Doctoral Dissertations

Orcid ID

0000-0002-0890-128X

Date of Award

8-2021

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Industrial Engineering

Major Professor

Xueping Li

Committee Members

Kai Sun, James Ostrowski, John E. Kobza

Abstract

Natural disasters can cause widespread disturbances/power outages within distribution networks and hinder a utility’s ability to provide uninterrupted power supply to the critical public buildings (e.g., hospitals, grocery stores, fire, police and gas stations) within the utility’s serviced region. Backup generators, which are typically relied on during power interruptions, have limited capacities and have been reported to experience failures during usage. Microgrids, defined as localized power grids that incorporate distributed generators (DGs) and energy storage systems (ESSs) to allow them to operate independent of the main grid (i.e., island mode), can help utilities provide disaster relief power supply to critical public buildings during such outages. This research investigates the optimization of utility-owned microgrids assumed to be operating in island mode and supplying power to a network of critical public buildings over the course of a week-long power outage. A deterministic and two-stage stochastic model (considering only DGs), as well as a multi-stage stochastic model (considering DGs and ESSs) are developed to optimize the investment economics, reliability and resilience of the microgrids. The models provides a holistic objective function that captures the investment, fixed operation and maintenance, power supply efficiency, reliability and resilience of the microgrid in terms of a minimized total cost to the utility. This is accomplished by optimizing the location, sizing, power supply assignment and total number of DGs and ESSs within a utility-owned microgrid. Hourly and weather (cloud coverage) uncertainty in daily DG power output and critical public building demand are considered. The final DG-plus-ESS multi-stage model provides an exhaustive solution framework, that analyzes the microgrid’s reliability across all possible weather (cloud coverage) scenarios (e.g., sunny, cloudy, overcast) of a week-long outage (3,279 total scenarios).

Comments

I updated the dissertation to state August instead of May as the graduation month (per your most recent edit request).

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS