Oil&Gas Industry

Predictive analysis of feed turbo pump condition

Projects

An energy holding company operating thermal power plants

Get the consultation

Predictive analysis of feed turbo pump condition
Проблема
Problem
  • Failure to early detect feed pump faults
  • High fines for power under-delivery and emergency repair of equipment in case of emergency shutdown
Tasks

Developing a predictive analysis model for the feed turbo pump condition

Adapting the solution to work with the power equipment maintenance and repair management system

Features of the project
  • Necessity to identify anomalies in equipment operation based on historical 3-year operation data
Features of the project
Solution

A system of risk-based maintenance and repair management based on the equipment actual state.

The solution implements
  • Collects, processes and stores telemetry data from equipment sensors
  • Ensures the engineering and mathematical model operation for units and assemblies
  • Assesses and forecasts the equipment "health index"
Learn more about the solution
How the solution works
  • Classifying of data samples on equipment operating conditions

  • Selecting data processing algorithm Base variable - feed water flow rate at the head of the feed turbo pump

  • Calculating the base variable value and comparing with the actual value, determining an "error" - the difference between them

  • Calculating a corridor of acceptable "error" values for five-minute intervals Detecting additional previously non-recorded anomalies

Эффект для бизнеса
Business effect
  • Early detection of hidden defects
  • Prevented equipment failures
  • Longer equipment life
  • Less fines and emergency repairs
Any questions? Ask our experts

Leave request and find out how predictive analytics and effective data management can help your business

By clicking "Send request" button, you agree to the Privacy Policy.