Alabama Power and E Source earn U.S. Department of Energy award for graph database project

Alabama Power and project partner E Source have been named by the U.S. Department of Energy Office of Electricity as a Phase 1 Data Track winner in the second round of the Digitizing Utilities competition, earning a $75,000 prize to help develop a plan around their project submission.

The project, titled Enterprise GraphDB to Better Data QA & Analytics, focuses on creating a graph database gathering data and information from Alabama Power’s electric grid to help strengthen and improve reliability.

In Phase 1 of the competition, project partners formed teams of developers and were tasked with finding utility partners and identifying data or cybersecurity challenges. After consideration by the Office of Electricity, the prize was awarded to nine teams that proposed solutions to the problems. In Phase 2, these teams will continue to work together during the next four months to develop their solutions for an opportunity to win additional funding dedicated to the further development and implementation of the project.

Jesse Woods is the data analytics manager for Alabama Power and took part in the project. Woods believes the graph database solution developed by his team and E Source would have a significant impact on grid reliability.

“As we’ve developed more and more applications over the last several years, we’ve started to see a significant opportunity to bring together more of the data sets — more of the information that we have in different systems — and bring that into a single place that can be kind of a gold standard for using for our analyses and projects,” Woods said. “By building a database, all of our information — historical and current — is going to allow us to make analytical projects and products to be developed much faster than they have been historically because we’ll have all the information we need to build projects and to do things like machine learning or the development of artificial intelligence.”

In simple terms, a graph database is a database that uses graph structures to link data and information together. The graph relates the data items together, creating relationships that allow data to be linked together and show their correlations.

Woods uses social media as a way to illustrate how a graph database functions.

“A graph database really became popular when social networks and social media came about because the idea of representing pieces of information as connections to others,” Woods said. “So, James who’s friends with Josh who’s friends with Amy who’s friends with Paul who’s friends with whoever — how we measure those connections between people.”

Woods went on to explain how a graph database could translate over to Alabama Power’s grid.

“If you think about it that way and when you think about that’s how social media works and that’s how connections are represented in a data form, it’s very similar to our power distribution or power transmission system,” he said. “A pole is connected to another pole by a wire, and that wire may lead to a transformer. That transformer is connected to Jesse’s meter, which feeds power to Jesse’s house.

“Being able to represent how our system works and is connected in a database-type format is exciting and we think is going to enable some new analytics that we haven’t been able to do before.”

The graph database could effectively and efficiently provide data to Alabama Power’s power delivery personnel to help strengthen the grid, thereby improving reliability for customers. Data and information added to the database would include a variety of both current and historical information, such as past causes of outages or their locations. Examples such as lightning, tree limbs — or even squirrels — would be added to specific outages in addition to other factors like weather conditions or time of day. All of this would be linked together in the graph database, giving real-time information to Alabama Power.

“What these graph databases do is they allow you to really traverse over the data very quickly and be able to understand ‘Ok, this is where the outage occurred, this is all the pieces of equipment that we have in power delivery on either side of that outage,’ and so we can take actions to harden the system based on that outage or do other types of improvements ultimately to help prevent future outages and improve our reliability,” Woods said.

He likened the system to receiving a scan at a doctor’s office. By being scanned or tested by a doctor, a patient can make changes ahead of time or be prescribed medicine to avoid a potential health issue. The graph database would serve the same purpose by predicting potential issues before they happen — improving reliability and impacting customers by reducing outages.

“Our goal is to be able to predict outages before they happen, with the idea being that we can use information that comes from this graph database,” Woods said. “And that’s things like where we’re seeing additional wildlife or where we have trees touching our lines that are potentially causing ‘blinks’ for our customers — or what we call ‘momentary outages’ — if we can use those momentary outages to ultimately predict future sustained outages, then we can take actions to mitigate those outages before they happen.”

To learn more about the Digitizing Utilities competition and see a full list of winners, click here. To learn more about Alabama Power’s commitment to innovation, click here.