Renewable power plant design with uncertain data: a model of decision making based on risk analysis

The recent widespread use of renewable sources has brought about a large quantity of analyses and studies on their design and reliability, but a thorough analysis of all sources of uncertainty affecting their behaviour is still needed. Such kind of analysis is deemed to be necessary in order to perform a proper "decision making" approach for the choice of the best solution among different possible alternatives (in terms, e.g., of number and size of generating units). In the paper a proper decision making methodology for this aim is presented, in the framework of wind power plants design. It is based upon analyzing the risk in terms of "loss of load" of a wind turbine system in order to perform the best decision between alternative generation systems, designed to meet a given power demand and accounting the uncertainty of data, especially reliability and power demand data. A random variable is used to represent the risk of failure meanwhile utility theory is used to describe the consequences in a Bayesian approach. The methodology is then illustrated presenting the results obtained in the comparison between the risk functions of different alternatives of power plant generation systems. Uncertainties of load and of wind turbine generator reliability are dealt with by means of a combined Monte Carlo simulation/analytical probabilistic, which allows - with an excellent degree of precision - the deduction probability density function of the risk, becoming itself a random variable in the proposed approach, which allows a rational and complete decision making process. This is shown by means of numerical simulations in Section 4.