Utility’s Big Data Challenge
The smart grid revolution is unleashing torrents of data. Utilities face an imperative to develop ways to transform those bytes into system improvements and innovative services
Say “high tech” and utilities are not likely to be your first thought. Yet from smart phones, to web servers to the broadband links in between, utilities crank out all the electricity that keeps our digital economy humming. Go behind the scenes and you’ll see that utilities have still more in common with typical high tech companies than first realized. In fact, they too are making large investments in their information infrastructure, which has to be shared, managed and supported by complex software, with the goal to drive increasingly advanced insights and services.
For utilities, these efforts come after a century of simply producing and selling electric power. As they replace aged analog systems with digital upgrades, electricity providers are recognizing there are new opportunities to transform bytes of information into improved services and new revenue sources, such as home energy efficiency sources. Just like high tech companies such as Facebook or Google, utilities are recognizing more opportunities in mining big data – from better asset management, to decreasing operational and capital expenses, offering new products and services, such as home energy efficiency services, electric vehicle charging, or demand response. Monitoring and analyzing the cascade of data collected from the variety of devices and sensors allows utilities to run their businesses more efficiently.
“Smart” is the keyword common to these efforts. Whether smart meters, smart grids, or smart buildings, the initiatives share a common goal: Instrumented, i.e. smart meters, or device sensors; Interconnected, i.e. communication network, and Intelligent, insights and optimization from data collected. These efforts are to enhance the performance of practically any device that’s plugged into the grid, from everyday kitchen appliances, to electric vehicles, to huge power plants. Already the efforts are beginning to save energy and improve grid reliability.
However, with progress comes challenge. As the grid grows smarter, utilities must adapt to the complexity of handling the growing data flow. To make the most of this emerging network of things, utilities face a call to deploy analytics, a class of advanced software that can help discover patterns in this data deluge, allowing them to take action based on these insights.
To understand the scale of the task facing utilities, consider the impact of smart meters. Typically, standard analog meters were read once a month, for a single number representing energy consumed. A digital smart meter, by comparison, might relay a variety of indicators every 15 minutes or more frequent. Those updates pile up fast, to 35,000 per year. If not properly managed, the 3,000-fold increase in data volume could be overwhelming. Add in the billions of other devices on the grid that are being networked globally, and experts predict that utilities could soon be handling more data than any other business, including the traditional communications industry.
Advanced analytics software is already helping utilities to tame this rising tide of data. Here are three types of programs, up and running today, that point to what will be possible as utilities go digital.
The first fruit of these efforts can be seen in the way smart meters are modernizing the way in which customers interact with power providers. In the past, a new request for a move-in or move-out was prone to delays as the central office had to dispatch a technician to switch power off at one address, and perhaps another service visit to switch power on at a new address. In the back office, the process for update and transfer of account information was disposed for delay and error. Digital meters help slash through most of this tedious process and thanks to remote control, in a single transaction, digital meters can be switched off, or on, at different addresses, while the related bills are calculated through an integrated and automated business process. It’s an upgrade that saves money, and speeds service on a very common type of transaction.
Behind the scenes, the huge flows of data are also opening up the possibility of more fundamental changes in the way power networks operate. For instance, smart meters can relay pricing information to help consumers shift energy-intensive tasks—such as laundry drying or pool pumping—to hours when demand is low. That translates into real savings for end users. For the utility, when this kind of shift happens across enough customers, the peak power needed from power plants declines, too. Utilizing the data flow from many meters can let the utility measure and confirm these shifts in energy use, reducing the need to run high cost power plants and lowering the risk of brown outs. In the long run, less spending on generating energy or building power plants means lower rates for consumers.
Less troublesome turbines
Big data is also transforming the way utilities decide when and where power is created. It may seem simple, for example, to decide where to locate a windfarm – build where the wind blows of course. But subtle variations in the location, height, and orientation of turbines can deliver substantial improvements in the amount of electricity a windmill can generate.
For Vestas, the world’s largest manufacturer of windmills, better data is the latest competitive feature that comes with its most advanced turbines. Vestas is tapping into the power of an IBM supercomputer and big data analytics software to model past, present, and future wind patterns to optimize the location and design of sites its customers are developing. Just a few years ago, site analysis of this sort was constrained by the huge amounts of data necessary to simulate weather patterns. Vestas’ current system is on track to digest 20 petabytes of information—the equivalent of more than 20,000 terabytes. The system takes hours to process the volume of data that not long ago would have taken months.
Using a supercomputer to crunch and analyze this amount of data Vestas’ engineers can assess an unprecedented variety of operating conditions, from wind speed at different heights, to humidity, moon and tidal phases, sensor data, seasonal shifts in weather to name a few, which is then used to predict the future performance of the turbines and the optimum time for maintenance.
Analysis of big data can also help existing wind turbines with real-time monitoring of the system performance. Servicing rotors or controllers of wind turbines high up on a tower is costly, and dangerous, so the incentive to detect faults before failure is enormous. Data drawn from hundreds of sensors on thousands of similar turbines can quickly detect worrisome behaviors before they cascade into more serious damage. For utilities’ customers, this means fewer power disruptions; for utilities, that means steadier revenues and lower operations and maintenance costs. |
By sheer geographic scale, the largest part of our power system—and hence the most vulnerable to trouble—are the hundreds of thousands of miles of cables that link power plants to millions of customers. Accordingly, utilities see extraordinary potential in adding sensors and networking intelligence to these far-flung devices. It’s estimated that 10 billion assets will be linked up as the smart grid matures. The bulk of these new nodes will be located between customer and power plants. The pulse of signals from these devices is already helping to prevent grid failures and as smart sensors multiply, systems can monitor feedback from thousands of devices simultaneously, watching for signs of troubled equipment before it fails. Detecting a fault early not only gives a utility the means to prevent an outage before it occurs, but also if an outage does occur, it provides crews a head start to avoid costly black outs.
No matter how well managed the grid is, storms and other natural disasters inevitably can cause power outages. In this case too, the data collected and analysis are helping to shorten the duration of outages when they do happen. On the east coast, for instance, a major utility is combining two types of data to fortify its network against natural disasters. It starts with a cutting edge weather forecasting system able to predict conditions minute-by-minute, in areas as small as a neighborhood. The utility can then superimpose those forecasts on virtual maps of its assets—from substations to utility poles—and pinpoint facilities facing the highest risk of storm damage. This advance knowledge lets the utility plan on vegetation clean-up, and pre-position repair crews and material in the highest-risk areas. That way, if a key distribution main line is knocked out by a fallen tree for example, the utility can speed repair crews and replacement equipment to the right spot.
Though going on largely behind the scenes, the scale of digitizing the grid ranks as one of the largest info-tech initiatives in U.S. history, on par with the creation of the Internet, or the automation of industries such as manufacturing, banks, airlines and telecoms.
The opportunities are just as great. More so than any time since electricity was commercialized, utilities have greater choices, and more potential, to innovate than any time in their history. Their challenge is that the deluge data streaming from an increasingly digitized grid is full of value not yet realized. To get the most from that flow of bytes, analytics systems offer a fast multiplying kit of tools and technologies to assess and convert that data into services and products not yet imagined.
The opinions expressed in this article are solely those of the author, Mozhi Habibi is responsible for leading IBM’s energy and utility solutions strategy worldwide, focusing on emerging solutions in the industry, including smarter generation, intelligent utility network, distributed generation and business analytics & optimization.
Other Posts by Benjamin Lack
What are the emerging energy and utility trends?
"The Future of Energy and Utilities: An IBM Point of View."
|More coming soon...|
The Energy Collective
- Rod Adams
- Scott Edward Anderson
- Charles Barton
- Barry Brook
- Steven Cohen
- Dick DeBlasio
- Senator Pete Domenici
- Simon Donner
- Big Gav
- Michael Giberson
- Kirsty Gogan
- James Greenberger
- Lou Grinzo
- Jesse Grossman
- Tyler Hamilton
- Christine Hertzog
- David Hone
- Gary Hunt
- Jesse Jenkins
- Sonita Lontoh
- Rebecca Lutzy
- Jesse Parent
- Jim Pierobon
- Vicky Portwain
- Willem Post
- Tom Raftery
- Joseph Romm
- Robert Stavins
- Robert Stowe
- Geoffrey Styles
- Alex Trembath
- Gernot Wagner
- Dan Yurman