Air quality excursion investigation workbook

Why a failed air quality result needs a structured investigation

A laboratory report, an online dew-point alarm or a complaint from the production floor all raise the same question: is the compressed air reaching our process clean enough? The temptation is to replace a filter, take one more sample and move on. That rarely answers what actually happened.

A number above a limit may matter a great deal without identifying the failed component. A doubtful sample may still need urgent action without proving that process air was contaminated. A good repeat sample does not tell you what the air was like last week. And a pass upstream of a long pipe run says nothing about a user connection that was never tested.

The workbook keeps five milestones separate, because each needs its own evidence:

The workbook supports the engineering record. Decisions about product quality, batch release or regulatory reporting remain with the plant’s quality team and its own procedures.

Step 1: protect the process without destroying the evidence

If there is a real risk to product or people, act first. Isolate affected users, switch to a standby supply or bring in rental air. But record the condition of the system before changing it.

• Photograph filter housings, drain valves, dryer displays and the sample point before anything is touched.
• Note dryer dew-point readings, compressor discharge temperature, pressure and alarms at that moment.
• Keep removed filter elements and drained condensate, labelled with time and location.
• Record who changed what, and when, in the incident log.

Replacing a saturated filter element straight away may solve the immediate problem, but it also removes the best clue to how much oil or water came through.

Step 2: confirm the requirement and the point where it applies

Quality requirements are usually expressed using ISO 8573-1 purity classes for particles, water and oil, or as a specific limit set by the plant or its product standard. Before comparing a result, fix:

• Which contaminant and which fraction the requirement covers, for example total oil including vapour or only liquid and aerosol.
• The point where it applies: compressor outlet, after the dryer, at the header or at the user connection.
• The reference conditions and units of the limit.
• The revision of the requirement in force at the time of sampling.

A result taken after the dryer cannot be compared with a requirement that applies at a filling machine 150 metres downstream, and the reverse is also true.

Step 3: check that the sample is valid

Doubt about a sample is not the same as a pass. Review validity on its own merits, then decide what the result can and cannot support.

Read the report exactly. A “less than” result is not zero; it means the value is below the method’s reporting limit. Where uncertainty is stated, a result close to the limit may be indeterminate rather than a clear pass or fail.

Step 4: rebuild the timeline and the affected period

Lay out what happened in time order, and separate the moment something occurred from the moment someone noticed it.

1. Last valid passing result at the relevant point.
2. Maintenance, filter changes, dryer service or compressor work since then.
3. Process or demand changes: new users, longer shifts, higher flow.
4. Alarms, trips and power interruptions from compressor and dryer controllers.
5. Ambient conditions; Indian summer heat and monsoon humidity often coincide with moisture problems.
6. Sampling time of the adverse result and the time it was reported.

From this, define three periods: the observed period, when an excursion is directly evidenced; the possible period, between the last pass and first failure; and any unknown period with no data. Do not describe the whole possible period as contaminated unless the evidence shows it.

Step 5: map the contamination pathway

Trace the air from intake to the point of use and ask, at each stage, how the contaminant could enter or pass through.

Moisture should be investigated with pressure and temperature recorded, because dew point depends on pressure. A refrigerated dryer typically gives a pressure dew point around +3 °C, which is not enough for a −40 °C requirement at any flow. A dryer display shows conditions at the dryer, not at every terminal point.

Oil should be investigated by fraction. Liquid and aerosol oil are removed by coalescing filters; oil vapour needs activated carbon or an oil-free compression source. A plant with an oil-injected screw compressor and good filtration can meet demanding classes, but it depends on every filter stage being healthy and correctly drained.

Step 6: test causes with a hypothesis register

List every plausible cause, then record evidence for and against each one. Keep contradictory evidence instead of discarding it.

• Cause: separator element failure. For: rising oil consumption, high separator pressure drop. Against: oil fine at compressor outlet sample.
• Cause: dryer overloaded. For: high inlet temperature on hot afternoons. Against: dew-point logger steady at −40 °C throughout.
• Cause: sampling artefact. For: lab flagged a control concern. Against: repeat sample at the same point also failed.

Plan repeat tests around the questions that remain open, for example sampling upstream and downstream of a suspect filter at the same time, rather than simply repeating the original test.

Worked examples

Particle count with a sampling pause

A terminal sample records 25 minutes of collection at 2 L/min, but the instrument log shows a five-minute pause with no flow. Active time is 20 minutes. Accepted volume = 2 × 20 ÷ 1,000 = 0.040 m³. With 12 particles counted in the size band, concentration = 12 ÷ 0.040 = 300 particles per m³. Using the full 25 minutes would wrongly give 240 particles per m³. The correction changes the number; it does not identify the source.

Oil fractions against a total-oil limit

A requirement is total oil no more than 0.010 mg/m³ at a user connection. The laboratory reports liquid and aerosol oil of 0.006 mg/m³ and vapour below 0.008 mg/m³. Adding them gives up to 0.014 mg/m³, so compliance cannot be shown, even though neither figure alone exceeds the limit. A more sensitive vapour method or carbon filtration may be needed.

These figures are illustrative examples, not recommended purity limits.

Step 7: correct, recover and close

Keep four actions distinct in the record:

1. Correction: the physical fix, such as a new separator element, repaired drain or dryer service.
2. Recovery: evidence the air now meets the requirement at the relevant points, tested under representative load and ambient conditions.
3. Product decision: made by the process owner under the plant’s procedures.
4. Effectiveness: follow-up results over the following weeks showing the fix holds.

Replacing a suspect component does not by itself complete the investigation. Close the record only when the cause is supported, recovery is shown and the follow-up plan is agreed.

How Reddy Compressors helps

Reddy Compressors services compressors, dryers and filtration of all major brands, and can trace contamination problems on the equipment side. We are an independent multi-brand specialist working from Tarnaka, Secunderabad, covering Telangana and Andhra Pradesh with more than 20 service engineers and technicians, and supplying equipment and spares across India. We are not an authorised dealer of any brand. We have tie-ups with all major and reputed brands.

• Inspect compressors, separators, dryers, filters and drains along the air path.
• Log dew point, pressure and temperature to test moisture hypotheses.
• Replace separator elements, filter elements and drains with parts identified from the nameplate.
• Recommend treatment changes, including oil-free compressors where the risk justifies it.
• Supply rental compressors, including oil-free machines, while the permanent system is corrected.

Laboratory analysis and product release remain with accredited laboratories and your quality team; we support them with the engineering evidence.

What to send us

• The test report, including method, point, qualifiers and uncertainty.
• The requirement or ISO 8573-1 class that applies, and where it applies.
• Compressor, dryer and filter makes, models and nameplate photos.
• A simple sketch of the air path from compressor to the affected user.
• Recent maintenance records and any controller alarms.

Frequently asked questions

Does one failing air quality sample prove the compressor is the cause?

No. A failed sample shows that contamination was measured at one point and time, assuming the sample is valid. The source could be the compressor, the dryer, a filter, drains, intake air, pipework or the sampling itself. Map the air path, collect evidence for and against each cause, and sample at more than one point before deciding which component to blame.

Why is there oil in my compressed air?

Common reasons are a worn or damaged separator element, the compressor overfilled with oil, high discharge temperature, saturated or poorly drained coalescing filters, and old oil residue in pipework. Oil vapour can pass through coalescing filters and needs activated carbon. Check oil consumption, separator pressure drop and filter condition, and sample both upstream and downstream of the filters.

Why is there water in my compressed air lines after the dryer?

The dryer may be overloaded by high inlet temperature or flow, have a refrigerant or desiccant fault, or a bypass may be open. Failed drains, branch lines taken from the bottom of a header and cold outdoor pipework are other causes. Measure dew point at the point of use, not only at the dryer display, and compare it with your requirement at line pressure.

Is a less-than result on a lab report the same as zero?

No. A less-than result means the value was below the method’s reporting limit, so the true value could be anywhere from zero up to that limit. When fractions are added, such as liquid oil plus vapour, use the limit value cautiously. If the combined figure could exceed your requirement, you cannot claim compliance and may need a more sensitive method.

Can a passing retest close an air quality incident?

A passing retest shows the condition at the time of the retest. It does not show what the air was like during the earlier period, nor why the first result failed. Close the incident only once the cause is supported by evidence, a correction is made, recovery is shown under representative conditions and the process owner has decided what happens to any affected product.

How often should compressed air quality be tested?

That depends on how critical the air is and on your industry’s rules. Pharmaceutical and food plants often follow a documented schedule agreed with their quality team, with extra tests after maintenance or any excursion. Continuous dew-point monitoring is common where moisture is critical. The air quality sampling plan builder helps set test points, contaminants and frequency in a consistent way.

Related pages

• Air quality sampling plan builder
• Compressed-air quality measurement
• Air quality and treatment architecture
• Air quality remediation projects
• Compressed air for pharmaceutical plants
• Compressed air knowledge and calculation centre