AI In Renewable Energy, Part 2: Grids and Storage Systems

Oh hey! Welcome to another blog series about my favourite thing in the world, AI in renewable energy. I researched this in great depth in January 2024 and went to a seminar at the University of Cambridge on it, which was the greatest hour in my life. 

Recap

In my last post, we talked about why use AI in renewable energy and we began to look at patterns and predicting generation.

Smart Grids

There is nothing more I love in the world than a smart grid. Modern grids have to integrate a greater amount of energy than ever before into themselves. The demand is higher than ever and therefore the supply has to be higher too.

This modernised smart grid quite simply cannot function without artificial intelligence.

As the grid relies on varied sources of power (in renewables alone we are looking at hyrdo, wind, solar, maybe one day tidal), live sensor data provides continual live feedback which feeds into the AI that helps us to manage the grid. Similarly to in the last post on generation, the AI that manages the grid feeds off of data such as:

• weather stations information
• patterns of consumption
• the amount of energy produced per station

This can help us and allow us to make real-time decisions quickly about the grid. This saves money and resources by helping avoiding excess energy, either storing or generating it. It also protects the grid by helping to avoid overloading it with excess power. Again balance must be kept at all times or else the grid just shuts down.

An Optimal Load Flow

By helping us to optimise the flow of the load of the electricity through our grid, AI once again makes the grid cleaner. However, as we saw last time, AI also allows us to increase the security of our grid. One of the biggest barriers to green energy is energy security. By allowing us to increase our energy security through predictive models, AI massively facilitates the transition to renewable energy. And this is one of the reasons why it is so crucial to the renewable energy transition.

Efficient Energy Storage Systems

Energy storage systems such as batteries have greatly benefitted the grid. AI can use its predictive models from patterns to decide when to store power and when to release it to the grid. This is such a crucial part of renewable energy and grid stability that I cannot even begin to explain it. The grid relies on balance to work and only AI can think fast enough to decide in real-time what to do based on the huge amount of data available.

As always these models have access to data on past energy consumption, as well as data on patterns in predicted energy surges, thus knowing - when do we store energy; when do we release it to the grid; and when do we power up our batteries.

All of this allows AI to decide whether to release stored energy at times of crisis; or to have it as something that we can fall back on in simple times of need and grid-rebalancing. This arguable improves the lifecycle of our batteries. I suppose the only question is - don't we, as humans, have to make the decisions on how urgently we want our grids to draw on batteries, before we can program these into our AI? Join me next time for more storage systems, and for AI-driven site analysis.