Compressed air blows containers clean before filling, conveys powders, actuates equipment inside cleanrooms, dries washed components and blankets product with nitrogen. In each of those it contacts product or product-contact surfaces directly — and in a large proportion of facilities it has no microbiological specification at all.
Why It Gets Overlooked
Water systems get attention because everyone recognises water as a contamination route. Compressed air is treated as an engineering utility, owned by maintenance rather than by quality, and it rarely appears in a microbiological monitoring programme.
The oversight matters because compressed air is a direct route. Air blown into a container immediately before filling deposits whatever it carries onto the product contact surface, with no subsequent step to remove it.
Where Contamination Enters
- Intake air — the compressor draws ambient air, and everything in it. Intake location matters enormously; a compressor drawing from beside a drain, a loading bay or a cooling tower is drawing exactly what you would not choose.
- Condensate in receivers and distribution lines. Compression and cooling produce liquid water, and a receiver with standing condensate is a growth environment feeding the whole system.
- Oil carry-over from lubricated compressors, providing nutrient as well as being a contaminant in its own right.
- Distribution pipework, particularly low points and dead legs where condensate collects.
- Point-of-use filters that have been in place too long, becoming a growth site on the downstream face.
- Hoses and fittings, frequently left on the floor between uses and then connected to a product line.
- Desiccant dryers where the media is degraded or the regeneration cycle is failing.
The condensate point is the dominant one. A dry compressed air system is microbiologically stable; a wet one is a water system with air in it.
What to Test
| Parameter | Method | Applies to |
|---|---|---|
| Viable count | Impaction sampler onto agar, defined volume | Product-contact air and gases |
| Non-viable particles | Particle counter | Cleanroom-supplied air, ISO 8573-4 |
| Water content / dew point | Instrument | All systems; predicts microbiological risk |
| Oil content | ISO 8573-2 methods | Lubricated compressors |
| Specific organisms | Selective media where indicated | Risk-based |
The viable count is the microbiological test and it needs a sampler designed for compressed gas — a pressurised gas stream cannot simply be directed at a settle plate. Dedicated compressed gas samplers reduce the pressure and control the flow so a defined volume impacts the agar at a velocity that does not desiccate or damage the organisms.
ISO 8573-1 Compressed Air Purity Classes
Compressed air quality is specified using ISO 8573-1, which sets purity classes for three contaminants: solid particles, water and oil. A specification is written as three numbers in that order. For example, ISO 8573-1:2010 [2:2:1] means particle class 2, water class 2 and oil class 1.
Solid particles
| Class | 0.1 to 0.5 µm (max per m³) | 0.5 to 1 µm (max per m³) | 1 to 5 µm (max per m³) |
|---|---|---|---|
| 0 | As specified by the user, stricter than class 1 | ||
| 1 | 20,000 | 400 | 10 |
| 2 | 400,000 | 6,000 | 100 |
| 3 | Not specified | 90,000 | 1,000 |
| 4 | Not specified | Not specified | 10,000 |
| 5 | Not specified | Not specified | 100,000 |
Classes 6 and 7 are defined by mass concentration (up to 5 and up to 10 mg/m³), and class X is above 10 mg/m³.
Water
| Class | Pressure dew point |
|---|---|
| 1 | -70 °C or lower |
| 2 | -40 °C or lower |
| 3 | -20 °C or lower |
| 4 | +3 °C or lower |
| 5 | +7 °C or lower |
| 6 | +10 °C or lower |
Classes 7 to 9 are defined by liquid water content. Dew point is the parameter that matters most for microbiology: air dried to -40 °C pressure dew point leaves too little moisture for organisms to multiply in the pipework, while air from a refrigerant dryer at +3 °C can condense in cold sections and support growth.
Oil (aerosol, liquid and vapour combined)
| Class | Total oil concentration |
|---|---|
| 0 | As specified by the user, stricter than class 1 |
| 1 | 0.01 mg/m³ or less |
| 2 | 0.1 mg/m³ or less |
| 3 | 1 mg/m³ or less |
| 4 | 5 mg/m³ or less |
| X | More than 5 mg/m³ |
"Oil-free" compressors avoid oil in the compression chamber, but the intake air still carries hydrocarbon vapour, so oil class still has to be demonstrated by testing rather than assumed.
The Parts of ISO 8573
ISO 8573-1 sets the classes. The other parts give the test methods:
| Part | What it covers |
|---|---|
| ISO 8573-2 | Oil aerosol content |
| ISO 8573-3 | Humidity (dew point) |
| ISO 8573-4 | Solid particle content by particle count |
| ISO 8573-5 | Oil vapour and organic solvent content |
| ISO 8573-6 | Gaseous contaminants such as CO, CO2, SO2 and NOx |
| ISO 8573-7 | Viable microbiological contaminants |
| ISO 8573-8 | Solid particles by mass concentration |
| ISO 8573-9 | Liquid water content |
ISO 8573-7 is the part that defines how to sample compressed air for viable organisms, by impaction onto an agar surface after reducing the pressure. ISO 8573-1 itself does not set a microbial class, so the microbial limit has to come from the industry standard or the cleanroom grade of the area where the air is used.
Sampling Without Invalidating the Result
- Use a purpose-built compressed gas sampler with a pressure-reducing inlet, not a general air sampler or an open plate.
- Sample at the point of use, downstream of any point-of-use filter, because that is what reaches the product. Sampling at the receiver tests the generation rather than the delivery.
- Purge the line before sampling, for a defined period, so the sample reflects the flowing system rather than stagnant air in a dead-ended sample port.
- Control the flow rate to the sampler's validated specification, since impaction efficiency depends on velocity.
- Sample an adequate volume — typically 500 to 1,000 litres for a low-count system, since a small volume gives a limit of detection too high to be meaningful.
- Include the hose and fitting actually used in production, since a clean pipeline delivering through a contaminated hose delivers contamination.
- Record dew point and pressure at the time of sampling, because a wet reading explains a high count.
Point three is the most commonly wrong. A sample port left dead-ended between monthly samples accumulates condensate and organisms that have nothing to do with the flowing system, and a result taken without purging measures the port.
Setting Limits
There is no single compendial limit for compressed air microbiology, which is part of why the programme is often absent. The workable approach is to set limits by what the air contacts.
- Air contacting sterile product or Grade A zones — treat as Grade A air, no recovered growth, and filter at point of use through a validated sterilising-grade filter.
- Air contacting non-sterile product directly — limits derived from the product's microbiological specification, typically aligned to the grade of the room it is used in.
- Air used for equipment actuation inside a cleanroom — aligned to the room grade, since leakage enters the room air.
- Plant air with no product contact — no microbiological specification needed, but keep it in a separate distribution system so it cannot be cross-connected.
That last point is worth enforcing physically rather than procedurally. Where product-contact air and general plant air share a distribution network, a single incorrect connection contaminates the controlled system, and the failure will be invisible until product results move.
What Purity Class Do Pharma and Food Plants Need?
| Application | Commonly specified quality |
|---|---|
| Pharmaceutical product contact or sterile area supply | Particle class 1 or 2, water class 2, oil class 1, with microbial limits matching the cleanroom grade at the point of use |
| Pharmaceutical non-contact (instrument air, actuators) | Particle class 2 to 3, water class 2 to 4, oil class 2 to 3 |
| Food and beverage, direct contact | Commonly 2:2:1 with a microbial limit, following the BCAS food-grade compressed air guideline |
| Food and beverage, indirect contact | Commonly 2:4:2 |
| Medical air in hospitals | Set by the pharmacopoeia monograph and medical gas standards (ISO 7396-1 for pipelines) |
| Breathing air | Set by respiratory protection standards, with limits for CO, CO2, oil and water |
For pharmaceutical use, the microbial limit at the point of use normally follows the grade of the room the air enters: for example, air or nitrogen supplied into a Grade A zone is expected to give no recovered growth, consistent with the Grade A air limit. Food safety schemes such as BRCGS and FSSC 22000 expect a documented risk assessment of compressed air and testing where the air contacts product or packaging.
Filtration at Point of Use
For any air contacting product, the point-of-use filter is the control that actually protects the product, and it needs treating like a filter rather than like a fitting.
- Sterilising-grade 0.2 µm rated for gas service where the air contacts sterile product.
- Hydrophobic membrane, since a hydrophilic filter wetted by condensate blocks and then may be blown through.
- Integrity tested where the application warrants, using the gas-service method rather than a liquid test.
- Change frequency defined and enforced, based on differential pressure and time in service rather than on when someone notices.
- Housing sanitised at change, since the downstream face and the housing interior are the growth site.
- Installed correctly — a filter fitted backwards is a common and completely invisible failure.
Compressed Air Testing in Food and Beverage Plants
Food and beverage plants use compressed air in ways that bring it into direct contact with product: blowing and drying bottles and cans, conveying powders, mixing, sparging, cutting with air knives, and blowing out bags and packaging. Air that contacts product or primary packaging is effectively an ingredient, and food safety auditors increasingly ask for evidence of its quality.
The British Compressed Air Society food-grade guideline is the most widely used reference. It recommends ISO 8573-1 class 2:2:1 for air in direct contact with food, 2:4:2 for indirect contact, and a microbial limit for contact air, with sterile filtration at the point of use where the risk justifies it. BRCGS Food Safety and FSSC 22000 require a documented risk assessment of compressed air and testing at a frequency based on that risk, usually at least once a year for contact points.
In Indian food plants, the most common findings are refrigerant dryers sized for the original demand rather than the current one, condensate drains that fail closed, coalescing filters past their change date, and plastic hoses at the point of use that are never replaced. Each one is cheap to fix and easy to miss without testing.
Troubleshooting a Failed Compressed Air Test
| Result | Most likely causes | First actions |
|---|---|---|
| High microbial count at one point only | Contaminated hose, fitting or sample port; stagnant dead leg | Replace or sanitise the hose and fitting, purge longer, resample |
| High microbial count at many points | Wet air in the distribution, failed dryer, colonised receiver or pipework | Check dew point at the dryer outlet and in the loop, drain receivers, inspect filters |
| Moulds rather than bacteria | Moisture and intake of outdoor air, condensate in cold sections | Check intake location and filters, dew point against the coldest pipe run |
| High particle counts | Filter element failure, corrosion in carbon steel pipework, poor sampling technique | Check filter differential pressure and elements, inspect pipe material, repeat with a clean diffuser |
| Dew point out of specification | Dryer overloaded at peak demand, desiccant exhausted, bypass valve open | Trend dew point against demand, service the dryer, check bypasses |
| Oil above class limit | Compressor wear, saturated coalescing or carbon filters, contaminated intake | Replace filters, check compressor condition and intake location |
Whatever the cause, document the investigation, the corrective action and a passing retest before the point of use is released back to product contact. Repeated failures at the same point usually point to a design problem, such as a low point with no drain or a long dead leg, rather than a maintenance one.
Nitrogen and Other Process Gases
Nitrogen used for blanketing, purging or pressure transfer contacts product as directly as compressed air, and it carries an additional consideration: the supply route.
Cylinder and bulk-delivered nitrogen arrives with a certificate covering chemical purity, which does not address microbiology. On-site generated nitrogen is produced from compressed air and inherits whatever the compressed air system carries. Either way, the point-of-use filter is what protects the product, and the certificate is not a substitute for it.
Qualifying a Compressed Air System
A compressed air system used for product contact should be qualified like any other utility, and in most facilities it never has been.
Design review establishes what the system is actually required to deliver — which points of use contact product, what quality each needs, and whether product-contact and general plant air are physically separated. Systems that grew over time frequently have cross-connections nobody documented.
Installation qualification confirms the compressor type, dryer, receiver, filtration train and distribution match the design, and that materials of construction are appropriate. Oil-free compressors remove one contamination route entirely and are worth the capital where product contact is involved.
Operational qualification demonstrates the system delivers specified pressure, flow and dew point across the demand range, including at peak simultaneous draw, which is when systems most often fail to hold dew point.
Performance qualification samples every product-contact point of use across a defined period, establishing the baseline microbiological performance and the routine monitoring limits.
The dew point element deserves emphasis because it is the parameter that predicts everything else. A system holding a pressure dew point comfortably below the lowest ambient temperature the pipework will see cannot form condensate, and without condensate the microbiological risk largely disappears. A system that holds dew point at peak load and loses it at low load — which happens with poorly controlled refrigerant dryers — forms condensate intermittently, which is harder to detect and just as damaging.
Compressed Gas Testing Procedure, Step by Step
- Map the system. Identify every point of use, the compressor, dryer, filters, receivers and distribution, and classify each point as product contact, indirect contact or non-contact.
- Select sampling points. Include the points closest to product, the furthest points of the distribution loop and any low points or dead legs where condensate can collect.
- Collect the microbial sample following the sampling rules above: a pressure-reducing compressed gas sampler at its validated flow rate, a defined purge, and typically 500 to 1,000 litres onto a general medium such as tryptic soy agar.
- Run negative controls. An unexposed plate and a handling control show the sampling did not introduce contamination.
- Incubate and count. Incubate per the site's environmental monitoring procedure, commonly at 20 to 25 °C and then 30 to 35 °C, count colonies and calculate CFU per m³.
- Identify isolates from product-contact points, at least to genus, and compare with the environmental flora.
- Test particles, dew point and oil at the same points, using a particle counter with a high-pressure diffuser, a dew point meter and an oil indicator tube or analyser.
- Report and trend. Compare results with the limits for each point and trend them over time.
Building a Compressed Gas Monitoring Programme
Qualification proves the system works once; monitoring shows it keeps working. A typical risk-based programme looks like this:
| Point of use | Microbial | Particles | Dew point | Oil |
|---|---|---|---|---|
| Product contact, Grade A and B areas | Quarterly to monthly | Quarterly | Continuous or monthly | Six-monthly to yearly |
| Product contact, Grade C and D areas | Six-monthly to quarterly | Six-monthly | Monthly | Yearly |
| Non-contact (instrument air, actuators) | Yearly or not required, by risk assessment | Yearly | Monthly | Yearly |
Frequencies should increase after a failure, a maintenance intervention on the dryer or filters, a change to the distribution, or a long shutdown, and they can be reduced with a documented history of stable results. Continuous dew point monitoring at the dryer outlet, with an alarm, is the cheapest and most effective microbiological control in the whole system.
Reading a Compressed Air Test Report
A useful report should state, for each point:
- The sampling point identity, its classification and the limit applied
- The sampling equipment, flow rate, sample volume and calibration status
- Microbial count in CFU per m³, with isolate identification where found
- Particle counts by size channel and the ISO 8573-1 class achieved
- Pressure dew point and the corresponding water class
- Oil concentration and the corresponding oil class
- Pass or fail against the specification, and any deviations during sampling
A report that gives only "complies" without the measured values cannot be trended and will not satisfy an auditor.
How DSTechnoverse Helps
Compressed gas is the utility we most often find missing from an otherwise complete monitoring programme.
That covers compressed air and process gas viable sampling with purpose-built samplers, sampling point selection at genuine points of use, limit setting derived from what the gas contacts, dew point and condensate assessment, point-of-use filter programme review, organism identification tracing contamination back to intake, condensate or distribution, and integration with the wider environmental monitoring programme.
Our teams work from Indore and carry out utility sampling campaigns at client sites across India.
Common Findings in Compressed Air Systems
The same handful of problems account for most of what we find when a facility samples its compressed air for the first time.
- No specification at all for air that demonstrably contacts product, so there is nothing to fail against and no monitoring history.
- Sampling at the receiver rather than at points of use, which tests generation and misses everything the distribution adds.
- Dead-ended sample ports never purged, so the sample measures the port rather than the flowing system.
- Point-of-use filters with no defined change frequency, left in place for years and acting as growth sites.
- Filters fitted backwards, an entirely invisible failure until someone dismantles the housing.
- Hoses left on the floor between uses and then connected to a product line without sanitisation.
- Condensate in the receiver, undrained, feeding organisms into the whole distribution network.
- Lubricated compressors on product-contact duty where oil-free would remove a contamination route entirely.
- Shared distribution between plant air and product-contact air, with a cross-connection nobody documented.
The first item is the most consequential and the easiest to fix. Writing a specification for each product-contact point of use, derived from what the air touches, converts an unmonitored utility into a controlled one and costs nothing but the decision.
Frequently Asked Questions
What is ISO 8573-1?
The international standard that defines purity classes for compressed air by three contaminants: solid particles, water and oil. A specification such as ISO 8573-1:2010 [1:2:1] gives the particle, water and oil class in that order. Microbial contamination is tested under ISO 8573-7.
What does compressed air class 1.2.1 mean?
Particle class 1 (no more than 20,000 particles of 0.1 to 0.5 µm, 400 of 0.5 to 1 µm and 10 of 1 to 5 µm per m³), water class 2 (pressure dew point of -40 °C or lower) and oil class 1 (total oil of 0.01 mg/m³ or less). It is a common specification for pharmaceutical product-contact air.
How do you test compressed air for bacteria?
A compressed gas sampler reduces the air to atmospheric pressure and impacts a measured volume, typically 500 to 1,000 litres, onto an agar plate. The plate is incubated, colonies are counted and the result is reported as CFU per m³, with isolates identified where the air contacts product. ISO 8573-7 describes the method.
Can compressed air testing be outsourced?
Yes. Many pharmaceutical, food and device plants use an external laboratory that brings calibrated samplers, collects samples at each point of use, tests particles, dew point, oil and microbes, and issues a report against ISO 8573-1 classes and the site's microbial limits. Outsourcing also gives an independent result for audits.
Does compressed air really need microbiological testing?
Where it contacts product or product-contact surfaces, yes. Air blown into a container before filling deposits whatever it carries directly onto the surface, with no subsequent removal step.
Can we use a normal air sampler?
No. A pressurised gas stream needs a sampler with a pressure-reducing inlet and controlled flow, or the velocity damages organisms and the impaction efficiency is unknown.
Where should we sample?
At the point of use, downstream of any point-of-use filter, after purging the line. Sampling at the receiver tests generation rather than what reaches the product.
What limit should apply?
There is no single compendial figure. Set it by what the gas contacts — Grade A air for sterile product contact, room-grade alignment for cleanroom use, and no specification for non-contact plant air kept on a separate network.
Is a supplier certificate enough for nitrogen?
No. Certificates cover chemical purity, not microbiology. The point-of-use filter is what protects the product regardless of the supply route.
Should compressed air be tested if it never touches product?
Not microbiologically, but it should be physically separated from any product-contact system. The risk is not the plant air itself but cross-connection — a shared distribution network means a single incorrect connection puts uncontrolled air into a controlled line, and the failure is invisible until product results move. Separation should be physical, using incompatible fittings where practical, rather than relying on labelling and procedure.
Does oil-free compression remove the need for testing?
It removes one contamination route and not the others. An oil-free compressor eliminates hydrocarbon carry-over, which is both a chemical contaminant and a nutrient source, and it is the right specification for product-contact duty. What it does not address is intake air quality, condensate in the receiver and distribution, filter condition, or hose and fitting hygiene — all of which remain live routes. Oil-free is a worthwhile control that reduces risk rather than a substitute for monitoring.
How often should compressed air be sampled?
Risk-based, and typically monthly to quarterly for product-contact points depending on grade. Air feeding a Grade A zone warrants the same frequency as the zone itself. What matters more than the interval is that dew point is monitored continuously, because dew point predicts microbiological risk far more reliably than a monthly count detects it — a system that stays dry stays clean, and a monthly sample can easily fall between two condensate events.
What is the commonest cause of a high count?
Condensate. A wet system is effectively a water system with air in it, and dew point is the parameter that predicts microbiological risk most reliably.
A monitoring programme that samples convenient locations produces a reassuring dataset and misses the organisms. DSTechnoverse designs and runs environmental monitoring programmes with risk-based site selection, limits derived from your own data, and statistical trending. Based in Indore, Madhya Pradesh, travelling to client sites across India. Talk to our microbiology team.
Related articles
- Cleanroom Classification as per ISO 14644-1: ISO Classes, Particle Limits and GMP Grades
- Cleanroom Grades and Environmental Monitoring Limits Explained
- Environmental Monitoring Programme Design: Sites, Frequency and Limits That Actually Work
- Gowning Qualification: Proving an Operator Can Enter the Cleanroom