In the practice of energy renovation of existing buildings, the uncertainty of renovation parameters has a significant impact on the renovation results. To support energy renovation decision-making, a Monte Carlo method combined with Latin hypercube sampling was proposed to evaluate different renovation schemes, and a tree-based Gaussian method was used to screen key variables that affect the renovation process. The results show that uncertainty analysis can quantitatively evaluate renovation schemes at the preliminary design stage. The energy-saving rates of two typical renovation scenarios fluctuate between 32.7% ~ 55.2% and 55.5% ~108.4%, respectively, with higher uncertainty in schemes with better energy-saving effects. The cumulative probability distribution was used to assess the probability of renovation success, with the probabilities of meeting renovation targets being 58% and 96.4%, respectively. The integration of renewable energy technologies ensures the renovation results. Sensitivity analysis results show that infiltration rate and equipment power density are the most important factors in office building energy consumption, accounting for 80% of the output variance, which provides a theoretical basis and methodological reference for the selection of more building energy renovation schemes in the future.