THE CHALLENGE

This global refinery had suffered chronic premature filter clogging, resulting in costly and frequent plant shutdowns.  The problem was unable to be solved in the 20 years the plant had been operating.  The plant process has numerous interdependent processing areas and complicated chemical reactions, making analysis of the problem particularly complex.

VROC processed 12 months worth of data from the 170,000 sensors 44.4 billion data points within hours.

THE VROC SOLUTION 

The analytics for this customer was performed on an offline historical data set. The client was interested in determining the root cause of the failure amongst the 170,000+ sensors on the plant. 

VROC ingested the historical data and provided insightful results as to the major factors responsible for failures. In addition, VROC was able to identify transient factors such as one month where the chemical supplier to the plant supplied incorrect chemicals. VROC correctly identified that the chemical was the significant factor causing failures for that particular month. 

VROC automatically identified the root causes of the issue which included equipment and processes in other areas of the plant. This advice informed engineers how to eliminate the issue.
 
Greater than $1.2m per year expected benefits in reduction of maintenance costs and downtime.

THE RESULTS

  • Rapid identification of the root causes of the problem informed engineers how to eliminate the issue
  • Greater than $1.2m per year expected benefits in reduction of maintenance costs and downtime
  • Increased site-based resource utilisation by re-allocating limited engineering resources to perform corrective maintenance tasks as oppose to collecting, analysing and modelling equipment data for failure and root cause identification