There is great value in big data if you know where to look—and that’s especially true for the electricity sector, which is well positioned for significant benefits from capturing, managing, and analyzing the tremendous amounts of data from the grid. By creating a smart grid—one that’s instrumented, intelligent, and interconnected—utilities are moving into the big-data world where they can predict and, to a certain degree, prevent problems from occurring before they happen and exploit opportunities they didn’t know existed.
To achieve these benefits, utilities must collect and analyze oceans of data to gain operational and customer insights. Data must be managed and stored in a way that maintains privacy and security, but allows analysts to work with it. When that architecture is in place, the data can be transformed into useful information.
How big a job is that? The world’s systems and industries are producing more useful data every day. The number of connected devices—this so-called “Internet of things”—is growing by some estimates at a 45 per cent compound annual growth rate through 2015. We are creating the equivalent of 2.5 quintillion bytes of data every day. Every two days, we create the equivalent of all the data that existed in the world up to 2003.
To manage and use this data, utilities must be capable of high-volume data management and advanced analytics that support data-driven decision-making and planning by transforming data into actionable insights. But it’s not an easy transition. Many utilities are simply overwhelmed by the sheer volume of data they now have access to, and utilities are all at different stages of adoption and transformation.
For utilities, the benefits of big data fall into four categories—and the potential returns are convincing.
Preventing equipment failure
Each electricity grid has thousands of devices attached to it—transformers, switches, circuit breakers, automatic reclosers, et cetera—and these devices are all engineered to perform at certain loads and weather conditions. Condition-based maintenance is the process of tracking the condition of those devices continuously, constantly assessing performance against their design parameters to predict impending asset failure and prevent costly downtime. Is it performing at the proper level? Has it been performing outside of its specified parameters for a period of time?
If a device has been operating at 150 per cent of its design range for an extended period of time, its likelihood of failure is greater. Data analysis enables us to predict and address the problem before it causes a larger problem that will be much more expensive to fix. It also reduces unplanned service interruptions.
Many jurisdictions have mandated the use of smart electricity meters—some along with a time-of-use price structure—to help consumers and small businesses manage their costs. However, what customers may not see is the other data smart meters provide that can help create improvements and efficiencies within the grid. For example, a utility can cross-reference information from a smart meter and from its outage management system. If the utility has smart meter info, it can link it with outage info and could predict and manage outages. With large outage, after a storm takes down a substation, a utility will know the substation is down, leaving an entire neighbourhood without power; however, a utility may not be aware of smaller outages (nested outages) where perhaps a tree fell on a power line knocking out power to only a few houses. With the ability to remotely interrogate smart meters to determine if power has been restored, the utility gets the full picture and can restore service much faster and more efficiently.
Understanding customer behaviour
The data from smart grid devices can provide companies with unprecedented capabilities for rethinking their approach to customer service and customer operations. From a customer perspective, this enables utilities to better understand how their users are segmented and offer services like personalized guidance on energy usage patterns and incentives to reduce consumption. Before big data, utilities couldn’t recognize a customer’s energy-consumption trends or anything else about their needs. In the past, a criticism about utilities has been that they can behave like a monopoly—that they don’t have to demonstrate they care about their customers because their customers can’t go anywhere else. But the message utilities can send back, as they embrace big data, is that they can tailor products and services to meet individual needs. They can also provide an enhanced customer experience via the web and social media.
Once utilities have good customer profiles, they can be much more responsive. For example, if a customer calls (or accesses the web portal) to discuss an issue or get information about a bill or an outage, the customer service representative has all the information to know what has occurred to that customer in the past or their preferences. The utility can tailor its message to ensure the customer knows what is planned to address their issue or provide them with additional information that can leave the customer with a positive experience.
Integrating renewables into the grid
The percentage of energy generation from renewable sources is slowly creeping up, and societal pressure will maintain that trend for the foreseeable future. But it is not an easy task to integrate distributed renewable sources with the grid. Renewables are variable; if it’s cloudy or still, the renewable source will not be generating as much electricity as planned. Utilities must be able to predict that variability and determine whether particular loads should be dispatched or other loads should be turned on to compensate, maintaining a consistent power quality and voltage so customers aren’t affected. Using analytics, utilities can not only improve wind turbine placement but also control and monitor the quality of wind power and solar energy.
For example, new advanced power and weather modelling technology is now available to help utilities increase the reliability of renewable energy resources. Hybrid renewable energy forecasting(HyRef) uses weather modelling capabilities, advanced cloud imaging technology, and sky-facing cameras to track cloud movements while sensors on the turbines monitor wind speed, temperature, and direction. When combined with analytics technology, the data-assimilation-based solution can produce accurate local weather forecasts within a wind farm from a month in advance to 15-minute increments. Using local weather forecasts, HyRef can predict the performance of each individual wind turbine and estimate the amount of generated renewable energy. This insight will help utilities manage the variable nature of wind and solar, and more accurately forecast the amount of power that can be redirected into the power grid or stored.
How do we get there?
The opportunities from big data analytics are significant, and they cross the three main utility domains: grid operations, asset management, and customer operations. In the past, these three domains have not been tightly integrated. Going forward each can benefit from better integration. For that to happen, there must be excellent governance. Who is going to collect the data? How will it be stored, accessed, shared, and protected? How will its accuracy be determined and maintained?
Big data offers big benefits, but only if it’s reliable and it can be mined for useful insights that utilities can act on. Right now, no utility has reached the big data nirvana state, but many have laid the foundation, and that will change utility behaviour, from both an operational and societal perspective, for many decades to come.
Bruce Orloff is the smart grid leader for IBM Canada.