November 7, 2024

Artificial intelligence is making a relevant contribution to the energy transition in Chile, especially in the renewable sector. We detail examples that demonstrate this.

Artificial intelligence (AI) amazes and astonishes, but it also generates fear and disbelief. In simple terms, Claudio Gutiérrez, an academic at the University of Chile and researcher at the Millennium Institute for Data Fundamentals, defines it as “the technology that allows computers to simulate human intelligence and the capabilities of people to solve problems.”

He specifies that on its own or combined with other technologies (such as sensors, geolocation or robotics), “AI can perform tasks that would otherwise require intelligence or human intervention. Digital assistants, GPS guidance, autonomous vehicles and generative artificial intelligence tools (such as Open AI’s GPT Chat) are just some examples of artificial intelligence in the daily news and in our daily lives.”

FACT

1 Suncast offers a soiling prediction service (accumulation of contaminants or dirt on solar collector surfaces). Using advanced analytics and satellite weather forecasts, its models calculate this factor and optimize the cleaning schedule for panels to maximize the profitability of photovoltaic parks.

GHG REDUCTION

In the energy transition that Chile is experiencing, AI is already making its contribution felt. Bernardo Severino, researcher at the Energy Transition Center (Centra) of the Faculty of Engineering and Sciences of the Adolfo Ibáñez University, points out that this technological tool has great potential to reduce greenhouse gas (GHG) emissions, being a key driver in the fight against climate change. “A 2021 Boston Consulting Group (BCG) study estimates that currently proven AI use cases could reduce global emissions by 5% to 10% by 2030. If that potential is fully realized, AI-powered applications could contribute 10% to 20% of the IPCC’s interim emissions reduction target for 2030, which is critical to achieving carbon neutrality by 2050,” he says.

He adds that, in the context of Chile, where the energy transition is driven by the growth of renewable energies such as solar and wind, artificial intelligence “plays a crucial role in enabling more efficient management of these sources. Its ability to predict demand and supply patterns, based on real-time data, optimizes the integration of these intermittent energies, improves the stability of the electrical system and maximizes the use of available resources.”

Along the same lines, Andrés Díaz, director of Mobility South America for Schneider Electric (SE), comments that AI and digital tools are useful for optimizing supply and demand, as well as for accelerating the shift towards energy systems with lower carbon emissions. He then highlights that, as a neutral and independent consulting company, they collaborate with clients from more than 100 countries, putting at their disposal more than 20 years of experience in machine learning and artificial intelligence, in which they have developed “15 solutions with AI capabilities, more than 20 internal applications, three global AI team centers around the world and more than 350 experts in this field and data.”

The executive also comments that the study “The role of artificial intelligence in achieving the Sustainable Development Goals”, by Nature (2020), revealed that AI could facilitate the fulfillment of 79% of the Sustainable Development Goals (SDG) and up to 93% of these if only the environmental SDGs are considered. Faced with this scenario, he highlights that AI, like machine learning, “has the ability to learn on its own. This will undoubtedly pose new technological challenges, driving the development of new processes, procedures and technologies aimed at improving energy efficiency. AI is also key to improving the availability of electric power, and our goal is to increase it through more efficient and sustainable networks.”

CAPABILITIES AND BENEFITS

According to Andrés Díaz, “it is important to keep in mind that in order to correctly execute AI, it is necessary to understand its true capabilities, the impact it can have on the business and the advantages it brings, together with digitalization, among which is reducing carbon emissions, optimizing energy demand and improving operational efficiency.”

Bernardo Severino, meanwhile, explains in detail some of the main benefits offered by artificial intelligence in different areas of the electrical industry:

• Energy efficiency: It allows the analysis of large volumes of data from multiple sources (such as energy consumption, real-time demand and environmental conditions). Thanks to this, it is possible to identify consumption patterns and adjust the energy supply dynamically, optimizing the use of resources. A clear example is energy management in smart buildings, where AI systems automatically adjust HVAC and lighting based on occupancy, external conditions and user habits, which can significantly reduce energy consumption.

• Renewable energy generation: AI improves the efficiency of renewable energy generation plants, such as solar and wind farms. These energy sources are intermittent and depend on factors such as solar irradiance or wind speed. AI algorithms can predict future energy production based on weather data and optimize plant operation, adjusting the need to operate storage systems. This not only improves the reliability of renewable energy, but also reduces the need to resort to conventional energy sources when production is low.

• Energy storage: This solution can optimize energy storage systems (such as batteries) by predicting when and how much energy will be needed, minimizing losses and improving battery life and performance. This is crucial to ensure that renewable energy can be stored when there is excess production and released when demand requires it.

• Development of new technologies: AI is accelerating the creation of disruptive technologies in the energy sector. One example is innovation in smart grids, which use AI to manage the flow of electricity more efficiently and ensure that energy reaches where it is needed at the right time. AI is also accelerating advances in materials research and the creation of new carbon capture technologies and alternative energy sources. Thanks to the ability to process large volumes of data and discover hidden patterns, AI is helping researchers quickly identify new materials for higher-capacity batteries, more efficient solar panels or advanced energy storage technologies. These are essential advances to accelerate the transition to a decarbonized and sustainable energy system in the long term.

RENEWABLE ENERGY

In a 2024 report, the electricity generation company Atlas Renewable Energy (ARE), with renewable energy projects in Chile, highlights the great contribution that AI is making to the renewable energy industry, as it provides it with “relevant information for the development, construction, operation and maintenance of projects.” It specifies that, thanks to machine learning algorithms and the increasing computing capacity of computer systems, AI is enhancing five key areas of wind and solar photovoltaic plants: design of predictive models for energy generation, monitoring and diagnosis of operation, performance and efficiency, battery storage and distribution of energy, and cost reduction.
The report adds: “AI improves the performance of renewable energy assets, as it streamlines and makes processes more efficient. The result is more competitive prices and a reliable supply of clean energy for generators and customers.”

The ARE document then details the benefits of AI for solar or wind plants in the five areas mentioned above:

• Predictive models: This technology helps develop models to predict the amount of energy that a wind or solar photovoltaic plant could generate. To this end, companies like ARE use data systems that adjust to changes at a much faster rate than conventional methods.

• Monitoring and diagnosis: Its application in this area reduces the need to carry out physical inspections at complex sites, such as the nacelle of a wind turbine installed at more than one hundred meters high. Using sensors and data collection devices installed in renewable energy plants, AI algorithms can continuously analyze the performance of the equipment. This remote monitoring capacity saves time, reduces costs and improves safety, as it minimizes the exposure of technicians to dangerous environments. Artificial intelligence also has applications for fault diagnosis, which allows for making the necessary corrections and increasing production.

• Increased performance and efficiency: AI algorithms can adjust equipment settings (e.g. solar panel orientation) in real time and make them operate at high capacity to increase their energy production. They also prevent energy losses as renewable electricity discharges and optimized maintenance programs minimize downtime and maximize equipment availability.

• Storage and lower costs: By using predictive analysis, it is possible to predict the demand for the resource and distribute the stored energy at more convenient times, thus optimizing the operations of renewable energy storage systems. AI also allows monitoring their performance and efficiency, detecting any potential problems and reducing the costs associated with batteries.

EXAMPLES

Various examples show the great impact that artificial intelligence is having in sectors related to the national energy transition.

Schneider Electric, for example, offers AI-based solutions “that take advantage of our architecture and platforms such as EcoStruxure for the benefit of our four end markets: buildings, data centers, infrastructure and industry,” says Andrés Diaz.

EcoStruxure Resource Advisor is a solution to manage the energy and sustainability footprint. “This platform allows companies to collect, analyze and automate important information for their sustainability goals. It centralizes that data in one place so that AI and human experience can get the most out of these figures. In this way, we demonstrate the use of digital innovation for resource and energy management, in order to make informed business decisions and improve business results, he adds.

For Claudio Gutiérrez, one of the most significant contributions is made by Suncast, a Chilean platform that uses AI to predict solar and wind energy production with high precision, allowing the operation of renewable plants to be adjusted based on real-time weather conditions. “This has allowed operators to maximize energy performance, reduce losses and improve the reliability of the electrical grid, which is key to the integration of renewable energies in Chile. Suncast currently manages more than 3,000 MW in renewable plants (solar and wind) throughout the country,” he says. Another relevant initiative is the intelligent management of buildings through the Enel X platform, which uses AI to optimize energy consumption in large infrastructures. The researcher from the University of Chile details that “this solution automatically adjusts heating, ventilation and air conditioning systems based on current conditions, reducing energy consumption and carbon emissions, and achieving significant economic savings.”

For their part, the companies Ecometric and Innergex Energía Renovable announced the creation of the first full-scale artificial intelligence laboratory for wind farms worldwide. The facility will be located in the Cuel wind farm, located in the commune of Los Ángeles, Biobío Region.

The project will implement AI-based noise monitoring devices to optimize the operation and maintenance of wind turbines. This technology will allow noise events to be detected in real time, facilitating decision-making to mitigate impacts on the community.

Energy and water consumption

Bernardo Severino warns that, despite the obvious contribution of artificial intelligence to the national energy transition, “there are also significant risks, mainly related to the energy and water consumption required by data centers and the infrastructures necessary to train and execute AI algorithms.”

It is estimated that the energy used to train large AI models is growing at a rate of 26% to 36% annually, “which represents a significant challenge in terms of sustainability,” he says.
The researcher from Adolfo Ibáñez University maintains that solutions to these problems include the optimization of AI models to make them more energy efficient and the adoption of more sustainable data centers that use renewable energy sources and more efficient cooling technologies.

He also highlights the need for mass monitoring systems that provide real-time operational data in the electrical system and in other key sectors for the energy transition, such as transportation, industry and agriculture.

And in the case of electricity networks, it states: “It is imperative to relaunch the installation of smart meters to improve the visibility and management of distribution networks and, at the same time, involve end users in the energy transition.”

Courtesy of Induambiente.