Antares has been designed for the medium and long term study of large interconnected power grids. Its goal is to simulate the economic behavior of the whole transmission-generation system, throughout the year and with a resolution of one hour.
The fact that the simulated period is not located in the immediate future (a few hours, days or even weeks) makes it easy to see that :
It would be meaningless to examine the system for a single set of assumptions matching only one of the countless possible scenarios: months with less rain than usual, or conversely more sun and wind than average, cold spells occurring at times when there are many outages of thermal plants, …
It would be hopeless to try to survey the whole combinatory of all possibilities, even with significant simplifications in the modeling of the physical phenomena at work.
This is why Antares is a simulator of the Monte-Carlo kind, which means that it defines sets of plausible scenarios by carrying out random draws reproducing various events that can affect the system operation throughout the year. The general consistence of the physical assets making up the system itself is defined by a set of assumptions postulated as realistic in the framework of the specific vision of the future that one wants to explore: size of the generating fleets, types and costs of available fuels, ratings of the interconnections, etc.
Moreover, the enactment of each scenario consists in playing one «Monte-Carlo year» from end to end before stepping to the next one, which means that Antares is a genuine sequential Monte-Carlo simulator. As a rule, simulators of that family yield better results than tools that are not fully committed to accommodate the actual chronology of events.