Cloud Stories: Energy Forecasting
The University of Texas cloud helps forecast our future energy needs
By Paul Navratil, primary investigator, Texas Advanced Computing Center (TACC) at The University of Texas at Austin, Academic Research
With the second- and third-largest academically hosted super computers in the world, we have a lot of experience managing large amounts of data for research purposes. But what’s really exciting is what we can do with that data, how we can use it to develop real solutions to some very real challenges. Using touch screen visualization tools based on Intel® technology, we help people see their data, making it accessible to researchers and consumable to the public.
Take the Pecan Street smart grid demonstration project as an example, where we collect usage data for electricity, water, and gas, as well as solar energy generation data, from more than 200 homes, schools, small businesses, and commercial buildings in the Mueller community of Austin, Texas. A number of metering devices and sensors measure energy usage in the buildings and then stream that data to TACC.
Eventually this project will capture data from more devices, such as Intel® Atom™ processor-based smart energy hubs and advanced energy sensors, involve as many as 1,000 homes and businesses, and encompass more data points, such as circuit-by-circuit information, to understand which appliances and devices are most efficient and where home-based energy production (such as solar) is best utilized. The Intel® devices will also enable testing of software to let homeowners manage their electric vehicles and HVAC systems so they can optimize utility-rate programs and potentially save money and/or energy.
The Pecan Street project is expected to amass several terabytes of raw data this year—data that must be dissected and analyzed and served to a variety of constituencies. Our supercomputer, based on Intel® Xeon® processors and soon the Intel® Many Integrated Core Architecture (Intel® MIC Architecture) with our upcoming Stampede system, has the power to crunch the data on multiple variables, enabling researchers to ask the kinds of questions that can lead to real solutions. For example: Which refrigerators in this neighborhood are nearing the end of their energy-efficient life? Is it more effective to deploy solar panels on the south or west side of the building?
As the Pecan Street project evolves, our ability to visualize and share the data will hopefully make a real difference in energy policy and user habits in the future. For example, consumer product manufacturers could use this information to design equipment that’s more efficient, together with an interface that provides updates to the consumer smart phone devices. Homeowners might consider alternative energy sources or change their energy habits once they can see how much energy it takes to run their big-screen TV. We will also be able to share this data with policy makers so they can learn which energy sources are most effective and where there is the greatest need. One day soon, the data we are collecting from this project might help the mayor of Austin assess the energy needs of the whole city. Personally, I think that's really exciting.
"One day soon, the data we are collecting from this project might help the mayor of Austin assess the energy needs of the whole city."