PhD Defense by Murat Yildirim

Primary tabs

Thesis Title: Predictive Analytics and Optimization for Improved Electric Power Network Reliability and Operation


Advisors: Dr. Nagi Z. Gebraeel and Dr. Xu Andy Sun


Committee: Dr. Shabbir Ahmed, Dr. Jianjun Shi, Dr. Antonio J. Conejo (Department of Electrical and Computer Engineering, The Ohio State University)


Date and Time: Monday, July 25, 2016, 09:00 AM.

Location: Groseclose 402 (Advisory Boardroom)




Advances in sensor technology, data storage and signal processing enable methods to indirectly monitor many complex engineering systems. The aim of this dissertation is to present novel statistical and optimization methods that exploit real time sensor information to derive predictive failure risk assessments and decision models to enhance reliability and profitability in electric power systems.


In the first part, we focus on developing and solving large-scale optimization models to compute sensor-driven optimal operational and maintenance decisions for a fleet of power plants. Operational decisions relate to the well-known Unit Commitment problem, which identifies dispatch and commitment profiles that satisfy demand requirements, yet are optimized against real-time degradation levels of each power plant. Maintenance decisions focus on deriving optimal fleet level condition-based maintenance schedules that exploit potential economic and stochastic dependence existing among the individual generating units. The decisions are performed while adhering to constrains, such as generation and ramping limits of the power plants, capacities of transmission lines, network reliability, etc.


For the second part, we present a maintenance and operations scheduling policy specifically for wind farms. Maintenance considerations in this problem differ significantly from the previous model. For instance, regardless of the number of turbines scheduled for maintenance, when crews visit an offshore location, they incur significant costs due transport of workboats and helicopters. In this work, we therefore consider the trade-off between sensor-driven optimal maintenance decisions for single-turbine systems, and the significant cost reductions arising from grouping the turbine maintenances together. The effectiveness of our approach, and the impact of electricity price and crew deployment cost are illustrated in extensive experiments. Recently, there has been a growing interest in sensor-driven maintenance policies for single-turbine systems. We show that for most practical cases, these policies may perform poorer than the traditional time-based fleet maintenance policies. Our findings clearly illustrate that to obtain the full benefit of sensor information, a policy should integrate the dynamics within the maintenance and operations of the wind farm as a whole.


Finally, we consider the interaction between the operational load on the power plants, and their corresponding rate of degradation. This interaction is particularly important since it significantly affects the remaining life of the power plants and the optimal maintenance decisions. Also, by deciding on the dispatch level of the power plants, a maintenance planner can intentionally alter the optimal maintenance time (i.e. by lowering the load on the cheap power plants and postponing their preventive maintenance, or by increasing the load on the cheap power plants and using more capacity before early maintenance). We propose optimal maintenance and dispatch decisions for power plants operating in this environment, and show that considering the load dependency can provide significant savings in both maintenance and operations cost, while ensuring a more reliable system.



  • Workflow Status:Published
  • Created By:Tatianna Richardson
  • Created:07/21/2016
  • Modified By:Fletcher Moore
  • Modified:10/07/2016



Target Audience