The thread 21121956 refers to makes an interesting read.
I should say that Electricpete's comments on the spreadsheet where explained in another thread.... The spreadsheet calculates the values of A and B from the input data, it does not look these values up and the input temperature may be in degC but the calculation converts to Kelvin.
The point Skogs was making about his discussions with Shell.
I think he meant that there was no equation that allowed you to find the viscosity at a temperature knowing only the viscosity at a single temperature; you need the viscosity at two temperatures to solve the Walther equation (which is the basis of the ASTM D341 calculation) but the more accurate description of temperature viscosity behaviour is in the Vogel equation.... but this is even less handy to use.
The problem for process engineers is that what they'd like to do is measure a property online at whatever conditions exist and calculate the property at a reference temperature.
You can dot his for density, for example, and many other properties, but not viscosity. It complicates process measurements no end and either you control the temperature as with the process capillary viscometers which have a temperature bath, or you are looking at multiple viscometers measuring at different temperatures.... a reasonable option when dealing with the Walther equation, you need two viscometers, but far less practicable when dealing with the Vogel equation.
So back to Skogskurra's discussion with Shell. This is kind of ironic becuase SHell produced an equation called the Shell V50 equation which was designed to produce the viscosity at 50deg C from the viscosity measured at one other temperature..... but for fuel oils. Skog's discussion was about lubes. Lubricants are complicated by the manipulation of viscosity index so you could have two oils with convergent temperature viscosity curves so that at one temperature they have the same viscosity.
With fuels oils this doesn't happen. They exhibit a vi that changes with the viscosity a 50degC in such a manner taht we can aproximate the calculation of viscosity at one temperature, based on the measurement of viscosity at another temperature, reasonably well - provided we stay within certain guidelines... this works fine for fuel oisl where the reference temperature is 50degC and the measured temperature is somewhere bewteen 40 and 60degC (e.g. on barges, in terminals) but not where the process temp is 90-130degC (e.g. in refineries).
JMW