When supplying energy from renewable sources, there is the question of intermittence of the different energy sources. Indeed, when there is no wind in a country, all wind farms are usually still which can induce an interruption of the service. A proposed solution to overcome this issue, is to diversify the locations to take advantage of the fact that when one energy source does not provide enough power, the other may be still productive. The first task of this project is to identify potential locations in which one could build an energy source (typically solar or wind) and to collect data about the production that could be obtained at various time periods with different weather conditions. This task has been already done and is collected in files. These files are matrices of potential locations with time periods. Since, no energy source has been yet built in these locations, we rely on a very rough assessment of their production through a binary value : 0 if the location would not produce enough in the time period, and 1 if the location would produce enough in the time period. This project aims to find a set of locations (less than 10% of the total number of potential locations) such that the number of time periods in which you have a certain number of locations providing enough energy is maximized. Several parameters were used in trials (the threshold to consider that a time period has enough energy produced for example), and the budget of locations that we can select.