UAI News
Welcome to the UAI News Community! UAI will now be posting all of our member communications to this... View more
Organizers:
- Organized by
-
-
Transformer Failure Prediction Model
-
Transformer Failure Prediction Model
How does your utility reduce outages caused by transformer failure? Members of Utility Analytics Institute (UAI), a utility-led membership organization supporting utility analytics professionals and utilizes of all shapes and sizes, are sharing their experiences and lessons learned with each other through community conversations and presentations.
Entergy has been exploring the opportunity to predict transformer failure using data from Advanced Meter Infrastructure (AMI) meters. This team developed an algorithm to detect voltage spikes above the nominal range for the transformer.
Bashar Kellow, Manager of Business Analytics at Entergy, and members of his team consisting of experts in business analysis, data engineering, data science, product management and solution architecture, presented aspects of the transformer failure prediction use case, “Predicting Asset Failure,” during the May 24,2023 meeting of the UAI Asset Health Analytics Community.
“I would like to publicly thank Yong Li from DTE Energy. Because of his openness to collaborate and share information, Entergy was able to save a ton of time on this project. Good example of collaboration between UAI members.” – Andy Quick, Vice President, Business Data and Insights at Entergy
With a major focus on customer satisfaction and utilizing the investment Entergy made in Smart Grid, this session demonstrated how the use of AMI meter data can help predict failure of distribution transformers, plan better for their replacement, and reduce outage minutes and overtime. During this session, Entergy demonstrated their Transformer Failure Prediction Tool, the business process, and the preliminary business value achieved since the release of this product.
This presentation exhibited the remarkable research and findings Entergy has conducted to proactively predict transformer failure and replace these transformers before the failure and resulting outage happens. The discussion recapped the team’s approach to the problem, the solution selected, and various aspects of the project as outlined below.
Problem
Transformer failures cause unplanned outages for their customers.
Solution
Developed an algorithm by using AMI meter data to measure the voltage on customer meters and detect voltage spikes above the transformer’s nominal range and a dashboard to visualize and drill into the findings.
Product Approach
· Transformer Failure Dashboard: Created dynamic visualization for identification and location of transformers.
· Prioritization: Selected the order in which the team would examine the transformers.
· Triggers: Determined the criteria to drive the decision to roll a truck to investigate and/or replace a transformer.
· Model Validation Efforts: The product team debated the pros and cons of various methodologies.
· Data Refresh Rate: The team implemented a schedule for refreshing the data within the model based on the frequency the data platform received AMI meter data.
· Data Management: Data engineers debated and resolved challenges associated with analyzing large volumes of voltage data.
Product Findings
· Time to failure.
· Results of investigations.
· Reviewed lessons learned in the field.
· Teamwork with field operations.
· Opportunities identified for future improvements.
Business Case & Business Process
· Identified customer satisfaction as the primary business driver.
· Developed a process to work hand in hand with the field crew.
· Discovered training opportunities in the field during installation and repair.
· Located and corrected inaccurate meter mapping.
· Collaborated with a UAI member utility to reduce time to market.
Product Roadmap & Benefits Anticipated
· Expansion to other operations.
· Product enhancements.
· Integrated reports.
· Enhancement to prediction model.
· Apply prediction failure model to other devices.
Categories of Use Cases in the Pipeline
· Asset Health
· Vegetation Management
· Capital Project Cost Management
· Safety
UAI offers exclusive monthly Community Conversations, providing members with 96 engaging sessions per year. Led by utility analytics professionals, these conversations focus on topics important to UAI members, advancing critical utility business processes and elevating members’ analytics capabilities.
UAI Asset Health Community members can access the “Predicting Asset Failure” presentation video directly by clicking on the link and scrolling down or through the UAI Asset Health Community Library where a downloadable version of the presentation is also available. UAI members who are not currently members of the Asset Health Community can easily request to join.
——————————
Sincerely,Sandi Joralemon
Sr. Research Analyst
Utility Analytics Institute
sjoralemon@utilityanalytics.com
830.832.5042
—————————— -
0 Replies
Bashar Kellow is Manager of Business Analytics at Entergy. He is leading the advanced analytics initiatives for the Power Delivery (T&D) organization. Bashar has been with Entergy for 9 years and held numerous roles in Distribution Operations and IT. Prior to Entergy, Bashar managed the portfolio of T&D technology projects at Pacific Gas and Electric.Sorry, there were no replies found.
Log in to reply.