Capturing the winds energy potential
A drive across the nation’s landscapes is revealing more wind farms cropping up on the horizons. As of last year, 37 states are home to at least one wind farm, and the more than 40 gigawatts of installed capacity accounted for about 3 percent of the electricity generation in the US.
Sandia is working closely with partners in the windpower industry to to improve the viability and reliability of wind as a major alternative energy source. (Photo by Randy Montoya)
A 2008 DOE report, however, points to much greater potential and suggests that by 2030, 20 percent of the nation’s energy needs could be supplied by wind turbines. While the idea of harnessing the wind’s energy is ancient, the global scale of the industry only began to be realized in the last five years. With this realization, the growing need for highly reliable wind turbines becomes paramount, but little data exists to point out opportunities and areas for improvement.
“Wind energy is leading our nation’s and the world’s clean energy movement. As we become more dependent on these energy sources, we must make sure that we are installing the most effective, viable, and reliable systems possible to transform the energy picture of the future,” says Jose Zayas, (6120) senior manager for Renewable Energy Technologies.
To address this need for data, Sandia is completing the development stage to create a Continuous Reliability Enhancement database for Wind (CREW). This database will be the foundation for analyses to identify primary failures and associated improvement opportunities, enable reduced operating and maintenance costs, and provide industry benchmarks. The DOE-sponsored project focuses on the nation’s utility-scale turbines of one megawatt and higher.
Data in 36 key operating parameters
“This project is the first effort to compile a comprehensive dataset that reflects the performance of the US wind fleet. With better understanding of current performance of the major turbine systems, wind operators can direct their efforts toward improvements in those areas that will drive increased reliability and efficiency,” says team lead Bridget McKenney (6121).
By tapping into turbines’ existing supervisory control and data acquisition (SCADA) industrial control systems, Sandia researchers are collecting information on 36 key operating parameters such as wind speed, blade angles, component temperatures, and torques. Every two seconds, a wind turbine’s SCADA system captures a picture of how the turbine and its components are performing relative to a defined operating envelope and its environment. Currently, four wind plant owner/operators are participating in the development phase of the CREW project by providing this SCADA data to Sandia’s CREW database via automated data collection software developed by Strategic Power Systems (SPS). SPS is a key partner with many years of experience in collecting high-volume data from steam and gas turbines via a proprietary software tool. SPS has converted this tool to collect data from wind turbines.
“Our assignment from DOE is to characterize the national fleet. We’re not looking at one technology, one location, or one company,” says Alistair Ogilvie, CREW database lead (6121). “We want to look at the entire US fleet and create baseline statistics for the industry to be able to say, ‘This is what you and your competitors should be trying to achieve.’”
To reach a statistically representative baseline, CREW will aggregate data received from all participating wind plants. The current data set represents approximately 2 percent of the nation’s wind turbines, but as the project grows, CREW researchers expect to include approximately 20 percent of turbines to establish representative benchmarks.
The volume of data to sift through is mind-bending. The print collections of the Library of Congress are roughly 10 terabytes; CREW’s data set, with dozens of variables taken at two-second intervals from a fifth of the nation’s wind turbines, is expected to dwarf that. Over the past six months, four pilot plants with a combined 345 turbines, or about 2 percent of the nation’s installed turbines, have generated two terabytes of raw data. This data has been provided to the CREW database by the SPS collection tool. To process this enormous dataset into a usable database that can readily support a wide variety of queries, CREW turned to Sandia’s Enterprise Database Administration team (9538).
Identifying likely failure points
“The goal is not to have the biggest database — the goal is to transform it into a useful dataset for the analyst,” says Michael Mink (9538). “We’re taking the raw data on weather, wind speed, angles of the blades, and so on, at two-second intervals, taking time chunks of that, summarizing it and putting it into another database that people can query easily and quickly.”
Determining which components are most likely to fail is an important part of benchmarking. Major turbine systems include a set of three blades, rotor, shaft, generator and gearbox, any of which have the potential to fail. Turbines that are down for maintenance or repair are expensive — in addition to lost productivity, the cost of hiring a crane for repairs can be upwards of $250,000, and because there are only a few cranes in the nation large enough to handle turbine heights and component weights, an operator might wait for months before the turbine is up and running again.
The CREW team will use the dataset to identify the top turbine systems and components responsible for the majority of downtime, which will then inform specific research and technology improvements. “If we can identify those components most at risk, we can provide the industry and DOE with information on which ones need further research and where funding should go,” says Alistair. Components that were once considered the most vulnerable become more reliable, allowing the industry to move on to addressing the next challenge.
“We’re excited about the results so far, and look forward to the next few years, as we make an important contribution to our industry to improve the reliability through a component-level focus,” Bridget says. “It’s an important project for the industry and the nation as a whole, and we could not have been successful without the outstanding partnership of Corporate Computing and SPS, and the support and leadership from DOE. Together, we can share our expertise to help shape the future of the nation’s wind energy generation.”
Dealing with 10,000 truckloads of data
The Continuous Reliability Enhancement for Wind (CREW) program is wrapping up its pilot phase, and has already generated 68 billion rows of data. Over its lifetime, the project is expected to produce more than 10 terabytes of data, roughly the equivalent of 10,000 pickup trucks filled with books. Such a fantastic amount of information would be nearly impossible to wade through, but the CREW team turned to Sandia’s Enterprise Database Administration team (9538) to turn it into something usable. The team has taken the 68 billion rows of data and transformed it into a structured database format to help CREW better use the information they’ve collected.
“Earlier this year, we formed a five-person Process Innovation Team focused on supporting customers working with a terabyte of data or more,” says Cynthia Huber (9538), manager of Enterprise Database Administration (EDA). “We have built new and innovative processes, designed to handle big data loads and transformations, on top of our already tried-and-true database support standards. CREW’s requirements for consuming big data fit well with our goal of enabling Sandia’s missions in managing large volumes of data, so we are enthusiastically supporting that effort.”
The techniques used by EDA are expected to translate to multiple other industries, such as solar, that have the potential to generate large quantities of data, but also need that data to be user-friendly. With the mission of providing valuable database administration services to Sandia’s applications development community that ensure the integrity, availability, recoverability, accessibility, integration, and security of Sandia’s corporate data, the EDA is continually working to meet Sandia’s database needs. Working with terabyte-scale projects is an evolving field, but the EDA is well-equipped to handle such challenges.
“We have a highly skilled group of database administrators with many years of multiplatform database management experience servicing both enterprise and mission level applications,” Cynthia says. “Additionally, our ability to scale up to handle these data volumes is made possible by the excellent support we receive from our enterprise IT partner Infrastructure Computing Services (9324), which provides the system and storage administration for the database servers that ultimately host this capability. They provide us the hardware footprint that enables the database administrator to perform the data integration and transformation necessary to deliver valuable and secure data to our customers. We’re experts in data integration, data transformations and securing data for customer use.” — Stephanie Hobby