With the current highly dynamic development of renewable energy sources (RES) in Chile, an accurate prediction of feed-in from wind turbines and photovoltaic systems is becoming increasingly important. Especially in sparsely populated northern Chile, there are good preconditions for (partial) energy autonomy of individual industrial plants and even entire regions. ZSW aim to work on economic viability of these conditions as part of a specific feasibility study. Experts of the Systems Analysis department will expand an existing wind power forecast prototype together with a newly developed solar power forecast to form one operational system. For this purpose, a RES forecast for Germany developed by ZSW is available, which is currently being further improved in the framework of various projects. A medium-term forecasting system adapted to the local geography and meteorology is being realised in cooperation with the Chilean partner Universidad de La Serena using several weather models and measurement data, combined with advanced methods from the field of machine learning. Its usefulness is not limited to the prediction of feed-in for grid control and trading purposes or future upcoming regulations of the electricity grid. It is also being applied to improve the (semi-) autonomous RES supply of a selected sub-grid in northern Chile. The hybrid power plant simulation and optimisation model P2IONEER developed by ZSW is being used as well, which allows for the determination of an optimal mix of renewable energy sources taking into account the fossil-fuel power station pool, transmission capacities and energy storage systems. The aim of the study is to achieve a signalling effect for the expansion of RES in Chile.