Overview of how the Poseidon Systems' sensor suite digitizes oil and machine health though an innovative and holistic approach. This next generation digitalization technology will revolutionize rail, mining, energy, wind, marine, etc., bringing asset operational cost down significantly.
24. Wear Rates – Why Online?
• Wear metal generation is not a continuous,
predictable process
• It is typically observed in bursts
• Samples drawn minutes can vary by orders of
magnitude
• Impractical to analyze frequently enough to
characterize gearbox health
• Online, inductive coil sensors are not skewed by:
• Presence of nonmetallic particles
• Oil changes
• Filter media changes
Wear metal concentrations vary dramatically based
on operating conditions. Accurate conclusions
cannot be drawn from periodic sampling.
25. Lab Samples & ISO Codes Don’t Predict Health
Viewing particle count data by ISO code or iron concentrations
from lab samples is unclear
• Case Study Across 140 Assets:
• 70ppm Alarm
• 0.8% false alarms (1/119)
• 100% missed detection (0/18)
• 100% nuisance (1/1)
• 50ppm Warning
• 13% false alarms (16/119)
• 72% missed detections (13/18)
• 77% nuisance (17/22)
• Lab recommended limits of 25/14/11
• Exceeded on every gearbox for >6 and >14µm
Little to no value for taking actionable insights or measuring
health
27. Replaced
DeratedFault identified via
online monitoring
Offline Sampling
Online vs. Offline Example
Offline sampling unreliable for capturing fault events
Case Study – Oil samples did not indicate signs of fault, online sensor alerted years earlier
Normal iron and ISO values
ISO codes decrease
ISO increase, but still within “norms”
42. This technology provided me with nearly
a million dollars in warranty claims in
the first 3 months of operation.
John McKay
Operations Team Leader – AES