General Electric announced that the data-driven insights drawn from its Monitoring & Diagnostics (M&D) Center translated to customer savings estimated at $70 million in 2014, up from $53.9 million in 2013.
Every day at center, the company collects more than 30,000 operating hours of data from a fleet of more than 1,500 gas turbine and generator assets, supplementing a 40-terabyte database representing more than 100 million fleet operating hours.
The company's M&D team has developed dozens of physics-based, proprietary algorithms that provide early warning of more than 150 potential failure mechanisms.
The Atlanta-based facility features a team of more than 50 engineers who analyze over 35,000 operational alarms per year.
Among the activities monitored at the center are the inlet temperature of a compressor, the thermal performance of a gas turbine, the temperature of combustion exhaust, dynamic tones of the combustion system, vibration levels of a rotor and the temperature of bearings.
"These capabilities enable power plant and fleet operators to tap into robust data analytics that, when blended with domain expertise, is transforming preventative maintenance processes into predictive ones," Dick Ayres, general manager of software solutions for GE's Power Generation Service business, told eWEEK. "The ability to foresee and forestall issues is at the very heart of predictive maintenance. This approach of seeing around corners is translating to significant benefits across the power generation industry, including less operational downtime, lower maintenance costs and longer asset life."
In addition to GE units, this remote monitoring can be applied across a customer's entire fleet.
Through device-agnostic predictive solutions, the M&D Center monitors technology and equipment not only from GE, but also Nooter/Eriksen, Flowserve, Emerson, Delaval, Byron Jackson and others.
Additional regional support with these services is provided to GE customers around the globe from other M&D locations in Scotland, France, India and its newest centers in Dubai, Saudi Arabia, and China.
"The key to effectively analyzing large volumes of operational data is breaking down silos that may exist across different data sources to be able to see across the entire spectrum of information," Ayres said. "Moreover, the data needs to be integrated with several other key elements—infrastructure, physics-based analytics, software, domain experts and mobility—to provide enterprisewide visibility that drives better decision making as it relates to optimizing operational processes and business priorities."
GE created a new enterprise software platform called Predix, a secure cloud environment that provides the foundation for all these elements needed to harness the more than 100 million hours of operational data the company has from the world's largest gas turbine fleet.
"Predix provides the complex scale, volume, velocity and security required for industrial-sized applications. We are migrating existing software and data applications, which we call Predictivity solutions, to this platform, as well as developing new offerings to expand our software portfolio," Ayres said. "Our vision for this portfolio is to collaborate with our customers to assemble the solutions they need to achieve their desired outcomes."