Accurate prediction of past repair and maintenance history, which is not available through current existing methods
“We use guardione® solution for the management of our transformer’s conditions and we trust it to prevent unexpected failures.”
[Account’s Electrical Maintenance Engineer]
Account Overview
This account is a public corporation that was established through the acquisition of KEPCO’s cogeneration plants and KDHC’s branch offices, is responsible for electric power generation and 24-hour continuous supply of district heating.
Challenge
The client had been maintaining the transformers through insulating oil gas analysis and detailed inspections, but both were insufficient when it came to preventing unexpected failures. Considering that the transformers were the most important part of power plant operation, the client needed a new approach.
Project
Onepredict has been officially recognized for its predictive diagnosis technology and algorithm excellence by participating as a joint developer with Korea Electric Power Corporation.
Through a PoC, the client verified diagnostics and prediction accuracy, which are guardione®’s key features. The detailed procedure for this was as follows.
1. A blind test was conducted using the prior-to-repair gas analysis data of a previously repaired transformer and prior-to-failure data of a previously malfunctioning transformer.
2. The data had been diagnosed as “Normal” based on international standards such as the IEEE or IEC, which was the reason for unexpected malfunctions.
Result
Guardione® solution successfully proved its performance in the following ways. The authentication led to years of long-term contracts for all transformer units.
1. Guardione solution assessed the transformer given as the PoC task and determined its status as “Warning” for 1 year preceding the actual repair and also predicted its status 1 year later as “Warning”.
2. In addition, the solution not only provided a simple diagnostics forecast, but also the cause and area of malfunction with a “Warning” diagnosis. The loosening of the fastening part was suggested as the cause of the problem, and the solution accurately pinpointed the bolt fume phenomenon that had occurred in the past.
Accurate prediction of past repair and maintenance history, which is not available through current existing methods
“We use guardione® solution for the management of our transformer’s conditions and we trust it to prevent unexpected failures.”
[Account’s Electrical Maintenance Engineer]
Account Overview
This account is a public corporation that was established through the acquisition of KEPCO’s cogeneration plants and KDHC’s branch offices, is responsible for electric power generation and 24-hour continuous supply of district heating.
Challenge
The client had been maintaining the transformers through insulating oil gas analysis and detailed inspections, but both were insufficient when it came to preventing unexpected failures. Considering that the transformers were the most important part of power plant operation, the client needed a new approach.
Project
Onepredict has been officially recognized for its predictive diagnosis technology and algorithm excellence by participating as a joint developer with Korea Electric Power Corporation.
Through a PoC, the client verified diagnostics and prediction accuracy, which are guardione®’s key features. The detailed procedure for this was as follows.
1. A blind test was conducted using the prior-to-repair gas analysis data of a previously repaired transformer and prior-to-failure data of a previously malfunctioning transformer.
2. The data had been diagnosed as “Normal” based on international standards such as the IEEE or IEC, which was the reason for unexpected malfunctions.
Result
Guardione® solution successfully proved its performance in the following ways. The authentication led to years of long-term contracts for all transformer units.
1. Guardione solution assessed the transformer given as the PoC task and determined its status as “Warning” for 1 year preceding the actual repair and also predicted its status 1 year later as “Warning”.
2. In addition, the solution not only provided a simple diagnostics forecast, but also the cause and area of malfunction with a “Warning” diagnosis. The loosening of the fastening part was suggested as the cause of the problem, and the solution accurately pinpointed the bolt fume phenomenon that had occurred in the past.