Turbine Vibration Monitoring & CMMS Integration: Power Plant Guide

OxMaint integrates with Bently Nevada 3500 systems through the OSIsoft PI or Aspen IP.21 historian layer that most power plants already use to aggregate data from the 3500 rack. The integration path is: Bently Nevada 3500 rack outputs alert relay states and channel data to the historian via Modbus TCP or the dedicated BN Data Manager software. The historian then triggers OxMaint via PI Notifications, PI Event Frames, or a scheduled REST API call when vibration tags exceed configured thresholds. Direct Bently Nevada to OxMaint connection without a historian layer is also possible using the 3500's built-in Modbus TCP interface and OxMaint's API endpoint. Book a demo to walk through your specific Bently Nevada configuration.

False alarm suppression is handled through three OxMaint configuration settings. First, a confirmation delay: the system only generates a work order if the threshold exceedance persists for a configurable duration (typically 2–5 minutes for Alert tier, immediate for Danger and Trip). This eliminates transient spikes from startup events, load changes, and electrical interference. Second, a speed-gated enable: many vibration alert criteria are only valid above minimum governing speed — OxMaint's trigger rules can be conditioned on the unit being above a defined speed tag value. Third, a duplicate suppression window: if an alert fires and a WO is already open for the same asset and parameter, no second WO is created — instead, the new alert data is attached to the existing open work order.

The data attached to an OxMaint vibration-triggered work order depends on what your monitoring system can export at the time of trigger. At minimum, the work order receives: the alert timestamp, the channel name and asset tag, the measured value at time of alert, the threshold tier that was exceeded, and the configured limit value. If your historian or monitoring system supports it, OxMaint can also attach: a 30-second trend plot exported as PNG from the historian, a spectrum plot from the vibration analyzer, and a raw data CSV of the last 60 minutes of the affected channel. The goal is that the technician who accepts the work order has enough context to make an initial assessment before physically reaching the equipment.

OxMaint's primary function is maintenance execution and work order management — not process data historian. For continuous vibration trending at high sample rates (100 Hz and above), a dedicated historian like OSIsoft PI or the Bently Nevada System 1 software remains the appropriate tool. OxMaint complements the historian: it receives structured alert events from the historian, generates and manages the maintenance response, and stores inspection findings and corrective action records against the asset history. For plants without a full historian, OxMaint supports manual vibration reading entry on a scheduled basis — a technician logs vibration readings directly in OxMaint during rounds, and the system tracks trends and generates work orders when manually-entered readings exceed configured limits.

Vibration-triggered work orders in OxMaint are created as corrective or condition-based maintenance work orders rather than PM work orders. They carry a higher default priority than scheduled PMs, appear on the technician's mobile queue immediately regardless of shift schedule, and include the sensor data attachment as context. They are linked to the triggering asset's PM history — so if a vibration alert fires 3 weeks before the next scheduled bearing inspection, the technician can review the PM work order history while responding to the alert. If the corrective WO finds a defect that requires rescheduling the next PM, the PM interval can be adjusted directly from the corrective WO screen.

For a plant with an existing OSIsoft PI historian and Bently Nevada 3500 monitoring system, the integration implementation follows this timeline: asset tag mapping and threshold configuration in OxMaint takes 1–2 days per turbine unit. PI Notification or Event Frame configuration by the DCS team takes 1 day per monitoring system. Work order template creation for each fault response type takes 1 day. End-to-end pipeline testing takes half a day. Total elapsed time from project start to live integration: 5–10 working days for a single CCGT unit. Multi-unit and multi-site rollouts scale linearly. OxMaint's integration support team provides configuration assistance at no additional cost. Start a free trial to begin the integration configuration.