Big data is a big topic when it comes to the smart grid. GigaOm’s Katie Fehrenbacher commented that it was one of the most obvious trends at the recent DistribuTECH 2012 event. The data generated by smart devices throughout the grid could reach truly monumental proportion. Consider all the types of data that may be collected by devices from the meter all the way back to the back office — time of use data, pricing data, weather data, power system data (like waveforms and voltage profiles), outage data. Turning this data into useful, actionable information will be no small feat.
There are approaches that can help manage this data deluge, though. Here’s a few thoughts on how to manage this data, apart from creating massive databases, data centers, and communication systems:
- Use data where you need it. Instead of building communication and data storage systems to transmit and remotely store data unnecessarily, you can use intelligent devices embedded in the grid to interpret and make sense of data generated out on the grid. These intelligent devices then use this data to respond to system disturbances and optimize grid performances in real time. This approach won’t eliminate the need to communicate data all the way to back-office systems. But by using and storing data locally, utilities can realize a measurable reduction in remote data storage and communication requirements — while simultaneously improving performance of the grid.
- Communicate only data that needs to be communicated, when you need it. A “network of networks” — different communication systems working together to support different smart grid applications — can also help better manage data. Meter data may only require readings every 15 minutes to support billing and dynamic pricing programs. But other applications, like fault detection, isolation, and restoration, will require a response in seconds. By building layered communication systems to support different application requirements, you’ll also help ensure data is managed effectively so smart grid applications function properly, while also minimizing costs. After all, it wouldn’t make financial sense to provide the same high-speed communications for meter readings as are required for DA.
- Make data easy to obtain through interoperability. Interoperability is an overriding priority for smart grid operators, and for good reason. Different systems need to work together to best manage data, to avoid redundancy in data capture and ensure that data can be easily used by smart grid applications as well as grid operators. At S&C, for instance, we’ve been working with Alstom Grid to jointly develop an integrated layered intelligence solution where information captured by S&C’s IntelliTeam® SG can be accessed through Alstom Grid’s IDMS. This approach provides a simplified user interface for control room operators so they can more easily access information they need to improve grid performance.
- Use appropriate communication protocols. To facilitate data sharing and management, use proven protocols designed for the task in hand. For example, a self-describing protocol designed for use within a substation (with fast Ethernet-based communications) isn’t going to be the best choice out on the feeder for many years. Appropriate Cybersecurity measures must first be put in place, along with communication systems with fast-enough latency—and we’re talking 4ms or less. It’s hard enough to get speeds of 100 ms, let alone 4 ms. There are communications solutions for communications that are fast enough, but they are few and far between. For most utilities, it makes sense to go with proven communication protocols, and then move onto other aspects of the data management puzzle.
Discussions on how to manage smart grid data are at early stages. How do you think we can best manage the data generated by smart grid devices?