RON calculations from component basis

[Mutt]

Does anybody have info on techniques and data for predicting by calculation of RON / MON values for mixtures of hydrocarbons based on the detailed compositional analyses? Mixtures of aromatics / olefins / paraffins and their interactions are of interest for tailoring final gasoline blends.

 

[22082002]

If you have detailed component analysis,
RONC for mixture =( Sum(RONC of component 'i'*volume of component 'i' ) where i 1 to n) divided by total volume of mixture. This works for similar components you mentioned above.

MON is difficult to predict.I have not tried, but you can try some books like 'refinery economics' (sorry for not recalling author's name).

If you have ProII or similar simulator try blending operations.

 

[Mutt]

Hi 22082002 :
The volumetric blend method proposed works more or less for components of same species - however in mixtures of 30-40 components of olefins, aromatics, oxygenates, paraffins some very divergent answers from the simple volumetric blending become evident! Even using blending RON/MON values doesn't give sufficient accuracy.

I suspect this is going to be proprietary info and not readily available - def. not found on the web anyway - have spent many hours of searching and emailing with no success.

I have developed own technique and still collecting data to verify fully but curious to know if other methods compare in procedure.

 

[22082002]

Hi Mutt,
when I used to do this things, I got very close result, I had only olefins, aromatics and paraffins(i & n).
I also suggest you to use the RONC-values given by GC, in laboratoy.

Many sources give different RONC for same component, so it is also impotant, to which you refer.

Regards
SAA

 

[enbee]

I'm working on the implementation of In-line blending in my refinery.The simplest method for RON calculation is linear by volume , where the RON value is given by

Blend RON=
(RON1*volume fraction1)+(RON2*volume fraction2)+...(RON n*volume fraction n)

for a n component blend.
Another method is using availabe correlations like Healy's correlation.
RON Index =3.205+(0.279*EXP(0.031*RON))

where RON is the volume percentage of the component in the blend.Calculate the RON indices of all the components (of course,knowing their RON) and find the volume average RON Index.
Then invert the above function on the volume average RON index to get the RON of the blend.Simple! I do not know how well the above works with MON.I guess its designed from RON data so any fit would be a coincidence.
Indexing methods like these based on empirical data.So they work best with the crude that they are based on.
I'd like to know what you people think and if you have any other ideas/discoveries.Lets share.Make the most of a place where learning doesn't cost an arm and a leg.
Any other blend engineers reading this? get in touch with your experiences.Maybe we can start a separate forum.

 

[seeksense]

glad to come across this thread. The web is quite thin on gasoline blending calculation methods. Linear volumetric blending of RON and MON has worked well in my personal experience for predicting the finished blend. I like the idea of starting a blending forum.

 

[phex]

Although I am no blending engineer, and my experience in this area is limited, nut I found that blend calculated this was (linear ly by volume) often give results as much as one to two points below the real (meaning tested) value. One way to fix this is by using socalled "blending MON"s. data like this is quite difficult to get (ask a friendly refiner, perhaps you're lucky...) but give much better results.

If interest about blending exists, there is a small booklet by HPI called "Guide to Petroleum Product Blending" which may be helpful. You can order it here:

http://www.hpiconsultants.com/blending/blendguide.htm

hth,
chris