Bioprocess deviations are likely to occur at different operating scales, leading in most of the case to substrate deviation from main metabolic routes and impact product synthesis. Correlating qS and qP is of utmost importance for bioprocess observability and control and can be modeled actually by advanced metabolic flux models. However, if most of these models are able to make prediction about metabolic switches, they still do not incorporate deviation due to biological noise, i.e. phenotypic and genotypic heterogeneity. These limitations impair observability and thus the use of fundamental knowledge about biological network for practical application, i.e. metabolic engineering or bioprocess scale-up.