Prediction of normal paraffins in Kerosene

[sgoel]

Hello All

Is it possible to predict normal paraffins in a Kerosene stream with any of the physical property parameters. Purpose is to have simple go/no go test for use as a feedstock.

Thanking you in anticipation please.

Best Regards

 

[mbeychok]

sgoel:

There are too many different ways of producing a kerosene blend with essentially the same physical properties. Therefore, I don't believe you could depend upon being able to predict the normal paraffin of a kerosene by using the physical properties. I think you would have to perform a PONA (paraffin-olefin-naphthene-aromatic) analysis ... but I could be wrong.

Milton Beychok
(Visit me at www.air-dispersion.com)
.

 

[25362]

Try ASTM D3238-95 (2005).

 

[mbeychok]

25362:

From what little I know about ASTM D3238, it is not to be used on oils which contain olefins. Some kerosene blends do use some FCC cracked materials and would therefore probably contain olefins.

The test also seems to have other restrictions as to the amount of ring structures, etc. It seems as if you must almost know the PONA of the kerosene before you can ascertain whether the test can be used or not.

I still think that SGoel will need to run a complete PONA rather than depond on physical properties.

Milton Beychok
(Visit me at www.air-dispersion.com)
.

 

[25362]

Since sgoel referred to kero as a potential feedstock (to what purpose?) I -probably wrongly- assumed a straight run kero was the query's aim.

On re-reading it, it is now not clear to me whether sgoel wishes to know the individual % of n-paraffins, branched paraffins and cycloparaffins (naphthenes) or otherwise just assess the content of paraffinic hydrocarbons present or, probably preferentially, the % of non-paraffinic hydrocarbons.

A bit of more information by sgoel would help us to better grasp the query's underlying motive.

 

[stevethermo]

?Lacking any better method, you may be able to do some screening to reduce the number of possible feeds, using data that would usually be available, such as the density and the cloud point or pour point. If a MW or boiling point curve is also available, the Watson K factor (or UOP K)with the cloud point or pour point could be used. The Watson K (or just the density if the samples are from similar sources and have similar boiling ranges) would give you an estimate of the total paraffins, and the pour point would be an indicator of Normal paraffins. Once you have run PONA’s on a few samples, you could refine the calculation a bit, and probably have a reasonable idea whether a feed would be a good candidate. This will work best if you have a data on what has run well in the past.