Methodology for Deriving Parameters for Optimization Models of Systems of Flexible Energy Resources

Abstract

The increasing share of renewable energy generation poses a challenge to maintaining the adequacy of power generation and demand. Planning the operation of flexible energy resources helps stabilize this balance. Optimization models are needed for operation planning. The use of a predefined and validated optimization model avoids modeling errors, but parameterization is still necessary and error-prone. This work presents a methodology for determining parameters of optimization models for various kinds of flexible energy resources. The parameters are derived from time series data of the resource operation after preprocessing. This methodology includes algorithms for determining operational boundaries, the input–output relationship, system states, and parameters for storage systems. Connections of flows between individual resources within a system are extracted from a standardized information model. A case study of a combined heat and power system demonstrates the applicability of the methodology by deriving a set of parameters from time series data and an information model of the system's structure. The model parameterized by means of the methodology shows very good alignment with a validation time series data set with a normalized root mean square error of 1% (generator), respectively, of 6% (heat exchanger).