Electrify Europe 2018
Tweet
Sessions
Speakers
Exhibitors
Products
Press Releases
Floor Plan
Buzz
Increased Reparability in the Non-Initial Set of Critical Gas Turbine Components: Analysis of Operation Experience Data for Asset Optimization
(
Room
Lehar 3
)
20 Jun 18
4:00 PM
-
5:30 PM
Tracks:
Track 3: Big Data, Big Opportunity.
Speaker(s):
Daniel Dagnelund, Data Scientist at Service R&D, Siemens Industrial Turbomachinery AB;
Davood Naderi, Product Manager at Service R&D, Siemens Industrial Turbomachinery AB;
Linnea Johansson, Data Scientist at Service R&D, Siemens Industrial Turbomachinery AB
Regular maintenance of a gas turbine is required to maintain the quality and reliability of the machine, but it is beneficial to keep the durations between the services to a maximum; as it maximizes the operation of the turbine. Siemens Industrial Turbomachinery (SIT) is attempting to depict the expected lifetime of critical gas turbine components as part of developing data driven service strategies. The objective of this work is to test the hypothesis that the reparability of the initially installed set of critical mechanical components is lower compared to that of the subsequent sets, even using the same component design and operation profile. This is done by analysis of operation experience data from over 50 sets of components sent to repair from the customers on the field. Repair procedures include a thorough assessment of the component status. The un-reparability of the component is categorized into one of several independent failure modes e.g. erosion and cracks. The relevant experience data available at SIT was divided into two datasets; initial and non-initial sets of components. Careful analysis of the distributions of the number of scrapped components (scrap rate) and the failure modes show that the scrap rate distributions vary significantly between the two data sets, giving a cautious nod to the hypothesis. Moreover, the “initial set” parameter is shown to have a large explanatory value for sets exhibiting high scrap rates. Further analysis also reveals which failure modes are responsible for the observed difference. Discussion of the possible origins for the observed differences and the potential implication for the operation optimization is included.
Home
Sessions
Speakers
Exhibitors
Products
Press Releases
Floor Plan
Tweet
Login
|
Back to top
ChirpE