How do I change a single letter in a large text file

I have a large file file.dat, a database dump in a CSV format that looks like this:

file_id,line_id,st_x,st_y,value1,value2,...value100
111111,1,22.22222,33.33333,0.1234567,0.246890,0.369258,...
.....

It has several million lines.

Suddenly I realized that I want to change the names of the columns: change st_x to lon and st_y to lat.

Just copying the file on my machine takes some time T_copy.

Editing the file with

sed -i 's/st_y/lat/' file.dat

takes T_sed where T_sed ~ 1.7 T_copy.

Is there a faster method? I would not mind if the replacement string must have the same number of characters as the target string.

Note: I am sure this question has been asked before, just do not see how to find it.