{"id":{"repo_id":"gatech","oai_identifier":"oai:repository.gatech.edu:1853/72985"},"canonical_url":"https://search.dev.ndltd.org/etd/gatech/oai:repository.gatech.edu:1853/72985","repository":{"repo_id":"gatech","name":"Georgia Tech","base_url":"https://repository.gatech.edu/server/oai/request"},"display":{"title":"Large scale data analytics for resilience of energy networks","abstract":"Massive power failures are induced frequently by natural disasters. Two fundamental challenges arise in face of such failures: First, how recovery can be resilient to the increasing severity of disruptions and their impact on service users in a changing climate. Second, how can we measure the impact of failures and recovery and its heterogeneity on customers with different characteristics. We conduct a large-scale study on recovery from 169 failure events at two operational distribution grids in the states of New York and Massachusetts. Guided by unsupervised learning from non-stationary data, our analysis finds that under the widely adopted prioritization policy favoring large failures, recovery exhibits a scaling property where a majority (90%) of customers recovers in a small fraction (10%) of total downtime. However, recovery degrades with the severity of disruptions: large failures that cannot recover rapidly increase by 30% from the moderate to extreme events. Prolonged small failures dominate entire recovery processes. Further, our analysis demonstrates the promise of mitigating the degradation by enhancing recovery of a small fraction of large failures through distributed generation and storage. Next, a dynamic resilience metric is developed using spatiotemporal failure and recovery processes incorporating the cost imposed on individual customers. The resilience metric is then combined with randomization inference to design a framework on how to study the dynamic cost and its heterogeneity on customers with different characteristics. Our framework is validated on publicly available data from multiple states in the US.","abstract_html":"Massive power failures are induced frequently by natural disasters. Two fundamental challenges arise in face of such failures: First, how recovery can be resilient to the increasing severity of disruptions and their impact on service users in a changing climate. Second, how can we measure the impact of failures and recovery and its heterogeneity on customers with different characteristics. We conduct a large-scale study on recovery from 169 failure events at two operational distribution grids in the states of New York and Massachusetts. Guided by unsupervised learning from non-stationary data, our analysis finds that under the widely adopted prioritization policy favoring large failures, recovery exhibits a scaling property where a majority (90%) of customers recovers in a small fraction (10%) of total downtime. However, recovery degrades with the severity of disruptions: large failures that cannot recover rapidly increase by 30% from the moderate to extreme events. Prolonged small failures dominate entire recovery processes. Further, our analysis demonstrates the promise of mitigating the degradation by enhancing recovery of a small fraction of large failures through distributed generation and storage. Next, a dynamic resilience metric is developed using spatiotemporal failure and recovery processes incorporating the cost imposed on individual customers. The resilience metric is then combined with randomization inference to design a framework on how to study the dynamic cost and its heterogeneity on customers with different characteristics. Our framework is validated on publicly available data from multiple states in the US.","abstract_has_math":false,"creators":["Afsharinejad, Amir Hossein"],"institution":"Georgia Institute of Technology","degree_name":null,"degree_level":"Doctoral","degree_discipline":null,"degree_department":"Electrical and Computer Engineering","school":null,"contributors":[],"advisors":["Ji, Chuanyi"],"committee_chairs":[],"committee_members":["Ganz, Scott","Davenport, Mark","Divan, Deepakraj","Thomas, Valerie"],"year":2021,"date_issued":"2021-12-13","date_published":"2021-12-13","updated_at":"2026-07-27T19:51:32Z","subjects":["Resilience of recovery services","Large-scale power failures","Data analytics at scale","Infrastructure enhancement"],"languages":["en_US"],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/1853/72985","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Ji, Chuanyi"]},{"key":"dc:contributor.committeemember","label":"Committee Member","values":["Ganz, Scott","Davenport, Mark","Divan, Deepakraj","Thomas, Valerie"]},{"key":"dc:contributor.department","label":"Department","values":["Electrical and Computer Engineering"]},{"key":"dc:creator","label":"Author","values":["Afsharinejad, Amir Hossein"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2024-01-10T18:37:09Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2024-01-10T18:37:09Z"]},{"key":"dc:date.issued","label":"Date","values":["2021-12-13"]},{"key":"dc:publisher","label":"Institution","values":["Georgia Institute of Technology"]},{"key":"dc:type","label":"Dc Type","values":["Text"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Doctoral"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Resilience of recovery services","Large-scale power failures","Data analytics at scale","Infrastructure enhancement"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language.iso","label":"Language (ISO)","values":["en_US"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/1853/72985"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Massive power failures are induced frequently by natural disasters. Two fundamental challenges arise in face of such failures: First, how recovery can be resilient to the increasing severity of disruptions and their impact on service users in a changing climate. Second, how can we measure the impact of failures and recovery and its heterogeneity on customers with different characteristics. We conduct a large-scale study on recovery from 169 failure events at two operational distribution grids in the states of New York and Massachusetts. Guided by unsupervised learning from non-stationary data, our analysis finds that under the widely adopted prioritization policy favoring large failures, recovery exhibits a scaling property where a majority (90%) of customers recovers in a small fraction (10%) of total downtime. However, recovery degrades with the severity of disruptions: large failures that cannot recover rapidly increase by 30% from the moderate to extreme events. Prolonged small failures dominate entire recovery processes. Further, our analysis demonstrates the promise of mitigating the degradation by enhancing recovery of a small fraction of large failures through distributed generation and storage. Next, a dynamic resilience metric is developed using spatiotemporal failure and recovery processes incorporating the cost imposed on individual customers. The resilience metric is then combined with randomization inference to design a framework on how to study the dynamic cost and its heterogeneity on customers with different characteristics. Our framework is validated on publicly available data from multiple states in the US."]},{"key":"dc:description.degree","label":"Dc Description Degree","values":["Ph.D."]},{"key":"dc:format.mimetype","label":"Dc Format Mimetype","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Large scale data analytics for resilience of energy networks"]}]}],"canonical_facts":{"dc:contributor.advisor":["Ji, Chuanyi"],"dc:contributor.committeemember":["Ganz, Scott","Davenport, Mark","Divan, Deepakraj","Thomas, Valerie"],"dc:contributor.department":["Electrical and Computer Engineering"],"dc:creator":["Afsharinejad, Amir Hossein"],"dc:date.accessioned":["2024-01-10T18:37:09Z"],"dc:date.available":["2024-01-10T18:37:09Z"],"dc:date.issued":["2021-12-13"],"dc:description.abstract":["Massive power failures are induced frequently by natural disasters. Two fundamental challenges arise in face of such failures: First, how recovery can be resilient to the increasing severity of disruptions and their impact on service users in a changing climate. Second, how can we measure the impact of failures and recovery and its heterogeneity on customers with different characteristics. We conduct a large-scale study on recovery from 169 failure events at two operational distribution grids in the states of New York and Massachusetts. Guided by unsupervised learning from non-stationary data, our analysis finds that under the widely adopted prioritization policy favoring large failures, recovery exhibits a scaling property where a majority (90%) of customers recovers in a small fraction (10%) of total downtime. However, recovery degrades with the severity of disruptions: large failures that cannot recover rapidly increase by 30% from the moderate to extreme events. Prolonged small failures dominate entire recovery processes. Further, our analysis demonstrates the promise of mitigating the degradation by enhancing recovery of a small fraction of large failures through distributed generation and storage. Next, a dynamic resilience metric is developed using spatiotemporal failure and recovery processes incorporating the cost imposed on individual customers. The resilience metric is then combined with randomization inference to design a framework on how to study the dynamic cost and its heterogeneity on customers with different characteristics. Our framework is validated on publicly available data from multiple states in the US."],"dc:description.degree":["Ph.D."],"dc:format.mimetype":["application/pdf"],"dc:identifier.uri":["https://hdl.handle.net/1853/72985"],"dc:language.iso":["en_US"],"dc:publisher":["Georgia Institute of Technology"],"dc:subject":["Resilience of recovery services","Large-scale power failures","Data analytics at scale","Infrastructure enhancement"],"dc:title":["Large scale data analytics for resilience of energy networks"],"dc:type":["Text"],"thesis:degree_level":["Doctoral"]},"updated_at":"2026-07-27T19:51:32Z"}