Global climate change not only causes extreme weather events but also impacts the consumer's energy demand, posing a concern for the safe and reliable operation of power grids. Furthermore, the high adaption rate of electric vehicles makes energy transactions volatile. Hence, it is crucial to evaluate the power grid's reliability using risk assessment studies. Here, we solve this open problem to predict the risk of blackout by developing a real-world data-driven probabilistic framework in conjunction with an expert-based power grid methodology. Utilizing the real-world energy demands, climate station's temperature data, and IPCC regional climate models, the risk of a city in California increases up to 8% in the summer of 2100. The winter, spring, summer, and fall seasons' risks also follow an increasing trend from 2000 to 2100. We also show that a better localization planning of load hot spots like EV charging stations reduces the risk by 1.8%.