As I understand the vessel is already there, so are known flow rates, temperatures, densities, viscosities, interfacial tension, etc.
The (non-identified) oil's hold time will depend on:
(a) as indicated above, the "separation" zone holdup volume but also on
(b) the volume of the supernatant oil phase between its high and low levels (alarms?), i.e. the expected operating surge volume range, which will depend on the processing purpose of the buffer tank.
I suggest you consult books on liquid-liquid separation to estimate (a), and tell us something about (b) to enable contributing some thoughts.