Duke Energy, the largest power generation company in the United States, owns 52,700 megawatts of electric generating capacity. Like others in the industry, Duke Energy struggles with manually monitoring machine conditions. And supplying roughly 7.4 million customers across six states means they have a lot of monitoring to do!
The Problem: Too Much Data, Not Enough Time
Condition monitoring is essential to control costs and avoid outages from equipment failures, and manual data collection (also known as route-based measurements) is very labor intensive. At Duke Energy, predictive maintenance specialists physically walked to each collection point, gathered hundreds of data samples manually, and then returned to their computers to view and analyze their collected data.
With almost 60,000 collections a month, analysts were typically spending 80 percent of their time collecting the data and only 20 percent analyzing them. The result was leading to inconsistent diagnosis and limited risk assessment—and a lot of walking!
Duke Energy needed technology that could identify problems and notify specialists, letting them focus on higher value tasks and do their jobs regardless of location.
The Solution: A Platform-Based Approach
Duke Energy worked with us, as well as the Electric Power Research Institute (EPRI), InStep Software (now part of Schneider Electric), and OSIsoft to create custom monitoring and diagnostic infrastructure to be implemented across its fleet and plants.
In 2012, they began building a new architecture to support this project, that incorporated our CompactRIO platform. Maintenance sensors connect to CompactRIO monitoring systems, which perform signal collection and processing that’s transmitted to its plant servers. The CompactRIO systems screen the large amount of analog sensor data continuously looking for trigger conditions that, when met, are sent as full waveform captures to Duke Energy’s data specialists for analysis.
Duke Energy also used NI InsightCM™ software for condition monitoring to make the data more user friendly, and more broadly accessible beyond the small technical team.
By connecting an FPGA and an onboard real-time processor to the sensor, raw analog waveforms are translated into conditions indicating the “health” of the system at the node itself.
this “prevents the data overload condition in which subject matter experts are stuck looking for problems that are difficult to locate.”
The Results: More Uptime, Less Cost
As of March 2017, nearly 2,000 CompactRIO systems have been deployed and are managed by Duke Energy’s custom Smart Monitoring and Diagnostics architecture across 30 facilities.
Within these plants, Duke Energy relies on automated data collection, allowing analysts to spend 80 percent of their time on analysis, rather than data collection. Over the course of four years, Duke Energy has avoided 130 percent of costs by avoiding the higher price associated with failures.
In the future, Duke Energy wants to save more by using more wireless sensors and hopes to implement tools that can diagnose problems upfront. The ultimate goal? A predictive maintenance solution that identifies the problem and provides recommendations on how to resolve it.