6 replies to this topic

[dhns]

Dear all,
Could you any one explain about my issue....i have run one issue in my HYSYS
in one spot in one stream i seen two different liquid phases (liq 1containing ~10 Cp of viscosity & Liquid 2 showing the viscosity value of ~950 Cp )it is correct for Practical cases ?how i can take a correct viscosity value ?
please Explain .,

Thanks in advance
Regards,
Dhnsekaran.

[Zauberberg]

Can you upload the simulation file?
Once I have encountered the same issue - appearance of a second liquid hydrocarbon phase - it was coming out from extremely heavy crude oil (though it was a full-range crude), rich in aromatic compounds. The values of density and viscosity of heavy hydrocarbon phase were enormously high. And I'm still not sure whether it was a mistake in Aspen, or it was a reflection of asphaltene precipitation in the crude oil (there are no such correlations in HYSYS, as far as I know). At that time, I remember that I prepared a report for AspenTech support desk, but then the client had provided a new source for fluid characterization and the cases previously built were simply discarded.

Maybe some other forum members have had similar experience, and hopefully they can help us to resolve the mystery.
Best regards,

[JoeWong]

I think this is commonly known of phase separation. In long distance pipeline with multiphase flow i.e. condensate and produced water, phase separation may occur and results three phase flow (vapor, condensate and produced water).

Phase separation is affected by density, surface tension, viscosity, velocity, etc

I have even encountered another complicated situation. The fluid contains

• vapor
• condensate
• produced water
• sand
• mercury

There is no easy answer to your question. There are number of way to manage it.

You may consider 3 phase flow. Vapor, light liquid and heavy liquid and use their dedicated viscosity for fluid flow.

You may consider to ignore the light or heavy phase with low flow and take the viscosity for high flow.

You may consider light and heavy phases are homogeneous and apply the avearge viscosity method. There are number of model for average viscosity method i.e. volume average, mass everage, momentum average, API method, empirical method.

You may consider 3 phases are homogeous and apply the average viscosity method.

I only can advise how i look at the issue and what have been done. However, i don't have any clue to tell you which method to use. There are plenty of information and analysis required before a proper method is used.

Good luck.

[Zauberberg]

Joe,

I believe DHNS was referring to the appearance of 2 liquid hydrocarbon phases, and not water as the 3rd phase in the system. The viscosity og 950cP can be attributed only to some high-molecular weight, multi-ring hydrocarbon mixture, in my opinion.

[JoeWong]

Zauber,
You are probably right.

My early statement refer to light liquid and heavy liquid. Both liquids could HC-HC, HC-H2O, HC-Heavy metal, etc...

[daryon]

Joe,

I believe DHNS was referring to the appearance of 2 liquid hydrocarbon phases, and not water as the 3rd phase in the system. The viscosity og 950cP can be attributed only to some high-molecular weight, multi-ring hydrocarbon mixture, in my opinion.

Hi,

You should try adding water to this stream so that it is slightly over saturated with water and therefore creates an aqueous phase. This should force HYSYS to combine the 2 liquid phases and report gas, liquid, and aqueous phases.

Saturating with water will change the compositon of the hydrocarbon phase, i'm not sure how much this will affect the viscosity. I think this approach may give you an idea of the predicted properties of the combined hydrocarbon liquid phases.

Alternatively, you can reduce the number of phases that HYSYS reports. >>>>> Go into the simulation basis manager and into fluid packages tab. Click view on current fluid package (probably PR for oil and gas applications), then have a look at the 'StabTest' tab. In the stability test parameters it shows the maximum phases allowed (should be 3), if you change this to 2 it should force HYSYS to combine the 2 liquid phases, so the stream in question reports just gas and liquid phase. This approach may give you a more accurate representation of the predicted properties for the combined hydrocarbon liquid phase.

P.s. HYSYS uses viscosity prediction models (Twu or modified Ely & Hanley) based on the corresponding states principle which are modified by Aspentech. There is an error associated with these prediction models. The errors become worse if hypothetical components have been used in the oil composition as HYSYS first has to predict the critical properties for the hypo's before the viscosities are predicted.

Gernerally if you have actual viscosity data avialable I would recomend you ignore HYSYS viscosity predictions. If you only have a couple of viscosity-temperature data points ASTM D341-03 method can be used to extrapolate the data.

(Probably too late to help now, but thought i'd post anyway)

[JMW]

Re last post,
the spreadsheet below lets you calculate viscosity at different temperatures based on the ASTM D341 equation.

Attached Files

•  RMI_ASTM_D341_rev01.xls   962.5KB   169 downloads