You need to figure out which columns are extra and delete them from the import

I might also use the header info from the 7.2 when doing this just in case that says something that the import uses

thanks guys

has anyone done this ?
how would you do this please.. I`ve tried comparing the columns of the csvs but it`s nightmare.

Has anyone done this - Yes
How would you do this - Carefully

The column headers have numbers as well as names. You will find that all the new columns in rel 8 have higher numbers.

Find the highest number on 7.2 and delete every column with a higher number from release 8

There is no easy way.

For best practice of comparing COS (or anything similar) that is in rows is to copy and transpose past them into a column format.
In this case, do it for rows of the ID number and the name for easier reading and into a new sheet.
Paste one set of COS options to column A and B, while the other to C and D, for example.
Use a simple comparison method as follows. Assuming your data starts in A1.
Type into cell E1: =if(a1=c1,"-","No")
Copy this to cells E1 to F1 and then down to the last row with data.
This compares the ID number and the description separately and shows a dash if the records match and 'No' if they don't.
All you have to do now is to scroll through the list and remove all the columns from the actual data of 8 that does not appear in 7.2.

Then comes the hard part, fixing the mess that Excel creates when saving 'csv' data.

Cheers guys