Data visualization is bringing
Delaware Electric Cooperative’s business into sharper focus.
The Greenwood-based co-op, which serves more than 102,000 members in Kent and Sussex counties, is heavily residential, serving a mixture of suburban, rural, and beach areas. It has been adding about 3,000 consumer-members annually, with most of the growth in its beach-side territory.
To manage that growth and its established accounts as efficiently as possible, Delaware Electric uses a digital dashboard through which staff can track roughly 150 key performance indicators (KPIs) in easily accessible formats.
“We saw the benefit of having good data, accurate data, timely data that told us a story,” says Delaware Electric President and CEO Bill Andrew, which requires the information to be visually compelling and easy to understand. “We wanted to not have the data printed out on green bar paper. … That’s why you see all the graphics.”
Employees across the co-op can use the dashboard. The front page has 15 visual presentations of data on system control, information systems, member services, safety, finance, strategic initiatives, and more.
The dashboard also incorporates a page of options of particular importance to the CEO, including member metrics; electric load forecasts; operational data from PJM, Delaware Electric’s regional transmission organization (RTO); and a marginal pricing map.
“I keep a lot of these KPIs up front because it allows me to determine what direction the business is going in,” Andrew says.
Staff can drill down to specific indicators within categories. For example, from the metering tab they can access specific data on new connections.
“This is a leading indicator for me to see what we’ve got coming down the road,” Andrew says. “I look at this every day.”
They can also see a snapshot of upcoming jobs.
“And I know what my backlog is,” Andrew says. “I know what I’ve got to keep my engineers working.”
Much of the data is updated hourly, while less time-sensitive information is updated daily or even monthly.
“We’ve set this up so each of my departments, each of my engineering teams, can utilize this in their operations,” Andrew continues. “They understand what the key performance indicators are telling them on a real-time basis, allowing them adjust to that. Do we need to work Saturday? Do we need to work overtime? They’ve got the information at their fingertips that allows them to make those decisions.”
Creating a data habit
David Pinney, NRECA’s analytics research manager, says such capabilities are becoming essential to electric cooperatives. Advanced metering infrastructure and other smart systems have dramatically increased the data available to cooperative managers, but without the structure that allows them to make sense of that data, to see what it reveals about their system, it has little value.
“It’s not possible to make certain data-based decisions without accurate visualization,” Pinney says.
A good example is right-of-way management, he notes. Getting a picture of how issues can be addressed efficiently and effectively is difficult when relying on field reports, but co-ops have found that integrating the data into a map through a geographic information system transforms the process.
Andrew says taking full advantage of data visualization requires a cooperative-wide commitment.
“We have to make it a part of our daily business, our daily routine, and we have to ask the questions, ‘What is this telling us? What don’t I know?’” he says. “The whole concept is, ‘How can I utilize my data to help me do my business better?’ And that’s a cultural thing as much as it is analytics. It has to be habit forming.”
To help create those habits, the data has to be easily and broadly accessible to co-op personnel.
“It’s on our intranet, so anyone can get to it through their computer or their mobile device,” says Dwayne Street, Delaware Electric’s chief information officer. “We also have what we call display boards—screens—posted throughout the buildings, and they run through multiple key performance indicators, so if you’re just walking down the hallway, you might see a KPI and something clicks in your head. That’s part of that cultural shift.”
Different co-ops, different approaches
Andrew says the embrace of a data-driven approach has played a significant role in allowing Delaware Electric to operate more efficiently. While the average electric co-op last year had about 335 accounts for each employee, he notes, Delaware Electric had 666 accounts per employee.
As a larger co-op, Delaware Electric maintains a seven-person IT department, which built and maintains the performance portal. The co-op uses a “best-in-breed” approach to software and technology, which means the dashboard is taking data from products that come from different vendors.
When Delaware Electric set up the first, simpler version of its dashboard about a decade ago, Street says, the biggest challenge was integrating data from different systems. Today, the performance portal captures data from a range of internal sources, including SCADA, the automated metering system, the GIS, the outage management system, and accounting.
“Our solution is great for us,” Street says. “But it’s not for everyone.”
He notes that co-ops with limited IT resources frequently use vendors that provide a comprehensive suite of software and tech services, and many of those vendors offer customizable data dashboards.
“Grabbing one of those out-of-the-can options can be an excellent place to start,” he says, “and then working with the vendor to individualize it for that co-op because every co-op is different, and every co-op has its priorities.”
Employing AI and VR
One of Delaware Electric’s recent priorities is making use of artificial intelligence (AI) to tease even more insights out of the data. Tracking and understanding changes occurring over many years “can be very hard for the human brain … but computers pick it up very fast,” Street says.
To that end, the co-op recently hired a data scientist who is designing computer models that will use AI and machine learning to spot long-term trends within the data that can be hard for humans to see.
The models should “give us some insights we might be overlooking,” Street says.
AI is a complementary technology that can work in conjunction with data visualization to help cooperatives track their operations, Pinney says. Although AI allows a co-op to automate operations that were previously controlled manually, “visualization is still going to be important for understanding what those AI algorithms are doing,” he says.
Looking ahead, Pinney sees “sensor fusion,” the combining of data from sensors out on the system, playing a larger role in data analysis. Sensors can significantly increase the amount of information about what’s happening on a distribution system, enabling a more granular view of operations. But that data needs to be presented in a comprehensible fashion for staff to take full advantage of it.
Co-ops are beginning to experiment with augmented and virtual reality (VR) to help people glean the data, Pinney says. The interactive visual representation of what’s happening on a system over time, for example, would show SCADA data when looking at grid assets in the field or in training simulations.
Andrew says data visualization is part of an embrace of smart systems and other technology that is critical for cooperatives going forward.
“In order for us to remain competitive, the next level of reliability and of cost control is not going to be through bucket trucks and linemen. It’s going to be through technology,” Andrew says. “This is foundational with respect to technology because it really highlights the value of the data.”
2020 NRECA Annual Meeting: Developing a Data Strategy Can Help Co-ops and Their Members, Experts Say
Video: Bill Andrew Discussing the DEC Performance Portal at the 2020 NRECA Annual Meeting