Estimation of Viscosity of a gas mixture

[JasonNicholson]

I am an entry level engineer. As an exercise, I was asked to confirm an orifice plate sizing calculation using International Organization for Standardization (ISO) 5167-2. The gas is natural gas with the following composition:

Methane, CH₄[tab] 98.26%
Nitrogen, N₂[tab] .79%
Ethane, C₂H₆[tab] .63%
Propane, C₃H₈[tab] .22%
iso-Butane, C₄H₁₀[tab].04%
n-Butane, C₄H₁₀[tab] .04%
I-Pentane, C₅H₁₂[tab] .01%
n-Pentane, C₅H₁₂[tab] .01%
n-Octane, C₈H₁₈[tab] .01%

Design temperature and upstream pressure are 365°F and 411.2psi gauge. The Fluid data for the natural gas is very limited; we do not have viscosity of the whole fluid. However, we do have properties for each of the individual gases.

1. How do you calculate the viscosity of the mixture given the composition?
[tab]1a. Searching the web, I found the Wilke Method mentioned. Is it good estimation? Where can I find information about it?
2. Is a good estimate to use the viscosity of methane since it is 98.26% of the mixture at 365°F and 411.2psi gauge?

-Jason Nicholson

 

[dcasto]

The GPSA data book has a method useing tables and calculations, Chapter 23.

From the GPSA book, I get a viscosity of .0156 Cp at 14.696 psia, and .0165 Cp at 411 psig


From my similator useing API, I get .0163 Cp

 

[Fudge30]

Id just use the viscosity of methane - you won't get an accurate value of viscosity - even HMSO publications state that.

If you want to do a slightly better job - just do a weighted average i.e 0.9825 x visc. CH4 + 0.0079 x visc. N2.....etc.

To go further into it, all the chemicals are going to be relatively non-polar, no H-bonding etc. so I wouldn't expect any interactions that would effect the overall viscosity.

 

[25362]

For pure methane NIST gives 0.0161 cP.

 

[vzeos]

I would also use the methane viscosity as Fudge30 suggested. I’m not familiar with the Wilke Method but I have used the Carr method given below.
[tt]
[µ]_m = {[Σ]_i x_i[µ]_i[√]M_i}/{[Σ]_i x_i[√]M_i}

[µ]_m = viscosity of gas mixture
[µ]_i = viscosity of the i1 component
x_i = mole fraction of the i^th component in the mixture
M_i = molecular weight of the i^th component
[/tt]
Carr, N.L., "Viscosities of Natural Gas Components and Mixtures." Institute of Gas Technology Research Bulletin No. 23 (1953) June.

 

[dcasto]

The GPSA graph is look up pressure vs Temperature at gas gravity. If you are really good at excel, you can do a polynomial regression.

 

[JasonNicholson]

Thank you for your help with this problem! I am going to print this and file it away for use through out my career. Again thanks.

-Jason Nicholson