Cotton yarn is wound onto bobbins, each of which takes $100$m of yarn. If the thread breaks before $100$m is reached, the bobbin is rejected. In a trial of a new spinning machine, $13$ bobbins out of $81$ are rejected. Calculate a $90$% confidence interval for the proportion of bobbins rejected.
I think when you convert to Normal it's $X~N\big(\frac{13}{81},\frac{13}{81}\big)$ but I'm not sure what to do from there.