Compressor Condition Monitoring and Reliability

Separate machine condition from air-system demand

The compressor package and the compressed-air network answer related but different questions. Machine monitoring examines the supplied condition indicators and operating history of the compressor. The compressed-air monitoring service measures delivery, pressure and demand across the agreed plant boundary.

Define the purchasing objective in those terms. A customer concerned about repeated machine trips may need a package-event record and specialist review, while a customer investigating overnight demand needs a network measurement boundary. Some projects require both, with a clear relationship between the two records.

The asset monitoring service supplies the wider monitoring and maintenance approach. Use the manufacturing overview to connect compressor availability with the production equipment and departments that depend on it.

Inventory the machine and its existing protection

Identify the compressor type, model, drive, controller and installed instruments. A rotary-screw package and a reciprocating machine can have different measurement locations, operating patterns and diagnostic requirements. Obtain the relevant documentation and current service information before specifying additional points.

Record the protection and control functions already provided by the manufacturer. The plant's authorised service team remains responsible for their configuration and maintenance. The monitoring scope should state which read-only observations or documented interfaces are used and how access is approved.

Collect the known issue history, recent service work and available event logs. A useful initial review may reveal that the required information already exists in the controller but is difficult to retrieve consistently. In that case, preserving and correlating the existing evidence can be more useful than adding a measurement without a clear decision attached to it.

Define operating states precisely

Agree the meanings of stopped, available, starting, loaded, unloaded, modulating and faulted for the actual machine. The controller's labels and the plant's reporting terms may differ. Retain a mapping that the operator and service specialist can both interpret.

Check the source of each state and the conditions under which it updates. A run command does not necessarily confirm a completed start, and a package-ready signal does not establish delivered air. Keep commands, machine states and process outcomes distinguishable in the event record.

For machines with variable output or speed, preserve the relevant reference and its unit. Comparing temperatures or vibration across different operating states may be misleading unless the specialist can see the duty context. Select the states needed for the intended review rather than merging every operating minute into one average.

Select temperatures for an identifiable purpose

Record the physical source and meaning of each temperature. An oil temperature, discharge temperature, bearing measurement and enclosure air reading describe different parts of the package. Preserve their identities and any conditions that affect how they should be compared.

Use manufacturer documentation and specialist advice to choose the appropriate interpretation and action rules. If an additional sensor is proposed, confirm its range, mounting, access and expected response. A convenient external surface measurement should be labelled accurately and assessed for the intended task.

Include the ambient and loading information needed to review changes. A warmer reading after a different operating period may require a different investigation from a persistent change under comparable duty. The report should show the associated conditions and the service owner's finding, not simply classify every deviation as the same fault.

Match vibration acquisition to the compressor type

Discuss the required measurement with a competent condition specialist. The motion and operating cycle of a reciprocating machine can require a different analysis approach from a rotary package. Define whether the project needs overall indicators, spectra, waveforms, speed references or another form of evidence.

Choose locations and settings for the actual bearings, structure and diagnostic question. Record orientation, attachment and installation photographs. The relationship between the machine and the sensor is part of the measurement and must remain reproducible after maintenance.

Confirm the available outputs of the selected device before purchase. A periodic summary value should not be described as equivalent to a detailed diagnostic capture. If a specialist needs an additional portable measurement after an alert, include the access and review arrangement in the response plan.

Preserve starts, trips and restart context

Collect the event sequence needed to investigate interruptions: command, state transition, fault indication, reset and return to the agreed operating condition. Align timestamps between the controller, monitoring system and any local operator record. Retain the original source of each event.

Distinguish a planned stop, a control response and a protective trip using documented meanings and reviewed evidence. The same absence of output can arise from different circumstances. An operator annotation or service report may be needed to establish the cause after the event.

Record changes made before restart. Resetting an alarm, replacing a component and changing a setting are different interventions. Keeping the sequence makes a repeated event easier to investigate and helps the service specialist decide whether the earlier action addressed the underlying issue.

Understand duty and standby operation

Map the compressor combination used during normal production and the role of standby machines. Record the actual lead/lag or sequencing arrangement where it is part of the task. A machine that runs infrequently may still require regular verification because the plant depends on it when another unit is unavailable.

Observe a planned duty change under the site's approved procedure. Record whether the expected machine starts, reaches the required state and supports the process boundary being checked. The monitoring acceptance should distinguish a successful state transition from a complete test of the plant's production resilience.

Assign ownership of standby readiness, routine service and the next permitted exercise. The record should identify open defects or unavailable equipment rather than imply redundancy solely because two compressor icons appear on the screen.

Establish a baseline and a review routine

Select representative periods with the plant and service specialist. Include the operating modes relevant to the purchasing objective and record known conditions at the start. The baseline describes the observed machine and settings; it does not replace a mechanical inspection or service assessment.

Agree how a changed trend is reviewed. The first check should establish identity, data quality and duty before an engineering conclusion is recorded. Give unresolved observations an owner and a next step, such as a reference measurement, controller-log review or planned inspection.

Use the condition-monitoring pilot guide to connect events with work orders and closure evidence. The useful output is an explained decision and verified follow-up, including periods where the evidence supports continued observation rather than immediate intervention.

Keep service work connected to the measurement history

Record service date, scope, replaced parts and relevant settings using the actual work reference. A generic service-complete event loses the detail needed to interpret a changed trend. Where lubricant or filter information matters, retain the approved record and the service specialist's interpretation.

After work, compare readings under agreed operating conditions. Identify any change in measurement position or configuration before treating the difference as an equipment outcome. Preserve the original and revised baselines where the responsible reviewer determines that a new reference is needed.

Avoid assigning an avoided failure or downtime saving to every completed work order. A credible outcome states what was observed, what the inspection established and what happened after the intervention. Financial attribution requires an agreed counterfactual and suitable production evidence, which should be recorded separately from the measurement result.

Verify the installed scope before handover

Set the expected results and tolerances before the test. Use safe simulations for functions that cannot be exercised directly during commissioning, and label the resulting evidence. Retain open dependencies so the acceptance record accurately describes what was verified on the installed machine.

Specify the data path and operating costs

Suitable LoRaWAN measurement points with 4G backhaul may support selected distributed observations, while documented controller interfaces can supply agreed package data. Match acquisition, storage and transfer to the required evidence. Detailed captures and periodic summaries can have different requirements.

Define what happens when a source, gateway or backhaul is unavailable. Test stale-data indications and the recovery of buffered records where local retention is included. Original event time must remain distinct from the time a delayed record reaches the platform.

The quotation should identify hardware, installation, configuration, specialist interpretation and recurring services. Assign user administration, communication subscriptions, routine checks and configuration changes. Provide the point list, accepted settings, export instructions and service contacts as part of the operating handover.

Prepare a compressor-monitoring enquiry

Send the machine types, controller details, duty arrangement, recent trip history and available service records. Identify the production consequence, local service specialist and permitted installation window. Discuss compressor condition monitoring around a machine-level question that the plant can verify with the agreed evidence.

The engineering method draws on compressor and condition-monitoring guidance in PNNL's Operations and Maintenance Best Practices guide. The manufacturer's instructions and responsible specialists establish the actual package settings, protection responsibilities and maintenance decisions.

Frequently asked questions