Deep Learning Will Keep EVs From Killing The Power Grid

can be reached at
can be reached at
Share on facebook
Share on twitter
Share on pocket

As more electric vehicles come on line, there will be increasing demands on the power grid. The International Energy Agency (IEA) predicts that the number of electric cars on the road will rise from 3.1 million in 2017 to 125 million in 2030.  That’s a huge amount of cars vying for juice.

  • The influx of EVs in the next decades will put a strain on the grid
  • Using Oracle’s Analytics, utilities can manage power supply and avoid outages
  • Analytics data can help customers know best times to plug in and save money

We’ve probably all encountered the situation on a blisteringly hot day when demand for aircon shut down the grid for a period of time because the system was overwhelmed. In about eleven years, the grid will have to service 40-times the amount of electric cars. A report by McKinsey estimated that EV’s could account for additional growth of 1-4% in peak load over the next few decades. That number jumps to 30% in urban areas with high rates of EV adoption.

Photo by Tim Mossholder on Unsplash

Beware the surge

Without a planning system in place, this surge in usage could wreak havoc on power supplies. So what are utility companies with decentralized and outdated management systems supposed to do?

Oracle Utilities Analytics Insights to the rescue. This new program, already being tested by utility companies, taps into deep machine learning to identify the presence of an EV, show the time and frequency of charging and disaggregate the energy being consumed by the vehicle with advanced metering infrastructure (AMI) data. Utilities will then know the amount of power EVs are using, when, and for how long. From this data, they can plan energy distribution to power EVs at scale.

Customers can benefit with savings

In addition, the data can help utility companies map out the best time for customers to plug in, providing programs to educate and reward owners who charge during non-peak hours. Charging an EV bumps up typical household energy use by 15%, which can potentially double during peak demand times. Through these types of engagement programs, Oracle also sees an opportunity for utilities to buy-back energy from customers’ EV batteries to balance out any major swings.

The paradigm shift from combustion engines to electric motors necessitates not just a new infrastructure, but management systems to make it practical. Tools such as Oracle Utilities Analytics Insights will make the transition easier for both utilities and customers.

About the Author

can be reached at
Close Menu

We use cookies and browser activity to improve your experience, personalize content and ads, and analyze how our sites are used. For more information on how we collect and use this information, please review our Privacy Policy. California consumers may exercise their CCPA rights here.