Experience with nuclear power plant accidents demonstrates that long-term consequences of widespread contamination and the associated societal disruption are considerable, which can outweigh the immediate danger to public health and safety. This dissertation develops a new approach to modeling nuclear disaster consequences, known as the "Fukushima-informed Recovery and Cost Assessment" (FIRCA) approach. The FIRCA approach enhances the realism of modeling nuclear disaster consequences in three ways: 1) consideration of a more complete set of nuclear disaster impacts, 2) a new cost assessment of impacts, and 3) new assumptions about modeling of post-accident protective actions (based on Fukushima recovery experience and EPA guidance). This dissertation compares the new FIRCA approach to a conventional approach to modeling nuclear disasters. Results find that modeling based on the Fukushima nuclear disaster leads to a markedly higher overall cost (by a factor of 12 for the case study). This difference is due to a larger anticipated relocation size, a more complete set of nuclear disaster impacts (including a measure of societal disruption from displacement), and new cost assessments (e.g., cleanup costs) based on the Fukushima nuclear disaster. This dissertation also uses the FIRCA approach to analyze certain recovery policies. Results of the case study indicate that the harm of post-accident impacts could be significantly reduced (by a factor of 7) if authorities were to implement a relocation delay, credit shielding in the dose calculations that determine relocation areas, and the use of relocation dose criteria higher than federal guidance currently recommends. Results and insights could usefully inform post-accident disaster preparedness, modeling of nuclear disaster impacts, regulatory safety measures, and anticipated recovery policies after a nuclear accident.