Tailor vibration-data collection frequency to the failure interval and consequences of failure.
The interval must fit the failure progression: a 90-day route can miss a bearing defect that develops and causes failure within 90 days, whereas monthly readings can identify that condition before the failure event [S4].
Choose the Collection Interval from Failure Development
Management commonly assigns monthly, quarterly, or yearly routes according to perceived asset importance and available resources, but arbitrary intervals can leave failure modes undetected [S4]. A defensible interval is shorter than the time in which the relevant defect develops into pump failure [S4].
For example, a bearing may develop a failure on day one and run for ninety days before causing the machine to fail. Collection every 90 days can miss the development window, while collection every 30 days can expose the defect while repair work can still be planned and budgeted [S4]. This is a failure-progression example, not a universal bearing-life limit.
Criticality still affects the route design. Management may increase vibration collection beyond monthly after a recent machine failure, but higher frequency should follow demonstrated degradation or consequence of failure rather than serve as an unsupported plant-wide rule [S4].
Route Monitoring, Event Capture, and Continuous Systems
Portable collection is the baseline for many pump programs: a technician uses a handheld sensor, usually an accelerometer, at predefined labeled points and uploads the readings for analysis. The route provides detailed snapshots but is labor-intensive and represents only the conditions during each visit [S1].
Portable analyzers commonly record a few seconds of data, enough to plot vibration across several pump-shaft rotations and resolve the frequency content associated with the source of vibration energy [S5]. A spot reading with a vibration meter is faster, while a vibration analyzer supports frequency, phase, time-domain, and trend assessment.
Continuous systems use permanently installed sensors to measure and log vibration continuously and can alarm when thresholds are exceeded [S5]. Wireless sensors may also transmit vibration and temperature data to a central gateway, local server, or cloud platform, providing 24/7 coverage and capturing transient events that periodic checks can miss [S1].
Criticality Is Not the Only Decision Variable
Continuous monitoring is strongest for critical assets with rapid-failure modes, severe downtime consequences, variable-speed operation, or transient excitation [S2]. Route monitoring remains practical for lower-consequence assets when the route interval is demonstrably shorter than the progression time of the failure modes being managed.
Vertical turbine pumps require an additional structural check because their cantilevered columns, high motor mass, foundation grout, bolted joints, and fluid-induced forces can lower structural natural frequencies near running speed. Their Reed Critical Frequency should be established by a field bump test, not inferred only from a CAD model [S3].
A monthly measurement can reveal a persistent condition, but it may miss the peak response during a specific speed or operating state.
Compare Collection Methods by Diagnostic Coverage
The collection method should be compared against failure progression, event capture, labor burden, and structural sensitivity. Monthly portable routes offer repeatable directional readings and frequency analysis with limited labor, whereas continuous systems cover every operating period and transient events at the cost of permanent sensors, communications, alarm management, and data infrastructure [S1][S2].
Decision criteria are straightforward. Use route-based vibration data collection when the inspection interval is shorter than the expected degradation window and operating conditions remain stable enough for trend comparison. Use continuous monitoring when the pump can fail within days, when a transient can cause consequential damage, or when resonance risk depends on changing speed. Use event-triggered review when process conditions are known but full continuous acquisition is unnecessary.
Equipment classes also affect the required interpretation. Horizontal pumps commonly show internal mechanical faults such as unbalance and bearing defects, while vertical turbine pumps can respond strongly to installation stiffness and mode shape [S3]. The measurement interval does not remove the need for the correct diagnostic method or asset-specific baseline.
Set Baselines, Alarms, and Review Intervals
Spot readings should be compared with published acceptable amplitude tables, comparable equipment, and the pump's established baseline. Manual plotting can reveal abnormal vibration levels and a likely vibration profile that identifies the probable source of the energy [S5].
Continuous installations need defined alarm thresholds, escalation paths, and sensor verification. Alarm thresholds should reflect the machine and installation, not a generic level copied across the plant. A condition-monitoring system can integrate permanently installed sensors with plant controls and operator alerts, but continuous logging adds value only when engineers review the resulting data and act on it [S5].
The condition monitoring system should also record operating context such as speed, load, suction conditions, and maintenance events. Without that context, a change in vibration may be attributed to mechanical damage even when it reflects a process change. The vibration sensor is therefore only one element of the route, diagnostic, and response process.
Apply a Risk-Based Collection Schedule
Use monthly portable vibration collection as the defensible starting point for critical pumps when no failure-history evidence supports a shorter interval. Shorten the route to capture defects that progress rapidly, and deploy continuous monitoring for fast-failure, high-consequence, transient-sensitive, or resonance-sensitive service [S2][S4].
Verify the schedule through two trackable signals: the proportion of routes completed at the assigned interval and the number of developing faults detected before unplanned failure. For vertical turbine pumps, also record operating speed relative to the measured Reed Critical Frequency and the ±15% avoidance band [S3].