Before installing a compressed air oil monitoring system, define which form of oil you need to measure and what decision the measurement must support. An online oil vapor analyzer can reveal a change early. Its reading alone cannot demonstrate the total oil content that also includes liquid oil and oil aerosol. This guide gives plant engineering, quality, and procurement teams in Thailand a practical way to specify sampling points, alarms, reference testing, RFP requirements, and FAT/SAT acceptance.
“Oil content” covers different physical forms
The word oil can mean liquid droplets, fine aerosol, or vapor in a compressed air system. Their proportions and behavior can change with temperature, pressure, cooling, drying, filtration, and pipe conditions. A catalog claim of “oil monitoring” is therefore incomplete until the vendor states which phase its device measures and which phases it does not.
ISO 8573-1:2010 defines compressed air purity classes for particles, water, and oil. The measurement methods are described in separate parts. ISO 8573-2:2018 covers sampling and quantitative analysis of liquid oil and oil aerosols; it expressly excludes oil vapor methods. ISO 8573-5:2025 specifies pressurized sampling and gas chromatography to determine oil vapor content. A proposal that lists these standards without describing the actual sampling and analysis method is insufficient as an acceptance plan.
A photoionization detector, or PID, can be valuable for continuous indication of changes in oil vapor. CS INSTRUMENTS describes its OIL CHECK 500 as indicative continuous measurement and explains that ISO 8573-1 classification requires laboratory determination of the relevant oil phases. The distinction is operational, not a dismissal of online monitoring: detect changes quickly with an online instrument and verify a specified purity class through an appropriate testing program.
| What is measured | Practical question | Appropriate role | Common mistake |
|---|---|---|---|
| Oil vapor | Has the baseline changed, possibly due to intake air or adsorption media? | Online trend plus reference analysis when required | Calling a PID value “total oil” |
| Oil aerosol and liquid oil | Is oil being carried downstream? | Sampling and analysis under ISO 8573-2 | Assuming a vapor instrument detects aerosol |
| Total oil class | Does air at a specified use point meet its agreed requirement? | Appropriate methods and documented report | Equating one sensor alarm with class validation |
Compliance with an ISO purity class, a customer product specification, and an internal maintenance trigger are three separate decisions. Where compressed air contacts a cleaned component, air testing may need to be accompanied by checks of product surfaces and process conditions. An air monitor helps make a quality decision; it is not an automatic guarantee about the finished product.
Begin with the failure you want to detect
A specification based only on the instrument name makes quotations easy to compare but incidents hard to manage. List previous quality events, the product’s exposure to air, the relevant use points, and the consequence of stopping the line. Air blowing, coating, packaging, instrument purging, and pneumatic actuation do not share the same contamination risk.
Include plausible failure paths: compressor condition changes, deterioration of filters or adsorption media, contaminated intake air, drain problems, an incorrectly positioned bypass valve, and contamination after pipe work. An oil-free compressor does not eliminate every possible downstream source. CS INSTRUMENTS describes the influence of ambient intake conditions and system operation even when an oil-free machine is used.
For each failure path, ask which signal is likely to change first. Vapor trends may give useful early notice for some events. If aerosol carryover is the chief concern, combine appropriate sampling tests with observations of filter differential pressure and drains. Differential pressure is a useful supporting condition signal; it is not a direct measurement of oil aerosol concentration. See our article on compressed air filter differential pressure monitoring for the related maintenance design.
Record pressure, temperature, flow, dew point, filter differential pressure, compressor status, and sampling valve position on the same timeline as the vapor value. A low-flow line shutdown and full production should not be compared without context. For the flow measurement side of the design, see our compressed air flow meter selection guide.
Design sampling points around the process
The first figure distinguishes an online trend point from reference sampling points along a physical system: compressor, dryer, filters, header, branches, and critical use points. It is a location plan, not a claim that every facility needs the same number of instruments.

One analyzer on a common header may show a plant-wide trend but miss contamination introduced on a downstream branch. An analyzer at every use point raises cost and maintenance. Start with a point downstream of common air treatment and a representative high-risk use point; finalize the choice against the actual piping drawing and shutdown constraints. Check that an operator can safely reach a reference sampling port.
Do not assume the online analyzer’s sample tube can double as a standards-based test port. ISO 8573-2 methods have sampling requirements that depend on the method and flow conditions. The RFP should identify port locations, connection sizes, materials, valves, cleaning procedure, and operating state during sampling. Ask the test provider to confirm method suitability before installation.
Mark compressor changeover, dryer regeneration, filter replacement, and line restart on the drawing. Give each point a stable physical tag and database ID. Keep its equipment ID, branch number, flow direction, photograph, and drawing revision. Otherwise a value from before a piping modification may be compared with a value from a different physical location.
Pressure and temperature influence how a vapor indication is interpreted. The sample line may also introduce losses or interference. Compare vendor limits for pressure, flow, temperature, preconditioning, zero adjustment, and calibration against site conditions. A catalog detection limit is not the same thing as the uncertainty of a complete installed measurement.
Selecting an online PID oil vapor analyzer
The online instrument’s job is to record everyday changes and trigger a timely investigation. Evaluate its measurement principle together with the actual air composition, environmental conditions, calibration procedure, and maintenance capacity.
| Item | RFP question | Acceptance evidence |
|---|---|---|
| Measured phase | Does the value represent vapor only? Are aerosol and liquid oil excluded? | Product specification and screen terminology |
| Operating range | What pressure, flow, temperature, and humidity are supported? | Site-to-specification comparison |
| Interference | How do water, cleaning agents, and solvents affect the indication? | Vendor information and site trial |
| Calibration | What are the zero, calibration, replacement, and restart procedures? | Certificates, procedure, history |
| Interface | Can raw value, diagnosis, status, and communication loss be transmitted? | Signal list and PLC history |
| Response | Who investigates and who can decide on product segregation? | Work instruction and exercise |
CS INSTRUMENTS describes a PID design for continuous indicative vapor measurement, measures to reduce certain interference, zero adjustment, and a replaceable sensor. Those features are an example from one product, not a performance promise for all vendors. Ask each bidder for its target compounds, range, diagnostics, maintenance parts, and local service arrangements. At hot or condensation-prone sampling points, ask whether sample preconditioning can remove the compound you intended to monitor.
Do not set a PID alarm equal to an ISO total-oil class limit. The measured quantities differ. Use initial normal data, process risk, the cost of a false alarm, calibration results, and laboratory findings to establish operational triggers. A single universal mg/m³ threshold is inappropriate for facilities with different equipment and products.
It may help to distinguish an advisory stage, an investigation stage, and a decision on product isolation. The exact duration before each action must come from the product’s exposure, hold time, and testability. When a value rises, first check instrument diagnosis, sample flow, valve position, pressure, temperature, and recent maintenance. Separate an instrument fault from a real change in air quality.
Connect online trends to ISO 8573 testing
Figure 2 depicts two complementary paths. A PID follows vapor indications over time. A purpose-designed sampling path supports laboratory tests for liquid oil and aerosol, as well as vapor. Keeping these paths distinct is the core of an honest specification.

If a customer requires an ISO 8573-1 oil class report, define the class, use point, operating conditions, date and time, standard editions, methods, and reporting responsibility. The report should make clear how liquid oil and aerosol under ISO 8573-2:2018 and vapor under ISO 8573-5:2025 were evaluated. The CS INSTRUMENTS FAQ explicitly says a PID vapor indication alone does not classify total oil.
Parker’s compressed air quality testing white paper distinguishes class validation from indicative tests. Some editions of standards in that paper reflect its publication date, so check the current ISO pages when drafting a purchase order. In particular, the current ISO oil vapor method is ISO 8573-5:2025. Do not reproduce an old standard edition from a vendor PDF without checking it.
Online monitoring does not make laboratory testing unnecessary. Define triggers such as commissioning baseline, a contractual acceptance event, sustained trend change, system modification, or a significant quality investigation. Set periodic test frequency based on product risk and customer requirements. Keep the online values and operating conditions from the same period as each laboratory sample. A single correlation is not enough to turn the online device into a standards test.
State clearly on the test request whether the objective is vapor only, aerosol and liquid oil, or total-oil class evaluation. “An ISO 8573 test” may otherwise leave out a required phase. Confirm the laboratory’s sampling method, analytical method, measurement range, uncertainty reporting, and example report before work begins.
Write an RFP for operation, not just a price comparison
Attach a system diagram, compressor and treatment equipment list, critical use points, production schedule, pressure and temperature range, flow conditions, PLC/SCADA interface, and power-loss behavior. If a value is an estimate, label it as such and assign the site survey to confirm it. An estimate silently presented as a guarantee is a frequent source of disagreement after installation.
Functional requirements should cover the vapor indication and unit, point ID, diagnostics, sample-flow state, relevant pressure and temperature, timestamp, alarms, and missing-data indication. The interface should help staff compare trends with maintenance events and reference test results. Choose data retention from customer traceability requirements and the plant’s quality record policy.
Nonfunctional requirements include buffering during network outages, visible gaps after reconnection, time synchronization, user permissions, audit logs, calibration records, spare parts, local support, and document language. A Thai site may have Japanese headquarters, Thai maintenance staff, and English vendor manuals. Point IDs, alarm names, and parts lists must remain consistent across them.
Ask bidders to separate analyzer hardware, sampling port work, piping and electrical work, communications, commissioning, laboratory tests, calibration, consumables, spares, shutdown work, training, and annual service. An advertised “free interface” may exclude PLC program changes, history storage, or operator notification design.
Include a matrix of what each proposed system can and cannot demonstrate: online vapor trend, quantitative aerosol, total oil class validation, and causation of product contamination. Require the method and any third-party testing for each row. If a proposal says “ISO-certified sensor,” request the exact standard, test method, and meaning of the claim.
Connect to PLC and SCADA without losing context
A single number on a dashboard is poor evidence for a later investigation. Record timestamp, point ID, indicated vapor value and unit, instrument range, sample-flow or instrument state, diagnostic code, alarm state, and plant operating mode. Distinguish zero adjustment, warm-up, calibration, and communication failure from ordinary air measurements. Do not silently delete bad readings: a quality review needs to know whether air was normal or the instrument was unavailable.
Define which decisions the PLC may automate. An advisory alarm could notify maintenance; an investigation alarm could ask quality to assess product exposure; a quality owner could decide on segregation. Any connection to a safety function or machine stop requires its own risk assessment and validation. A quality trend design alone does not validate a safety interlock.
Figure 3 connects the sampling point, PID analyzer, PLC, and operator screen to a response workflow. When the alarm occurs, staff check the instrument and arrange reference sampling if needed. Keep the laboratory report beside the alarm history. A normal vapor indication does not rule out an aerosol or liquid-oil problem.

Keep a record of who acknowledged an alarm, when it happened, which sample was taken, which product lots were held, what filter work was performed, and who approved restart. A trend graph alone cannot explain a decision to a customer. If a paper checklist remains in use, link its point ID and time to the digital history.
Split FAT and SAT into meaningful evidence
At FAT, verify the supplied instrument and software against the agreed function: signals, units, ranges, diagnostic codes, communication loss, alarms, storage, and restoration. Simulated inputs can distinguish rising concentration from sensor failure or inadequate sample flow. FAT does not prove the quality of the factory’s real compressed air.
At SAT, verify the installed port, piping, power, communications, actual sample conditions, time sync, alarm recipients, display, and stored history. Record representative operating states such as normal production, compressor changeover, and line restart. If the contract includes a standards-based acceptance test, treat its sampling and laboratory report as evidence separate from functional SAT.
| Item | FAT evidence | SAT evidence | Owner |
|---|---|---|---|
| Analyzer and sample line | Specification, diagnostics, simulation | Running at actual pressure, flow, temperature | Engineering and maintenance |
| PLC and screens | Tags, units, alarms, history | Site network and recipients | Controls and IT |
| Test point | Drawing and parts list | Physical location, tag, demonstrated access | Quality and engineering |
| Reference test | Agreed scope and method | Sampling and report at stated conditions | Quality |
| Incident response | Reviewed procedure | Notification and record exercise | Quality and production |
Acceptance must cover visible data gaps, correct handling of old values after reconnection, and clear separation of calibration periods from normal measurements. Record residual issues with an owner and due date.
An alarm response that protects product decisions
When the vapor indication rises, check instrument self-diagnostics and the sampling line first. Low flow, drain trouble, an incorrect valve position, recent solvent cleaning, or temperature change may affect a reading. Review compressor status, differential pressure, dew point, and production state at the same time. Note whether the rise persists or appears at another point.
At the same time, identify the products and time window exposed to the air. Quality decides whether to stop production, hold product, or request additional inspection based on product risk and customer requirements. A PID alarm alone does not prove every lot defective; the absence of an aerosol reading does not prove the air acceptable. Order appropriate aerosol and vapor testing when the decision needs that evidence.
Afterward, review the time from alarm to acknowledgment, hold, sampling, analysis, and restart. Record the cause of a false alarm and decide whether to adjust the instrument, sample point, or operational trigger. Simply disabling a noisy alarm defeats the purpose of monitoring.
Pilot one system before expanding
Start with one critical system or use point. Collect the piping diagram, quality requirement, past events, and current inspection data. Separate online changes you wish to detect from standards-based evidence you need to produce. Run the pilot through representative day and night conditions and compressor changeovers. Set the duration to cover the plant’s operating cycle.
Write down the compared conditions before the pilot: line, point, use point, compressor combination, filters, dryer state, working days, and available sampling hours. Log changes alongside the readings. Review variability and missing data as well as average values. A repeated variation during every process change may call for an operating-state-aware alarm.
Train maintenance to check the sample line, quality to distinguish an indicative value from a class test, and production to record the notification and affected lots. Practice three different scenarios: instrument fault, rising vapor indication, and an unfavorable reference test. Confirm which role records and approves each action.
Use the findings to decide on more points, revised triggers, laboratory test frequency, and maintenance procedures. When expanding across lines, standardize IDs, units, state codes, timestamps, and links to test reports. Measure benefits through time to detection, time to identify a cause, sampling delay, product traceability, alarm usefulness, and instrument availability. Do not claim “zero oil” from installing a sensor.
Remove ambiguous wording before comparing proposals
Agree terminology before comparing bids. “Residual oil” may refer to a vapor indication on one screen and total oil including liquid and aerosol in another document. “Real-time measurement” may describe only the refresh interval, while a reader might assume it delivers an automatic quality decision. Define the displayed quantity, phase, measuring principle, unit, point, refresh interval, and conditions under which a value is invalid. A short terminology table connecting English vendor documents, Thai operator screens, and Japanese quality reports reduces handover mistakes.
“ISO 8573 compatible” is also too broad. State whether a claim relates to Part 1 purity classes, Part 2 liquid oil and aerosol testing, or Part 5 vapor testing, and identify the edition. An online instrument described as providing an indicative vapor measurement under Part 5 is not thereby a total-oil class test under Part 1. The test request should name the phases and purpose. Ask bidders to separate values generated by their own equipment from reports that require an external laboratory. If the lowest bid simply omits a phase, it is not a like-for-like price comparison.
“Stop on alarm” does not define a workable response. An instrument diagnostic, insufficient sample flow, communication failure, and rising vapor value need different actions. For every alarm, name its recipient, first check, provisional product action, need for reference sampling, and restart approver. If quality staff are absent overnight, specify who can place product on temporary hold and who reviews it the next morning. Link measured values with notification, acknowledgment, and affected lot times.
Separate performance responsibilities as well. An instrument vendor can warrant equipment specifications under stated conditions and its service scope. A system integrator implements wiring, display, history, and alarm routing. A purity-class test is a distinct sampling and analytical activity. Final product disposition belongs to the plant’s quality system. Replacing all four with a single promise of “quality assurance” makes investigation and accountability difficult if a test fails.
Frequently asked questions
Can an oil vapor sensor alone prove an ISO 8573 oil class?
No. A PID vapor indication and total oil address different quantities. ISO 8573-2:2018 covers liquid oil and aerosol; ISO 8573-5:2025 covers vapor. Agree the sampling and analysis program with a competent test provider when an oil class must be validated. Use the online value for trends and early investigation.
Can a PID quantify oil aerosol in compressed air?
Do not use a PID vapor analyzer as an aerosol quantification method. Check the specific product’s scope. ISO 8573-2:2018 addresses liquid oil and aerosol testing. Filter differential pressure and drain observations are supporting signals, not aerosol concentration measurements.
How many ISO 8573 oil test points do we need?
There is no universal number. Design the plan around the customer-specified use point, common treatment equipment, and high-risk branches. Document each location, operating state, and sampling method. Do not present a compressor-room result as evidence for every distant use point.
Is monitoring still relevant with an oil-free compressor?
Decide from the product requirement and system risk. Intake air, downstream equipment, piping, and maintenance can affect air at the use point. Compressor type alone does not establish use-point purity. Separate online monitoring from periodic reference testing where both are needed.
What belongs in the RFP and FAT/SAT?
Specify phases measured, test points, site conditions, signals, diagnostics, history, service, and reference testing. FAT verifies instrument and software behavior; SAT verifies real installation, operating conditions, notification, and records. If a class is contractual, require an appropriate test report as separate evidence.
Conclusion
A useful compressed air oil monitoring system connects early warning from online oil vapor trends with appropriate verification of liquid oil, aerosol, and vapor. Define the sampling points, measured phases, operating conditions, and alarm response before comparing instruments. Never turn an indicative PID trend into a claim of total-oil class compliance.
If your Thailand plant is still selecting sampling points, defining PLC signals, or preparing an RFP and FAT/SAT checklist, contact TOMAS TECH. A piping diagram and the quality requirement at the use point are enough to begin a discussion about how monitoring and reference tests should work together.