The Remaining Useful Life (RUL) determination for compressors NIA project (NIA_NGGT0079) “phase 1” concluded in 2016 and made an initial investigation into whether prognostic techniques, used to determine the Remaining Useful Life (RUL) of assets in other industries, could be applied to a gas transmission compressor fleet for the first in the UK.   The project showed some positive results and significant potential for the underlying approach in compressor unit failure modelling.

However, further work, including physical examination and confirmation of manifestation of predicted degradations in at least one compressor unit, is required to develop a reliable and proven model that can be used to improve our understanding of compressor turbine asset deterioration enough to inform future investment planning and targeting, including maintenance and replacement policies for asset management. The phase 1 project suggested a number of ways to proceed and after internal review of the outcomes this “phase 2” successor project has been developed with the original project partner to drive the further development required to achieve this goal.

This project is also distinct and complementary to the recently launched Compressor Data Analytics (CDA) NIA project (NIA_NGGT0079) that leverages some of the same data sets and analytical systems.  While the CDA project is focused on accurate prediction of short term “non-fatal” failures, such as failures to start and running trips on “day or hour ahead” timescales, this project is focussed on accurate prediction of end of life failures that could lead to permanent or prohibitively-expensive-to-repair failures.  The two projects will therefore take different approaches to solve different problems around compressor unit failure prediction.

Objectives

To develop, refine and prove a prognostic Remaining Useful Life modelling approach for gas generators/turbines.  To develop a successful fleet-specific approach for unrecoverable failure prediction and operational response to the mitigation of this risk.