I read your post from a few days ago: for some reason your system's logic addresses bit-level instructions using bit sub-elements from inside a REAL.
http://www.plctalk.net/qanda/showthread.php?t=133685
That is a bad method, and the sort of confusion that it creates is probably why RA removed the ability to use it.
From my reading of the problem, the best way to resolve this issue is to edit the program to use a UDT that mixes arrays of BOOL and DINT and REAL elements, rather than attempting to pack BOOLs into bits of a REAL.
Please post your replies to that original thread to keep the information in one place.