In this paper, we study cascading failures in power grids under the nonlinear AC and linearized DC power flow models. We numerically compare the evolution of cascades after single line failures under the two flow models in four test networks. The cascade simulations demonstrate that the assumptions underlying the DC model (e.g., ignoring power losses, reactive power flows, and voltage magnitude variations) can lead to inaccurate and overly optimistic cascade predictions. Particularly, in large networks the DC model tends to overestimate the yield (the ratio of the demand supplied at the end of the cascade to the initial demand). Hence, using the DC model for cascade prediction may result in a misrepresentation of the gravity of a cascade.
Original languageEnglish
Title of host publication35th International Symposium on Computer Performance, Modeling, Measurements and Evaluation (IFIP Performance 2017)
Number of pages6
StatePublished - 2017
EventIFIP WG 7.3 Performance 2017 - New York, United States


ConferenceIFIP WG 7.3 Performance 2017
CountryUnited States
CityNew York
Internet address

    Research areas

  • Power grids, AC versus DC, power flows, cascading failures, contingency analysis

ID: 32870001