Artificial Intelligence Boosts Renewable Energy Sources | Venturus

Artificial Intelligence Boosts Renewable Energy Sources

How machine learning can be used to add value to the energy generated by intermittent renewable energy sources such as wind and solar 

Renewable energy sources, such as wind and solar, have been highlighted as alternatives to meet the growing demand for clean and sustainable energy. However, despite being a great opportunity, the intermittent nature of these energy sources also presents challenges for the sector. Artificial Intelligence (AI) techniques have been applied to mitigate these challenges and add even more value to these energy sources. 

The evolution of wind and solar energy sources 

The growing demand for energy, along with the risk of global warming, has motivated several countries around the world to seek clean and renewable energy sources as alternatives to make up their energy matrix. According to the International Energy Agency (IEA), the share of renewable energy sources in the global energy matrix is expected to exceed 40% by 2040, against the current 25% 

Wind and solar photovoltaic energy sources have gained prominence among renewable energy sources, presenting a significant growth in recent years. In 2018, the installed capacity of global wind energy reached the mark of 600 GW, presenting a growth of 9.8% compared to 2017. The installed capacity of photovoltaic solar energy reached the mark of 512 GW, 27% higher than the previous year. Together, the two renewable energy sources accounted for 7% of global electricity production. 

In Brazil, energy is largely derived from hydroelectric plants that, despite also using a renewable energy source, generate a dangerous dependence on droughts that have affected several regions of the country in recent years. Thus, the use of wind and solar energy is also gaining space here, currently representing around 9% and 1.2%, respectively, of installed electricity capacity in the country. 

Government incentives have boosted the installation of wind and solar parks in several regions of the country, and according to the National Electric Energy Agency (ANEEL), the installed electrical capacity from these sources already surpasses the 16 GW mark. In order to have a size of this number, the Itaipu hydroelectric plant, one of the largest in the world, has a capacity of 14 GW. With the decrease in production costs of wind and solar energy sources, the expectation is that in the coming years the participation of these energy sources in the composition of the national electric matrix will be even more significant. 

The intermittent nature of wind and solar energy sources 

Wind and solar power sources will play an important role in meeting the growing demand for energy in the coming years in a clean and sustainable way. However, the growth of energy generation through these sources must also bring challenges for the sector. Unlike other energy sources, wind and solar energy have a strong intermittent aspect, which can bring complications to the operation of the electrical system. 

The power generation capacity of a wind or solar power plant depends intrinsically on local environmental conditions. Environmental conditions for the production of wind and solar power can vary seasonally or even throughout the day, which makes it difficult to predict the capacity to generate energy at a given time. This unpredictability brings technical and commercial difficulties for the operation of the electric power system. 

The generation and consumption of energy of the electric system must be balanced, that is, there must be a control between the generation and demand of electric energy. Both the lack of energy generated in relation to demand and its excess can cause problems to the system, which can damage equipment, impair the quality of energy or even lead to a drop in supply. 

The demand for energy varies throughout the day, so the system operator must increase or reduce the amount of energy generated to meet this demand. The use of intermittent energy sources, such as wind and solar power, can make the problem more difficult to manage because the capacity of these sources can not be guaranteed at any given time. The greater the complexity, the greater the participation of these intermittent energy sources in the electric matrix. 

In addition, the low reliability of the amount of energy available at any given time requires operators to have an alternative power source that can be triggered quickly to meet the demand of the system. This backup is usually done through other, more reliable sources of energy, such as fossil fuels, which can increase system operating costs and harm the environment. 

Artificial Intelligence can make energy production more predictable 

Some companies have invested in technological innovations to mitigate the problem of the unpredictability of intermittent renewable energy sources. Recently Google through DeepMind, a wholly-owned subsidiary of the artificial intelligence company, announced that it is applying machine learning techniques to predict the company’s wind farm power generation capacity in the central United States. 

Historical data from wind turbines and weather forecasts were used by DeepMind to train a neural network capable of predicting the power generated during the day by turbines of the plant in advance of 36 hours. Based on these predictions, models were used to create optimal recommendations for trading the generated energy. 

This predictability in power generation adds value to power system operators who can properly plan how to use it to meet the demand of the system. According to Google, the preliminary results obtained indicate an appreciation of approximately 20% of the wind energy generated in relation to the base scenario other than the generation forecast. 

Colorado’s National Center for Atmospheric Research (NCAR) has also been working with the use of artificial intelligence to predict wind power generated by wind farms in the state. Predictions with unprecedented accuracy have been generated by applying machine learning algorithms in the analysis of wind turbine data combined with information from weather stations and satellites. 

The reliability in the use of the energy of the wind farms allowed that the operators could increase the portion of renewable energy in its operation. Xcel Energy, one of the largest energy companies in the Colorado region, has expanded the amount of energy used from renewable sources to more than 30%.  In addition, the reliability that energy will be available allows operators to reduce the backup infrastructure of other power sources, which reduces the cost of power generated. 

NCAR and Xcel are now working on the forecast for solar power generation. The project is even more challenging than forecasting for wind power because it aims to predict the power capacity generated even by residential installations that are connected to the grid of the operator. The solution will use data from satellites, sky images, pollution monitors and public solar panels to infer the amount of solar energy generated. 

Especially for Brazil, which has enormous potential for wind and solar power generation, initiatives like this can add enormous value to the electric sector, further boosting the use of these energy sources. 


Veja mais

Entre em contato

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Campinas / SP - Brasil

Estrada Giuseppina Vianelli di Napolli, nº 1.185
Condomínio GlobalTech Campinas
Polo II de Alta Tecnologia
CEP 13086-530 – Campinas – SP
+55 (19) 3755-8600

+55 (19) 3755-8600