ABSTRACT
Percolation theory and fractal geometry play an im portant and increasing role in elu cidating the behavior of complex systems. Both concepts are related to the observa tion of critical phenom ena and the problem of non-robust formulations. Thus it is im portant to identify critical concentrations of solubilizing agents, which are related to percolation thresholds in a percolation phenom enon for an optimal design of qual ity. Due to finite size effects such thresholds may not be very sharp and thus may remain hidden for a long time. However, a small change in the quality of one of the components may induce a large change in the system property if the concentration of that com ponent was close to a percolation threshold. To avoid such situations using a formulation approach based on the concept of percolation theory and abandoning as much as possible the so-far most popular “trial and error” m ethod are recommended.