In today’s multifaceted energy world, a growing number of prosumer assets are increasing the complexity of power grids (towardsdatascience.com/no-fast-enough-energy-transition-without-intelligent-energy-storage). Decentralized systems with solar generation, wind turbines, and electric vehicles provide promise for a decarbonized future, but also bring along challenges for both utilities and prosumers.
The transformation of energy grids, the emergence of new services, new players: prosummers, consum’actors and new models such as self-consumption, alters the operating requirements and constraints of the grids themselves and imposes the management of increasingly massive data that would be unworkable without recourse to AI.
Distributed Vs. Centralized
The energy market is moving away from a model with centralized power plants only and entering the era of distributed grids and peer-to-peer markets. Multiple elements of the energy ecosystem are evolving at a dizzying speed. We are seeing a very complex market emerging, where the distribution company needs to allow more and more renewables and flexible energy assets to be installed behind the meter while maintaining a stable local grid. At the same time, prosumers who have installed such flexible assets want to optimize their energy flow to maximize the value of their investment.
A steadily growing challenge is the emergence and accelerated the growth of a decentralized generation, where private users, bigger or smaller, generate and use their own electricity from renewable sources, such as wind and solar power. This complicates supply & demand oblige utilities to buy surplus energy from private users, who produce more electricity than they consume and send it back to the grid. Since 2010, the use of solar energy has substantially increased and this exponential trend is expected to continue with photovoltaic cells, devices generating electricity from sunlight, reducing costs and increasing efficiency.
An extending decentralized production
The current systems have generally not been designed to take into account this diversification of energy sources, particularly the increase in renewable resources. For example, in many American jurisdictions, when demand outstrips supply, utilities activate fossil fuel-based power plants, known as “state of the art” power plants, just a couple minutes in advance to avoid a cascading disaster. This procedure is the most expensive and, but also, the most profitable part for these companies. It results in a higher electricity bill for consumers and an increase in greenhouse gas emissions into the atmosphere. These problems will be exacerbated as energy demand is expected to increase substantially in the coming years. To avoid these non-optimal (for the least) operating mode with Intelligent Energy Storage (IES), AI can enable automatic learning algorithms, combined with data on these complex networks and real-time meteorological data (from satellites, ground observations and climate models), to be exploited with the full potential to predict the electricity generated by RES, such as wind, sun and oceans.
Combined with other technologies such as Big Data, the Cloud and the Internet of Things (IoT), energy storage with AI can play an important role in power grid management by improving the accessibility of power sources. renewable energies.
The (Deep) Learning curve
AI can greatly help to manage electricity consumption so that big utilities or an even smaller grid with DER can sell when it’s expensive and buy when it’s cheap. Machine learning, and especially deep learning, algorithms can be applied in the energy sector in a very interesting way in this context. As the end-users are becoming “prosumers”, smart devices are proliferating, big data is available for analysis, renewable energy sources are growing, and business models and regulations are adapting.
Combining it all together can help get to the point where energy flows and/or is stored at the optimal timing, direction and volume. With artificial intelligence algorithms to determine when to produce, consume, store and trade energy, to the cost-benefit of the end-user, the service provider and the grid operator. With thousands of emerging energy communities, this vision might become clearer and perhaps even the main reality in the coming 5 to 10 years.
More and more sustainable communities, utilities and operators are currently under simulation or in the first phases of pilot projects. With Internet of Things (IoT) demanding more than 10 billion smart devices with over 100 million electric vehicles (buses, trucks and passenger cars), with more than 1 billion prosumers (private and industrial) having their own “production” of kWh (solar or else), all predicted by the year 2025 — it’ll be a huge challenge to maintaining reliability and secured supply and grid stability.
Expectations of how DERs will evolve in the coming years are multiple. But these changes require a completely new operating paradigm, and there is no better test for technology than real life. New models involving Artificial Intelligence, Energy Storage and Renewable are already being applied at various levels in many states on all continents, not to mention Australia, California, Germany, China, Costa Rica, Israel and many other countries around the world. It’s not required to be psychic to envisage that AI and DER will be the Transformational technologies that will soon be making the new grid model.
Source: Novacab Multiple DERs with Hybrid Energy Storage and Artificial Intelligence.
This article is an addition of a series on Artificial Intelligence, Grid, Renewable and Energy Storage by Stephane Bilodeau, ing., P.Eng, PhD, FEC. Founder & Chief Technology Officer, Smart Phases (Novacab), Fellow of Engineers Canada and expert contributor to Energy Central Network and Medium.