This paper targets to show a practical realization of such contemporary notions of modern industrial IT as ‘digitalization’ and ‘big data’ in referent example of energy producing enterprise. Huge historical databases of technological parameters accumulated by modern industrial control systems has transformed to a precious asset of modern companies. Deep knowledge extraction approach applied to the accumulated data can bring additional benefits. The research grounds on classical methodology of mathematical simulation. Using a programmatic realization of numerical Monte-Carlo procedure we create empirical simulation models of real energy producing assets based on their observed behavior. As a result we are able to conduct so called ‘what-if’ analysis and compare different production scenarios for modeled assets. The investigation approach includes the following steps: approximate the output power characteristics of the power generating units; calculate cross-correlation for all output powers with all available measurements collected by power station SCADA system; create several day working scenarios for research; for each possible approximation and all combinations of influencing parameters conduct Monte-Carlo recalculation sessions to define the scenario that gives better day margin for power plant.