Predictive Maintenance for Process Plants
Process plants rely on vast numbers of expensive and crucial pieces of equipment that operate throughout the plant. Replacing them is both costly and time-consuming, and often requires shutting down the plant, leading to lost production time and delays. Getting surprised by a sudden failure can lead to even greater losses, since the root cause must be investigated and the replacement part or machine ordered and then installed, all while the clock is ticking.
PREDICTIVE MAINTENANCE GIVES EARLY WARNING OF FAILURE
Predicting when a part or machine needs to be maintained or replaced allows process engineers to plan more effectively and keep the plant running smoothly.
SAM GUARD Predictive Maintenance uses human-enhanced machine learning to track the data from your plant, to predict wear and tear on your equipment (static & rotating) and parts that can lead to failures.
By alerting you early to parts or machines that require maintenance or repair, maintenance engineers can streamline your maintenance schedule for greater savings and efficiency, and improve your plant’s uptime.
- Carry out repairs faster and at lower costs when you detect them early
- Make better budget decisions with improved data
- Save money by preventing last-minute, unscheduled shutdowns
In a phosphate processing plant, SAM GUARD detected changes in the vibrations in pump number 3 while spotting that the power of the pump remained stable, and generated an alert with both pieces of information.
Because flow/RPM remained constant while vibrations increased and became more unstable, it was clear that the problem wasn’t related to higher performance by the pump. The problem was verified using an external vibration measurement device.
When the reliability engineer investigated, it was found that the pump showed signs of imbalance, so the engineer suspected material crystallization on the impeller. They decided to stop the pump for cleaning, thereby preventing critical damage to the equipment and the pump’s concrete structure.