Xformics Verticals
Energy

Energy

Problem:

Early anomaly detection and failure prediction in Gas & Steam Turbines with Power Plants

Problem:

Power plant turbomachinery such as Gas and Steam Turbine are complex equipments that have hundreds of thousands of parts and each can failure in many different ways. For eg., compressors, blades and rotor can have multiple failures modes due to numerous reasons. Though traditional physics based approaches can diagnose failures, not all failure modes can be predicted early in advance to prevent unplanned outages or catastrophic failures.

logo

Benefit:

Catastrophic failures can be prevented that may lead to serious/irrepairable equipment damage. Multi-million dollar business productivity loss can be avoided Environmental pollution that could result as of equipment malfunction/damage can be averted or significantly reduced. In certain cases human lives can be saved from accidents due to explosions or contact with dangerous components. Systematic analysis and early prediction of failure leads to effective management of the asset and hence the risk pertaining to insurance and finance of operation can also be accordingly well managed.

Solution:

Various machine learning techniques can be used along with physics models to predict failures. Especially, Deep Learning is a specialized area of Machine Learning that shows promise in detecting anomalies earlier ahead of a forced outage event thereby preventing several mechanical and financial damage to the business. Deep Learning techniques rely on complex networks of Neural networks that can uncover hidden relationship between parameters that significantly contribute to anomalies. These techniques can be developed generically such that they can be customized to target multiple failure modes with minimal effort along with domain expert input.

For our comprehensive cyber security services

Get appointment