Bently Nevada 991-25-70-03-05: Building a Reliable PLC/DCS Axial Thrust Alarm System

The performance of the alarm system begins at the machine, not inside the PLC. A 3300 NSv non-contacting proximity probe is positioned opposite a suitable metallic target, typically a thrust collar or prepared shaft surface. The probe responds to changes in the physical gap without touching the rotating component, avoiding mechanical wear at the sensing point.

The probe, extension cable, and transmitter must be considered one coordinated transducer system. For the 991-25-70-03-05, the required electrical system length is 7.0 m. This length is formed by the probe lead and matching extension cable together. Choosing components only according to the physical cable route can create an incorrect combination and affect system linearity, scale factor, and calibration.

A strong mechanical mounting arrangement is equally important. The probe bracket should remain rigid during acceleration, normal operation, process transients, and thermal cycling. Movement of the bracket relative to the casing will be measured as though the shaft itself had moved. Loose hardware or inadequate mounting stiffness can consequently produce unstable trends and false alarms.

The target surface must offer sufficient area and suitable material properties for an eddy-current proximity measurement. Surface irregularities, metallurgical differences, residual magnetism, scratches, or excessive mechanical runout may influence the indicated position. These factors should be evaluated during installation and, where practical, during slow-roll testing.

The 991 operates as a two-wire, loop-powered instrument. It requires 12 to 35 Vdc at its terminals and delivers a 4-20 mA output representing the configured displacement range. Its typical loop accuracy is within +/-1.5% of full scale under the manufacturer's stated conditions. For most installations, this output can connect directly to a standard industrial analog input.

However, increasing current should not automatically be associated with a particular thrust direction. The relationship depends on which side of the thrust collar is observed and how the probe is mounted. A controlled field test must establish whether increasing current represents movement toward the driver, driven machine, active bearing face, or inactive bearing face.

The control system database converts the received current signal into engineering units. Regardless of the units adopted by the plant, consistency must be maintained across the operator interface, alarm setpoint lists, historical records, maintenance procedures, and trip logs. The interface should not merely display "axial displacement." It should identify the monitored bearing, machine end, and positive direction. This prevents confusion when operators compare measured values with mechanical drawings or discuss potential faults with reliability engineers.

Alarm thresholds must be set according to the specific characteristics of the machine. Relevant design parameters include internal axial clearance, thrust bearing configuration, seal location, rotor end play, normal process loads, allowable transient displacement, and measurement uncertainty. The full-scale range of the transmitter is an instrument characteristic and is not equivalent to the recommended alarm or shutdown setpoints.

Different setpoints may be required for the two directions. The available clearance when the rotor moves toward one end of the machine may differ from the clearance at the other end, and the consequences of displacement may also vary. The PLC or DCS should therefore support independent thresholds for high positive and high negative displacement.

Appropriate time delays can prevent unnecessary trips caused by momentary disturbances. However, thrust bearing failures may escalate rapidly, so protection delays must be justified through machine dynamics and risk analysis. The functional design should address alarm deadband, latching, first-out indication, reset authorization, bypass control, and Sequence of Events recording.

A clear distinction must be made between a measurement fault and actual rotor displacement. The 991 transmitter includes a Not OK response. When it detects an abnormal sensor condition, the output current drops below 3.6 mA. The controller should recognize this state as a channel fault instead of interpreting it as extreme negative displacement.

A dedicated "Axial Position Instrument Fault" alarm should indicate that effective monitoring has been lost. Depending on equipment criticality, this condition may require a start inhibit, immediate maintenance, or a controlled shutdown. The required action should be documented in advance rather than left to the operator during an incident.

The available loop voltage must also be verified. Total electrical load includes field wiring, analog input resistance, intrinsic safety barriers, isolators, and surge protection devices. Permissible loop resistance is calculated as follows:

For a 24 Vdc supply, the calculated upper limit is approximately 522 ohms. A practical design should remain comfortably below this value to accommodate supply tolerance, cable-temperature effects, and component variation.

A complete commissioning test should demonstrate that known probe movement produces the correct value, direction, alarm, and protective response at the control-system level. Checking only the transmitter output does not reveal incorrect PLC scaling, reversed polarity, disabled interlocks, or inappropriate time delays.

Before testing, technicians should confirm the transmitter model, compatible probe, total transducer length, wiring polarity, terminal voltage, target condition, bracket rigidity, and connector protection. The initial gap or gap voltage should be documented as a baseline for future maintenance.

The transmitter offers independent zero and span adjustments and includes a test input for checking loop operation. Electrical simulation is useful for validating analog conversion and alarm logic. When the machine arrangement permits, a controlled physical displacement test provides stronger confirmation because it includes the sensing gap and complete transducer chain.

During service, maintenance personnel should inspect probe security, cable connectors, shielding, moisture protection, and terminal tightness. Trend review should complement periodic hardware checks. A slow movement in average axial position may reveal bearing wear, changes in coupling force, process deposits, or altered compressor loading before the alarm threshold is reached.