Fig. 4: Resilience of a globally interconnected system to extreme or unforeseen conditions.

a Impact of climate-induced variations in power generation and demand on solar-wind penetration at different stages. Maximum, 90th, and 50th percentiles of changes observed across 10,000 simulations are presented. b Reduction in global solar-wind penetration under S-I, S-A, S-C, and S-G scenarios when one of the 20 regional grids experiences a complete outage. Each circle represents a separate simulation in which a different regional grid experiences an outage, resulting in n = 20 simulations per scenario. Each simulation has identical modeling settings except for the grid where the outage occurs. Boxplots represent the median (center line), the 25th and 75th percentiles (box bounds), and the minimum and maximum values (whiskers) of outcomes across these 20 runs. c Changes in solar-wind curtailment and alternative energy supplies when one of the 20 regional grids (denoted by bar colors) deploys solar-wind resources based on the S-I scenario. The horizontal gray inset illustrates the extreme case where all 20 grids remain locally interconnected. d Changes in global solar-wind curtailment and penetration when all transmission lines associated with a single regional grid (denoted by bar colors) are disrupted. The horizontal inset depicts the extreme scenario where all transmission lines fail simultaneously. e, Power supply gaps resulting from supply interruptions from interconnected grids (left panel) and aggressive market competition (right panel). Circles represent scenarios of a single grid disruption, while bars indicate simultaneous disruptions across all interconnected grids. Circle colors denote disrupted grids, while bar colors indicate affected grids. The colors of circles in (b) and (e), bars in (c), (d), and (e), and lines in (e) correspond to the legend at the bottom.