A Stochastic Framework for Short-Term Operation of a Distribution Company

Abstract

This paper presents a stochastic framework for short-term operation of a distribution company (disco). The proposed framework optimizes disco's operational decisions in two hierarchical stages. The first stage, called day-ahead operation stage (DAOS), deals with the operational decisions on purchases from the day-ahead market and commitment of distributed generation (DG) units. The objective of this stage is to minimize the expected operating cost while the financial risk exposed by uncertain real-time prices and loads is restricted to a given level. The model associated with this stage is based on the mixed-integer programming (MIP) format. The second stage, named real-time operation stage (RTOS), deals with disco's activities in real-time. In RTOS, decisions are made on real-time market transactions, dispatch of online DGs, and invocation of load curtailments (LCs) such that the expected operating cost is minimized. This stage is formulated as a nonlinear programming (NLP) problem. To investigate the effectiveness of the developed framework, it is applied to a typical Finnish 20-kV urban distribution network.