Predicting rejection rates of electric distribution wood pole assets
Author(s)
Kelchev, Boyan Lyubomirov
DownloadFull printable version (7.579Mb)
Other Contributors
Leaders for Global Operations Program.
Advisor
Georgia Perakis and Patrick Jaillet.
Terms of use
Metadata
Show full item recordAbstract
Pacific Gas & Electric Companys (PG&E) electric distribution system includes approximately 2.4 million wood utility poles. The Pole Test & Treat (PTT) program at PG&E is responsible for inspecting these poles, prolonging their service life through the use of chemical treatments or structural reinforcements, and identifying poles that need to be replaced. Following industry best practices and taking advantage of the vast knowledge and experience of the PTT team, PG&E inspects poles every 10 years. The company believes that the next step in improving the performance of the PTT program is to leverage the data collected since the inception of the program and utilize modern statistical methods to better understand and predict decay in their wood pole assets. In this thesis, we describe the possibilities and limitations of using PG&E's current data to predict the results of future inspections. We study both the possibility of making predictions at the individual pole level, predicting whether a pole will be rejected during the next inspection cycle, and at the aggregate level, predicting what the overall rejection rate in a subpopulation of poles will be in the future. In order to accomplish this, we first studied the available data sources and performed exploratory analysis to understand the characteristics of the different variables and form hypotheses about the main drivers of rejections during pole inspections. Next, we attempted to build a classification model to predict the results of future inspections. This showed us that our current data cannot be used to yield an accurate prediction at the individual pole level. Then, we developed a model to estimate the overall rejection rates of subpopulations of poles. The result was a prediction with a Mean Absolute Percentage Error of about 30%. While not ideal, this model gives PG&E the ability to budget and plan for future work better. Finally, we leveraged the results of the prediction model to simulate the evolution of rejection rates in the future. The simulation highlighted a well-known problem in the utility industry - the problem of aging infrastructure. The relatively low average age of poles and the low replacement rates observed in the past few inspection cycles mean that PG&E will likely experience a drastic increase in rejection rates as the average age of its pole population grows. Planning for the accompanying increase in manpower and work hours required will be of great importance to PG&E in the next few decades.
Description
Thesis: M.B.A., Massachusetts Institute of Technology, Sloan School of Management, in conjunction with the Leaders for Global Operations Program at MIT, 2017. Thesis: S.M., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, in conjunction with the Leaders for Global Operations Program at MIT, 2017. Cataloged from PDF version of thesis. Includes bibliographical references (page 60).
Date issued
2017Department
Leaders for Global Operations Program at MIT; Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science; Sloan School of ManagementPublisher
Massachusetts Institute of Technology
Keywords
Sloan School of Management., Electrical Engineering and Computer Science., Leaders for Global Operations Program.