> How can I anticipate when this error will occour.
>
> Why do the precisions ADD when i multiply...?

Because this is arithmetic rules. When you multiple 1.5 and 1.5
you'll get 2.25, not 2.2 or 2.25013.

> What is my solution ?

If you don't need _precise_ arithmetic, cast one of numbers to DOUBLE
PRECISION to get imprecise floating-point result.

SY, SD.