Is Artificial Intelligence the Next Step for Smart Grids?
by Bob Shively, Enerdynamics President and Lead Facilitator
"If I were starting out today and looking for the same kind of opportunity to make a big impact in the world, I would consider three fields. One is artificial intelligence. We have only begun to tap into all the ways it will make people’s lives more productive and creative. The second is energy, because making it clean, affordable, and reliable will be essential for fighting poverty and climate change. The third is the biosciences, which are ripe with opportunities to help people live longer, healthier lives."
~ Bill Gates, writing in GatesNotes, “Dear class of 2017…”
Many electric utilities are moving quickly to implement the smart grid. Distribution systems are no longer passive systems moving power from the transmission-distribution interface to the customers. Instead, they are active networks that manage the movement of power from both substations and prosumers to customers or, in some cases, even back into the transmission grid. This requires new data to understand what is going on. The result is that electric grids must become networks that blend power delivery, communications, and data management.
Source for concept: IEEE 2030 Smart Grid Interoperability Reference Model Webinar, Sept. 6, 2018
These networks are becoming complex entities with intelligent devices that analyze great amounts of data and, in some cases, act autonomously without human intervention.
Source: PG&E Corporation Business Update, July 28, 2017
The number of active devices connected to the grid will skyrocket, as will the amount of data requiring transmittal and analysis.
Indeed, researchers recently estimated that in one decade the Dutch electric grid must be able to make sense of:
- 8,000,000 smart meters each being sampled every 15 minutes
- 800,000 electric vehicles (EVs)
- 45,000 behind-the-meter batteries
- 160,000 grid nodes
- 10 intelligent devices per household (equaling approximately 80,000,000 smart devices)
Needless to say, utilities are quickly finding themselves overwhelmed with so many active devices and the need to make sense of so much data. It is impossible for distribution operators to continue traditional ways of operating the systems. Many utilities that less than a decade ago were operating the system based on a physical map on the wall will need to now implement artificial intelligence (AI). We will see AI permeating through the energy sector very soon. 
But interestingly, as noted by the Brookings Institution: “Few concepts are as poorly understood as artificial intelligence. Opinion surveys show that even top business leaders lack a detailed sense of AI.”
Unfortunately, there is no accepted universal definition of AI (let alone, AI in the energy sector). Typical definitions tend to speak to computer systems able to perform tasks that normally require human intelligence and with the capability of imitating intelligent human behavior. Key factors include:
- Making decisions without human intervention by using data from various sources, analyzing the data, and acting on insights derived from the data.
- Continually studying the relationships of data and outcomes through machine learning by collecting data, looking for relevant trends, and analyzing how those trends affect outcomes.
- Adapting algorithms to continually improve outcomes by adjusting responses.
So, for example, let’s imagine a distribution feeder with continuing growth of rooftop solar, price-responsive demand devices, electric vehicles, and behind-the-meter batteries. The behavior of each of these will change based on numerous factors, but all will affect the voltage profile on the circuit. It would take a long time for a distribution engineer to perform a study to figure this all out, and by the time he or she did, customers may have already added thousands of more new devices. Instead, we can imagine distribution assets with enough intelligence to continually model expected voltage behavior and automatically adjust settings even moment by moment to maintain acceptable voltages.
As we often tell our seminar participants, utility companies must think of themselves as technology companies, not just energy companies. The importance of this distinction will continue to grow as artificial intelligence is required to manage the smart grid.
Want to learn more about Distributed Energy Resources and their impacts on the transmission and distribution grid? Schedule our Distributed Energy Resources and Microgrids classroom seminar. Email us or call 866-765-5432 ext. 700 for more details.
Footnotes:
The Smart Grid’s Data Generating Potential, Marco Aiello and Giuliano Andrea Pagani
