The paper presents stochastic optimisation technology as applied on separation sequencing problems. Simulated annealing is employed along with a synthesis representation in the form of a dynamic vector. Permutations on the vector and perturbations on the current state generate new sequences and layouts that are accessed, accepted and rejected. The approach can handle problems of industrial complexity. Operational variables (pressures, reflux ratios, product composition) can be optimised simultaneously with the structure. The potential of the methodology is illustrated with an FCC example.