Stochastic model for optimal declaration of day ahead station availability in power pools in India

This paper evolves a strategy for optimal declaration of the day ahead generation availability in the ´Availability Based Tariff regime in India´. The mathematical formulation presented in this paper maximizes the expected revenue of the generator by considering various stochastic parameters. The computational model developed in this work considers three random variables viz, the generating unit´s availability, unscheduled interchange (UI) and load. The state transition of generating units is assumed to follow the first order Markov´s process. The load and unscheduled interchange related to grid frequency follow known discrete probability distributions. Monte Carlo state duration sampling is applied for simulating the unit transitions. Both UI as well as load uncertainty are generated by state sampling approach from their probability distributions. The expected revenue generated in a day, comprising of 96 time blocks of 15 minutes duration, is statistically computed. Finally by iterative approach, the expected optimal station availability for next day´s declaration is computed. A numerical example of a generating station is illustrated to show the revenue maximization strategy by this optimization