Strongly correlated and spatially concentrated curtailment of power plants strongly affects the electricity market. Such curtailment is observed during heat waves in middle Europe, for example. First, curtailed power plants need to be substituted by more expensive ones. Second, additional congestion of the electricity grid may constrain substitution. These consequences and their spatial incidence have yet not been thoroughly assessed at the level of a national electricity system. Does congestion excessively amplifies curtailment costs? Do costs remain localized? How does the cost incidence depend on the market design? We employ a calibrated DC load flow model of the German electricity system that simulates an energy-only market followed by redispatch, as well as nodal prices, for a representative week and renewable feed-in scenarios. We find that spatially concentrated curtailment by 10% of Germany's installed non-renewable generation capacity leads to a 3% welfare loss of the market value, but that loss is not driven by congestion. The electricity price rises by 14% in average, and up to 17% in peak load hours. Consumers bear the burden of curtailment, whereas producer gain in the aggregate. Effects considerably spill over to other regions. While consumers in Southern Germany always lose, consumers in Eastern and Western Germany may gain welfare. Nodal pricing reduces loss by up to 1.5%, and shifts a larger burden to consumers and to Southern Germany. The aggregated economic effects of curtailment are manageable in Germany, but its distributional effects are multiple times larger.