Vibration Alarm & Trip Setpoints
From Diagnosis to Action Thresholds
Modules 4.1 through 4.5 covered how to measure and diagnose vibration. This module covers how those measurements translate into actual operational response — the tiered setpoint structure that determines when a reading calls for investigation and when it calls for an automatic protective trip.
The Tiered Setpoint Structure
The normal band represents the expected vibration range for a healthy machine, typically established from baseline data collected when the machine was known to be in good condition — "normal" is specific to each individual bearing location on each individual machine, not a single generic industry value.
The alert level sits above the normal band and triggers investigation and increased monitoring attention, without necessarily requiring an immediate shutdown — it's a "look into this" signal. Reaching alert level is exactly when the diagnostic tools from this whole track — frequency analysis, phase angle, orbit plots — get applied to determine what's happening and whether the trend is stabilizing, worsening, or approaching genuine danger.
The danger (trip) level is a threshold that, when exceeded, triggers an automatic protective action through the independent protection system (Module 3.4) — typically a unit trip — without waiting for operator judgment or diagnosis. This connects directly to Module 3.4's protection system independence principle: vibration danger trips are handled by the same fail-safe, independent protection logic as overspeed or low lube oil pressure trips, not by the control system's normal judgment-based response.
Alert-level responses involve operator judgment and diagnosis. Danger-level trips are automatic, fail-safe protective actions independent of that judgment — the same control/protection split established in Module 3.4, now applied specifically to vibration.
Not-to-Exceed and Where Trip Setpoints Come From
The not-to-exceed (NTE) value is a manufacturer-specified absolute limit representing the vibration level beyond which mechanical damage risk becomes unacceptable. NTE is typically used as the basis for setting the danger/trip level with some margin, rather than being the trip setpoint itself — the trip is deliberately set below NTE to stop the machine before it ever actually reaches the manufacturer's absolute limit.
Setpoint Logic — Beyond a Simple Threshold
Real protection schemes often use more than a single absolute level. Voting logic (such as 2-out-of-3 agreement among multiple sensors) protects against a spurious trip caused by one failed or malfunctioning sensor, while still ensuring genuine, multi-sensor-confirmed dangerous vibration reliably trips the unit. A rate-of-change alarm triggers on a rapid vibration increase over a short time period, even if the absolute level hasn't yet reached the standard alert threshold — a rapidly worsening trend can be more concerning than a stable, elevated-but-steady reading, since it suggests an actively developing problem.
Beyond real-time alarming, baseline shift tracking — trending vibration data over weeks and months — catches gradual condition changes that might never individually cross an alert threshold. This directly parallels Module 1.4's heat rate trending discussion: a slow-developing mechanical change may never trip a discrete alarm, and only trending catches it, which is why long-term vibration trend review is standard practice, not just real-time alarm response.