The fuel oil portion of ethylene cracker effluent is an incredibly complex, wide-boiling mixture. It is full of olefins, diolefins, triolefins, aromatics, polynuclear aromatics, etc. throughout the boiling range. I would guess there are probably millions of compounds in such a mixture, 99.99% of which can only be guessed at.
For design of the so-called gasoline fractionators (right after the furnace transfer line exchangers), highly simplified methods are used to characterize the heavy fuel oil portion. Typically, a true boiling point (TBP) distillation curve is assumed for the fuel oil components. This is based on the predicted yields from a pyrolysis program. The TBP curve extends all the way to 1500 degF. The shape of this curve is uncertain and much experience is required to obtain the proper shape.
Then, as in petroleum refining, pseudo-components are generated and used in column simulation. Unfortunately, the methods used in most process simulators to estimate the thermophysical properties of these pseudocomponents were originally developed for straight run petroleum fractions, not highly olefinic mixtures of the type you find here. So, the simulation results can be considerably in error, but it's hard to know in advance.
I hope this reply sheds some light on the incredible molecular complexity of the fuel oils from ethylene crackers.