Based on historical data, Sulzer identified pre-failure conditions which occurred weeks in advance. The customer requested Sulzer to deploy BLUE BOX on all 15 identical pumps on this pipeline to pro-actively identify the early warning signs and prevent any further loss in production.

The availability of oil pipeline pumps is critical to the performance of the business. Investing in new equipment should bring long-term peace of mind. However, in this case, shortly after installation, the pump began experiencing high axial acceleration that is not well identifiable with direct instrument readings. Ultimately, as this vibration increased, a resulting seal failure restricted productivity of the line.

BLUE BOX was used to interrogate and analyze the raw instrument output from the data historian for the six months preceding the failure.

BLUE BOX uses machine learning, in-depth analysis and algorithms backed-up with expert human intelligence and experience to filter out the noise and eliminate false red flags. A definitive alarm and recommended resolution could have been issued as
early as one month before the failure occurred, giving ample time to resolve the problem and avoid lost production.

BLUE BOX now monitors all the pipeline assets with no unexpected failure of pumps
since deployment.

Improved analysis of pump data can:

• Increase pump availability
• Extend asset reliability
• Reduce maintenance costs