CTCs are found in the region where $r < 0$ . That should be just inside the ring singularity, since in Boyer Lindquist coordinate system $r = 0$ means ring singularity. Does that mean this image showing CTC outside the ringularity is wrong, or I am missing something?
You are missing something. In a text where $r < 0$ is possible, the singular ring is a curve, not a surface: its equation in B-L coordinates is $(r = 0, \cos \theta = 0)$.
In contrast, $r = 0$ is the equation of a surface: it consists of two disks $(r = 0, \cos \theta > 0)$ and $(r = 0, \cos \theta < 0)$ together with the singular ring, their common boundary. The 2 disks are made of regular points: they are actually 2 manholes through which you can reach into negative space.
The picture is ambiguous: if the text from which you got it claims the CTC is in negative space (not sure, the pic is clipped), then it lies below the region $r = 0$ rather than just outside.