Toward a Self-Healing Grid

Data analytics have been part of the utility industry for decades, giving companies the information they need to run plants safely. This data holds great value, as it allows us to maintain our extremely high level of safety, and more effectively generate and deliver power to customers.

The information also allows utilities to upgrade and adapt our technology to embrace the analytics race for real-time data. Unfortunately, change cannot be rushed in the utility industry.

As chief information officer and former chief information security officer at the New York Power Authority, the nation's largest state power utility, my background in cybersecurity allows me to see the bigger picture and analyze the risk of opening that door much more studiously.

There are sectors of the economy where results are subjective and well-suited to analytics, so we see data analytics offering suggestions on items to buy, or people to date, or movies to watch. In sectors where fraud detection or behavioral analysis on a massive scale are needed, as in the utility industry, even the nascent capabilities of data analytics seem useful, but perhaps sometimes offer a false sense of security. This is because data analytics focuses on optimizing existing processes, which can be of value under current conditions but has the potential to cloud the broader picture or even lead to complacency about future needs.

The energy sector has long dealt with competing priorities. That is, the tradeoffs between delivering very high reliability and the inefficiencies associated with redundancy and readiness to satisfy peak demands under worst-case conditions.

As we look at data analytics, we may ask ourselves what problems do we have where data can offer useful insight, unlock opportunities we would have otherwise abandoned for lack of insight, and alert us to risks and opportunities we otherwise wouldn't know about.

Many vendors are starting to offer "predictive analytics" tools, which pull information from existing data sets to determine patterns, and predict future outcomes and trends. This is a very exciting opportunity for the industry to be more proactive in the management and maintenance of grid equipment. A representative from GE, for example, at its recent "Minds and Machines" conference, told a story concerning an oil rig that the company was monitoring with predictive maintenance tools. GE uncovered a problem with a piece of equipment that was about to fail and notified the owner, which saved several million dollars in downtime. That is where analytics tools basically pay for themselves - predictive proactive maintenance resulting in minimal downtime and cost savings.

At NYPA, analytics help us provide a high level of service to our customers and protect our assets. We can plan for the future, help grow New York State's power grid, and provide existing and potential customers with data that can help them plan for their future. Predictive analytics on the grid will improve response times to outages, and - more importantly - get the electric utility industry to the stage where we can replace equipment before it fails.

Another example of how data analysis provides benefits is the NY Energy Manager (NYEM), which is New York State's first energy management network operations center. NYEM was recently introduced by NYPA as part of New York Gov. Andrew M. Cuomo's BuildSmart NY program to reduce energy use 20 percent in public facilities by 2020. Data analytics allow NYEM - which is operated by NYPA at the Colleges of Nanoscale Science and Engineering at SUNY Polytechnic Institute in Albany - to provide real-time data on energy use to more than 3,000 public facilities across the state with the potential to serve even more. Participating facilities can improve building energy performance, lower the state's utility bills and reduce greenhouse gas emissions. The data analytics gathered by NYEM will allow government building operators to be better informed when planning for future energy use needs. NYEM can also provide technical expertise and ongoing training to participating organizations for managing their facilities.

NYPA also uses analytics to look for weather patterns, allowing us to optimize the operation of our power plants and transmission lines. Trend forecasts are used by the New York Independent System Operator, a not-for-profit corporation responsible for operating the state's electricity grid, to anticipate energy supply and demand and facilitate grid optimization.

While data analytics can be extremely helpful, they are not infallible. Data analytics can't replace a human's intuition or interpretation of the numbers. Despite that limitation, however, continuing developments in data analytics will allow utilities to predict the impact that changing supplies and demands, such as electric vehicles and renewable energy sources, will have on existing equipment. We can take data and produce information that will allow us to improve the system with a more holistic approach. Today, most of the energy industry is relying on equipment to tell us about outages after the fact. The future grid should be a self-healing network, with predictive analysis used to proactively manage the electric grid's assets.