Probability Estimation Models to Predict Risk Occurrence in Smart Grid Projects
Open AccessSuccess in implementing smart grid projects in the electric utility industry has been mixed. Considering the strategic importance of these projects, practitioners in the industry must understand and plan for management of project risks to avoid cost overruns and delays. This praxis presents 8 models to estimate the probability of occurrence of 8 types of risks common to smart grid projects, as identified by the United States Department of Energy (DoE) in its final report on the Smart Grid Investment Grant (SGIG) program, spanning from 2009 to 2016. Non-parametric statistical analysis and binary logistic regression-based model building techniques were utilized to build the models. DoE’s publicly available documents provided input raw data on parameters of the 99 smart grid projects completed under the SGIG program. Information on occurrence of the 8 risk-types was extracted from the lessons learned sections in SGIG project description documents. Three different models were built for each risk-type, based on the ‘Purposeful Selection of Covariates’, ‘Forward Selection’, and ‘Backward Elimination’ methods of variable selection. The model with the largest area under the ROC curve after 10-fold cross-validation was selected as final output of the praxis for each Risk-Type. This praxis improves the current state of the project risk management practice by augmenting traditional, qualitative risk assessment techniques, with a quantitative technique, to estimate the probability of occurrence of 8 risk-types in future smart grid projects.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.