Upload
others
View
6
Download
0
Embed Size (px)
Citation preview
Thomas Anthony, P.E.Industry Intelligence
ManagerOncor Electric Delivery Co.
USA
Thomas DamonAssistant VP–Smart Grid
Science Applications International Corporation
USA
Failure Detection and Predictive Analytics System
• Tom Anthony
• Industry intelligence manager with a U.S.-based electric utility
• Currently deploying 3.1 million smart meters
• Tom Damon
• System developer who uses power system engineering expertise to assist utilities in optimizing business processes by leveraging their technology investments
Project Background
• U.S. Department of Energy’s Modern Grid Initiative
• Knowledge transfer opportunities
• Leverage existing data sources to improve analytics and predictive capabilities
• Use the cloud to control data processing costs
Experience on Smart Grid Technology Demonstration and R&D Project
• Located in West Virginia
• Integration of distributed resources and advanced technologies
• Included fault location and prediction• Advanced fault location
• Neural network methodologies to identify pre-fault conditions and anticipate failures
Feeders and zones
The Utility Found the Effort Promising
• Continue enhancing its smart grid abilities
• Collaborate with a company that could develop needed algorithms and software for the fault detection and prediction system
• Analyze existing data from the utility’s broadband-over-powerline (BPL) system to automatically identify pattern interruptions and threshold breaches that might indicate incipient customer issues within the distribution system
A Two-Phase Project
1. Develop an initial “proof-of-concept” data analysis application and build a prototype
2. Develop a functioning data analysis tool and install it in the utility’s data center
Phase 1: Develop Prototype
Phase 2: Develop and Install Tool
Data Analysis Tool: Issues Screen
Data Analysis Tool: Work Request Screen
Data Analysis Tool: Details Screen
What the Future Holds
• Both the utility and the tool development company will continue to invest resources to develop the data analysis tool, part of an overall smart grid deployment program
• The utility plans to install the data analysis tool in the second quarter of 2011, after completion of its data center move
• Once the tool is installed and operational, the process of automating the recognition of actual issues on the BPL system will begin
• If the tool proves its value on the BPL system, it should easily be able to be adapted to other future remote data acquisition platforms
What We Have Learned
• The value of applying analytic algorithms to grid data analysis has been proven
• Predictive analytics is a core ingredient in future system integration programs
• Real-time analysis is necessary for high-value devices
• This effort increased the data sources to build a neural network capability to improve maintenance prescription and failure prediction