Development of an electricity trading algorithm for a virtual power-plant

Abstract

This thesis is based on a project of RDI in Centria UAS named CoPower and is about developing a community-based virtual power plant and a working business model starting in all involved coun-tries- Finland, Ireland, Iceland and Faroe Islands. The main objective in this work was to develop a software to enhance the profitability and efficiency of Virtual Power Plants (VPPs). The software integrates data analyses and AI to make informed trading decisions based on a real-time and histori-cal data of electricity price, weather forecasts, production and consumption patterns. To achieve this, the software collects and processes data from multiple sources such as ENTSO-E API, Finnish Meteorological Institute (FMI), and sensors implemented of CoPower staff into the farm in To-holampi, Finland were the Biogas Power Plant WeKas Oy is located. Based on the received infor-mation, algorithm makes a decision to sell, store the produced electricity or buy from the grid when the prices are low or negative and manage the charging and discharging cycles of the Battery Energy Storage System (BESS), all depending on the farm consumption.

This thesis work provides a practical software solution to VPPs to enhance their electricity trading strategies using data analyses and AI integration. This software offers a framework to be adapted and optimize other VPPs in today’s dynamic electricity markets.