Sushant Varghese Email & Phone Number

New York, United States

Working on resource adequacy modeling as part of the Market Structures Department.

University Park, Pennsylvania, United States

Development of a redesigned stochastic market design for electricity markets-Collaborating with PJM to develop a novel stochastic market design to replace deterministic models used in current practice. -Performing simulation of electricity markets as a stochastic MILP, using stochastic dual dynamic IP for problem decomposition and unsupervised learning algorithms for scenario reduction to demonstrate performance of proposed design.-Developed MILP model to simulate current market of PJM, the largest electricity grid operator in the US, with an error of less than 1% to be used as test bed for analysis. Development of modified mixed integer unit commitment model with enhanced combined cycle gas turbine modeling-Contributed to formulation of configuration-based combined cycle gas turbine modeling that explicitly represents the transition between modes as a distinct configuration to enable dynamic ramping.-Demonstrated that standard models do not capture all the flexibility of generators, and therefore result in higher system operating costs and lower revenue to generators. Demonstration of the value of flexibility enhancements to natural gas power plants-Studied the impacts of flexibility retrofits on the value and operational dynamics of existing natural gas power plants.-Collaborated with GE to develop multiple mixed integer linear programming models to simulate the electricity markets of Texas, California, and northeast US as test beds for analysis. -Assessed the effects of distinct flexibility upgrades to different classes of natural gas power plants on their operational dynamics, asset owner profitability, and system savings.-Demonstrated that current usage pattern and type of upgrade has impact on the value of the upgrade, analyzed the sensitivity of results to higher renewable scenarios and demonstrated diminishing returns due to market inefficiencies.

United States

-Developed a python package that can pick representative and outlier scenarios from a GE MAPS output. The toolkit consists of over 25 statistical tools and unsupervised machine learning algorithms which users can pick from for their analysis. -Developed a novel modified K Nearest Neighbors algorithm which is capable of picking anomalies in the data while ensuring that the selected outliers are qualitatively different from one another.

Columbus, Ohio Area

-Determined the optimal design of a hybrid energy system consisting of solar and battery energy storage under various pricing and subsidy regimes in California and demonstrated the distortionary effects of such incentive mechanisms.-Applied advanced discrete optimization techniques to approximate the nonlinear programming model with a mixed integer linear programming model.

Bachelor Of Technology, Production Engineering With Minor In Economics, Gpa 8.36/10

Activities and Societies: Vice President of Marketing and Events at AIESEC, Event Manager at Festember, Event Manager at Pragyan

Doctor Of Philosophy, Dual Degree In Industrial Engineering And Operation Research, Gpa 4/4

Activities and Societies: Research Assistant

High School Diploma