A risk bidding methodology is proposed to help prosumers formulating optimal quantity-price bids for the day-ahead energy market. A prosumer is the manager of a Low Voltage (LV) Micro-Grid (MG), connected to the main electric grid, where generators are paired with renewable energy sources (RES). To present the optimal bidding in the wholesale electricity market, the prosumers need to resolve a short-term management problem and need to identify all influencing variables (i.e. energy exchange, internal production, level of storage, Photovoltaic power plants (PV)). They also have to take into account the uncertainty in RES energy production to evaluate different risks associated with their tolerance preferences. A heterogenous MG which pairs traditional thermal and electrical generators with a PV power production is simulated. An economic model based on genetic algorithms is proposed to formulate the optimal bidding. Although in literature it is possible to find similar decision support models, one of the main original contributions of this work is to estimate the RES input of the proposed model with Analogs Ensemble (AnEn) approach, which is used here to provide day-ahead PV energy forecasting. The results of the model are analyzed evaluating the risk associated with the different prosumer's choices by the expected utility theory. The analyzed case study uses on residential MG and different prosumer risk tolerances (adverse, neutral and incline). Results are shown to demonstrate the effectiveness of the proposed methodology.