If you are expecting double, decimal, float, integer
why not use the one which accomodates all namely decimal (128 bits are enough for most numbers you are looking at).
instead of (double)value
use decimal.Parse(value.ToString())
or Convert.ToDecimal(value)