GEMA: research project combines predictive and prescriptive analytics for the intelligent management of variable energy

This Friday, a FONDEF IDeA Energy Management in Microgrids with Storage project of more than 340 million pesos developed during 2020 and 2021 will be closed. The multidisciplinary investigation was led by Rodrigo A. Carrasco, researcher and academic at the Center for Energy Transition ( CENTRA) of the Faculty of Engineering and Sciences UAI.

This July 8, the closing event of the FONDEF IDeA research project called “GEMA” is held, an acronym that refers to Energy Management in Microgrids with Storage, which completed its execution on April 12. The project received a total of $341 million in financing, of which the National Research and Development Agency (ANID) financed $200 million through its program to promote scientific and technological development, and the Adolfo Ibáñez University (UAI) acted as beneficiary.

Likewise, the project had the participation of the companies Metric Arts EY and MiroSolar, the first of them dedicated to data science issues, while the second is one of the national companies with the most experience in the field of solar thermal energy and photovoltaic. Additionally, during the development of the project, the companies Impacto Renovable, Victron Energy and Agrosuper participated in different aspects of the investigation.

Rodrigo Carrasco, Associate Professor and researcher at the Center for Energy Transition (CENTRA) of the Faculty of Engineering and Sciences, and Director of the FONDEF IDeA “GEMA” Projectexplains that “Our project successfully sought to improve the developments we have been making in recent years, which combine predictive and prescriptive analytics for the intelligent management of variable and intermittent energy when there is storage”.

Solar energy has been identified as one of the main renewable sources we have to reduce our need for fossil fuels. Its main disadvantage is that it is variable (it changes throughout the day and is non-existent in the hours of darkness) and intermittent (it can disappear inadvertently), which limits its usability. On the other hand, storage systems, such as batteries, are one of the main technologies that can reduce the effects of these disadvantages, also counting on a significant reduction in their cost in the last decade.

In the opinion of Rodrigo Carrasco, “if there is no energy storage, the decisions in a solar generation system are simple: if there is excess energy, it is sold to the grid if possible, and if it is lacking, it is purchased from the grid to supply the missing demand.

With the appearance of storage, there are now a large number of new options, since we can store the surplus and not sell it, and then decide when to consume what is stored. Current battery management systems typically do not take advantage of any data science or optimization tools to make their decisions, managing energy with basic policies that do not consider historical data, future forecasts, or energy costs“.

The GEMA project was developed in the midst of a pandemic (2020-2021). Rodrigo Carrasco specified that the objective of this project was to improve predictive models and connect them with more sophisticated prescriptive models in order to further improve our results. “At the same time, we hoped that this would allow us to adapt the models for other types of networks and storage systems, being able to deploy these systems in a wide variety of geographical locations and network configurations”, the researcher also pointed out.

The research also contributed to developing advanced human capital, training engineers to be able to use this type of tools to apply to different types of problems.

The work carried out also made it possible to demonstrate the economic benefits and flexible solutions that can be replicated in other contexts such as power charging systems, peak charges, among others. In addition, today it has a patent that allows it to protect the intellectual property developed and that will soon be licensed to interested organizations.

In this sense, the researcher pointed out: “We believe that what has been developed will be a contribution to improve the operation of photovoltaic systems and, with this, we will optimize the incentives for their penetration in different types of clients. The success of the project would not have been possible without a multidisciplinary vision that was achieved thanks to a variety of experiences in the team”.

By way of challenges, in Carrasco’s opinion; Although we have an implementation in a real solar plant, there are multiple aspects that we can develop in the future. In particular, optimization opens the door to analyze different functions that not only seek to reduce costs, but also improve resilience, increase battery life or face complex scenarios from the climatic or environmental point of view.

Participating in the GEMA project were: Tito Homem-de-Mello, academic from the UAI Business School; Gonzalo Ruz, Carlos Silva and Jocelyn Olivari, academics from the Faculty of Engineering and Sciences UAI; and Francisco Jara, professor at the Diego Portales University. Benjamín Bastidas and José Luis Ortiz also participated as project engineers and Helena García, student of the Master in Engineering Sciences, Industrial mention.