So, the real issue is the sensor and I have no clue...
If you try to do it optically, my fear is that the oil will accumulate around the sensor (or around the glass, etc.) and that would foul-up your measurements... In fact, that might happen with any kind of sensor.
If you don't need continuous measurements the best solution might be to use a filter (or use the existing filter) and weigh it (with a precision lab balance) before & after to see how much oil you "catch".
And of course, oil is not a gas. 
There are gas & particulate components of burned oil and there may be gas components of evaporated/boiled oil.