status
Not open for further replies.

KernOily

Petroleum
Jan 29, 2002
711
OK I got no replies to this after a week in the Piping and Fluid Mech forum, so I'm posting this one here.

What do you fellers do when calculating physical and transport properties of crude oil-water emulsions? Specifically, this is a crude oil-produced water emulsion of 13° API heavy sour crude.

I did some rough calcs with my simulator (Pipephase) and I get wildly different numbers for a dp/100' pipe friction loss calc depending on which model I use to determine the emulsion viscosity.

The one that most people use is Woelflin. However, it gives HUGE viscosities with correspondingly huge dP's.

Interestingly, the method which seems to most closely match a lot of the field data I've seen is to calc the properties just using a weighted volumetric average. While not very 'scientific' this seems to give decent results.

Well, I am now in the front-end engineering phase of a plant capacity upgrade in which I need to calc properties for a crude I have not dealt with before. It is 13° and sour, like the ones I have field data for, but who knows whether this one will behave like the others?

Just looking for some ideas and discussion here.

Thanks Guys ~ Pete

Thanks!
Pete
 
Replies continue below

Recommended for you

74Elisinore:

If it's mostly viscosity numbers you are looking for, while this may be too simplistic, consider running a viscosity versus produced water content curve assuming there is a varying water content over time.

I would tend to agreethat for other properties, your weighted volumetric average approach is as good as any.

Orenda
 
Viscosity prediction is pretty difficult at the best of times as you will have found from your web search.

Emulsions are in another category altogether.
The viscosity of the emulsion is higher than for the components.

Data from some Fuel Water Emulsion tests shows that for a 380cst (at 50degC) fuel oil with water the viscosity varies from around 8cst at 150degC (5% water) to 20cst at 150degC (40% water).

Now I don't have access to all the data so I can't do more than suggest that 8cst isn't far from the expected viscosity for a 0.5% water fuel.

However, while the data at 150degC shows the 40% emulsion as having a viscosity of 20cst at 150degC, i.e. little more than double that of a 5% water emulsion, at 50degC the viscosity could be as much as 2500cst i.e. 6-7 times the viscosity of the fuel without water.

In other words, I don't trust the calculations for fuel and cutter stock and I certainly wouldn't trust any predictions for an emulsion without a substantial body of work to support it.

I guess it is pretty unusual for engineers to come up with a problem that hasn't been reduced to a formula and a set of rules and corrections but i would suggest that when it comes to predicting viscosities then this is one of those situations.

You've already demonstrated a big lack of confidence in the data and methods so i recomend you do as Orenda suggests: run the tests.

Be sure to get them to cover a range of water % that includes your range and get them to run the tests across the temperature range you expect for the pipeline (you don't want to add any temp Vs viscosity uncertainties either).

JMW
 
Just for the heck of it, I thought I'd throw in the fact that emulsions can have wildly different viscosities depending on which phase is the dispersed phase and which is the continuous phase, i.e. a water-in-oil emulsion vs. an oil-in-water emulsion. The two cases can have viscosities that vary by an order of magnitude or more, and many factors, not only the phase ratio, determine which is which.

Fran McConville
 

I once copied a formula for finding the dynamic viscosity of an emulsion of two components.

[μ] = ([μ]c/[δ]c)[1 + 1.5 [μ]d[δ]d/([μ]c+[μ]d)]​

where
[μ] = the desired viscosity (any units)
[δ] = volume fraction

subscripts
c = continuous phase
d = dispersed phase

I'm sorry I cannot vouch for its precision, nor can I recall the source.
 
Thanks 25362. This is the same formula that HYSYS uses to estimate crude oil-water emulsion viscosity.

Since I originally wrote this, I have gotten actual lab test data of absolute viscosity vs. T of a 60 v/o water-heavy crude mixture-emulsion. It is not a true stable emulsion, rather a sort of mixture that will separate after a few minutes. The lab data was MUCH lower than the viscosity predicted by the above formula, by a factor of 4+. At 195° F, HYSYS predicted a value of 55 cp and the lab data came in at 12 cp. Interesting, don't you think?


Thanks!
Pete
 
74Elsinore:

Interesting but not particularly surprising given the instability of the emulsion, which largely negates developing any meaningful viscosity data.

Orenda

Orenda
 
Orenda sez - "Interesting but not particularly surprising given the instability of the emulsion, which largely negates developing any meaningful viscosity data."

...except for the fact that I must needs have numbers for design work. This type of unstable emulsion is much more representative of what one sees in actual upstream production operations. THe nature of the fluid in the pipe is not emulsion, but rather two more-or-less bulk separate phases whose properties affect one another in some manner. My field dP measurements and observations seem to substantiate this lab data I got as far as prediction of dP goes across a given pipe line using store-bought pressure gauges.

Obviuosly I could not just use the viscosity of water, or the viscosity of crude, for this application. Both would be in error. SO the actual number lies somewhere in between. The lab test data is closer to predicting real-world dP's vs using the too-high numbers predicted by HYSYS or Woelfiln, for example. If I was pumping mayonnaise (a stable emulsion), the results would b far different. Thanks!



Thanks!
Pete
 
74Elsinore - You might want to look for data on Orimulsion. This is a 30% water, 70% heavy oil emulsion sold by PDVSA as a boiler fuel. 0.2 % surfactant keeps the emulsion stable. They have probably done a lot of viscosity work.

HAZOP at
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor