Abstract
This report describes testing of selected use cases for future improved asset management at Statnett and
concludes the work in WP5 of the SAMBA project. The selection of use cases based on input from Statnett'
s experts in workshops, available data at Statnett today, and available models/methods from industry and
research partners. Testing of the selected use cases has been carried out by ABB, GE Grid Solution and
SINTEF Energy Research in cooperation with Statnett. The tested use cases were developed in WP2 of the
project, as documented in a separate report.
The testing consists of developing algorithms/codes to carry out the functions described in the use cases,
and then applying this to real data from Statnett. For some of the use cases the models to be used for testing
were readily available for testing, i.e. asset health indexes for transformers. The testing highlights possibilities
and benefits of implementing the use cases in Statnett, as well as identified limitations and problems. For
some use cases, poor quality of the available data has been identified and/or desired data is not available
today.
Use cases are not static descriptions and must be updated and further developed after the SAMBA-project.
The data input sources and accessibility will probably change as Statnett makes decisions on installing
sensors, new ICT systems and architectures. For the use case testing, data was provided by Statnett in Excel
spreadsheets. In the future, data exchange is expected to change as ICT architectures are built with exchange
of data in mind. In addition, some companies are establishing test platforms for external actors to perform
testing instead of transferring data. There are important decisions to be made when implementing the use
cases from SAMBA in Statnetts operations:
• Which use cases to prioritize for implementation?
• How much resources should be utilized to find, and quality assure historical data?
• Which of the suggested new measurements should be performed and for how many components?
• Who in Statnett should be responsible for each use case? How will the results will be implemented?