AI Energia Group

Energia Trading Student Project                                                                                                                                    A rapid increase in data & digitalisation has changed the global energy trading industry with AI & machine learning tools being used to trade in increasingly complex markets with an ever-increasing volume of data. Energia Trading identified a requirement for increased insights and decision support from market data utilising AI and thus developed and leveraged a relationship with Queens University’s MSc in Data Analytics course. Our goal is to deploy AI to optimise our portfolio utilising the fundamental market data and as such developed an algorithm that creates strategies and provides pathways to show the most economical ways to trade electricity in the Irish Single Electricity Market (ISEM) which ultimately enables us to better serve our customer from a cost perspective whilst also embracing digitalisation.

Energia is of the view that projects should be co-ordinated with a view to adding value to the business. In this instance there were a number of commercial and non-commercial links in the value chain including delivery of technical analytics, improved system efficiency and strategic decision support with the aim of creating the potential for commercial opportunities while giving the students a real challenge with SMART objectives both commercial and in terms of personal development. The students have benefited from the support and belief of the company through investment in this project in terms of personal development opportunities of the students, the development of the AI solution and in making a tangible financial commitment through letting the AI model run a live market position with real commercial exposure (Profit and loss).

Energia set out with the goal of deploying AI to improve market knowledge, remove psychological bias and to provide state of the art decision support tools for its existing and future trading activities. The structure of the trading desk is lead by an experienced trader or domain expert whereby a fundamental understanding of the data is communicated and interpreted in technical terms by the data analysts/students, applied to the data and trading signals are returned via interactive visualisation.

Technological / Technical Achievement – At Energia, we leverage advanced analytical approaches in volatile short term markets that contribute substantially to profitability , which is far from the reach of human computational power of making rational decisions, especially when dealing with large amounts of data. This not only transformed the trading processes, but progressively shifts the traders from merely just executing the trades based on their fundamental decisions towards focusing more on market forecasting & bringing improvements in advanced analytics models continuously.

Societal
As a result of our passion for consumers and businesses alike, Energia Trading, can look to leverage the in-house systematic algorithm that develops trading strategies, to one day potentially provide more affordable, reliable, sustainable power to our customers via sharper consumer pricing due to better market optimisation.

Models deployed
Post Brexit/ Inter-connector flow prediction model – informing a better understanding of the direction that the inter-connector between Ireland and GB will flow aiding increased understanding for the traders and its impact on the wider portfolio

ML model 1 – trading the ex-ante auctions against the Balancing market with considerable success to inform the wider portfolio (decision support) and as an outright trading signal for discretionary real time trading in ISEM Competitor analysis model- allowing Energia to view and predict the behavior of other trading participants I the ISEM market to better optimise our own portfolio.