Viscosity Correlation Calculation 1

[Falcon03]

Hello,

I would like to have some sorts of correlation to calculate the Viscosity at different Temperature, since we measure our Viscosity for Heavy Fuel Oil @ 50 °C.

I would like to know the expected viscosity at 100 °C. Thank you.

 

[Falcon03]

Let me add, I will use the correlation for the Fuel blend calculation. tks

 

[25362]

There are various charts in the market to help you out.
In order to be able to find viscosities of fuels at various temperatures one needs to know two viscosities at two different temperatures.

By using charts such as the BP viscosity/temperature chart or the Refutas chart one can find the wanted value at other temps. This is needed since the VI (viscosity index) of the fuels isn't known. The VI or V/T dependence varies with chemical structure of the hydrocarbons. You probably heard of paraffinic (high VI) and naphthenic (low VI) lubes.

The BP chart, for example, enables converting among various viscosities such as Redwwod I, Saybolt Universal and Saybolt Furol, all of them measured in seconds, to degrees Engler or to kin. visc. in cS. One can use either ^oC or ^oF. The BP chart is helpful in making blends on a weight basis.

One can also use the ASTM Standard Charts for Liquid Petroleum Products (D341) that give kinematic viscosities vs temperatures. There are charts for low and high viscosity ranges.

Perry suggests for visc. blending of non-polar liquids:
ln visc_blend = summation of (w_i x ln visc_i) with a 5-10% precision,
where w_i = weight fractions of the components.

Coming back to your question, measure the viscosity at another temperature, and then extrapolate or intrapolate as needed to get the value at 100^oC. Kinematic viscosities appear to be loglog dependent on temperatures.

 

[jmw]

Go to

http://www.solartronmobrey.com

and follow the bunker fuels link and downlaod the Bunker Fuel ASTM.xls spreadsheet. This is quite useful but you may have to try different values for the viscosity at 100degc till you get appropriate A & B values.
However, there is a program available from Det Norske Veritas Petroleum Services in Holland which is free and which allows you to calculate all sorts of related factors. Sorry, you'll have to find their contact details from their web site.
You can also go to

http://www.steamesteem.com

where there is a shareware program you can download.

 

[jmw]

PS. Norcross (

http://www.viscosity.com)

has a table of conversions from all sorts of instrument values to international units.
The function we have found to be the most reliable for everytiing from quench oil to bitumens (asphalts) is the ASTM D341 equation log10.log10(v +0.7) = A-B log10(T+273) wher V is kinematic viscosity in cst and T is the temperature in degC.
When it comes to fuel oils we have seen all sorts of functions used with widely differing results!

 

[jmw]

PPS: the best way to blend fuels is to measure the viscosity of the blend and use this for feedback control. There are now a number of very affordable solutions for this. Using calculations has too many unpredictable assumptions which can lead to widely varying results. Fuel is blended in the refineries at 90 to 130degC referenced to 100deg C and measured using process capillaries backed by laboratory capillaries (ASTM D445)In terminals and on barges the blend temperature is usually between 30 and 60degC with 50degC as the reference temperature, something CIMAC is apparently going to codify (see the CIMAC web site and fuels working groups pages). On barges it is usual to use Hoffler falling ball viscometers and hydrometers, if they do not have online continuous vioscosity measuremnt. As is known from the problems that were highlighted in Singapore a year or so back, this is not always reliable nor is the sample taken for independent analysis as reliable as it should be because of the vulnerability of this systme to operator "error". The

http://www.solartronmobrey.com

spreadsheet is set up using the CIMAC fuels data.
If you are blending fuel and measuring viscosity at 50degC i presume you are using the Hoffler falling ball viscometer and a hydrometer method. These can give good results but by the time you get them, it is often too late. Then too you are basing your calculations on samples. Samples are a problem. Spot samples are unrepresentative. The same applies to the three samples taken during bunkering (begining middle and end) which are averaged. Continuous viscosity measurement online revealed a significant viscosity variaton during blending which suggested that the required standard of 180 +ocst -10cst was not easily met. Continuous drip samplers tend to be biased toward a lower apparent viscosity because they generally tend to collect more sample at lower viscosity.
A number of bunker suppliers we have spoken with report that calculations based on the assumed viscosity of the fuel in the storage tanks can vary significantly due to stratification. Overall, the cost per ton of heavy fuel oil due to inaccurate blending based on calculation is more than $3. Not insignificant.
You may care to visit some other sites such as

http://www.chl.chalmers.se

. They have some studies based on fuel delivered over a two year period to a ship. CBI Engineering in Denmark produce fuel blenders as do ALBA shipping. Also,

http://www.Jiskoot.com

And

http://www.VAF.nl