How to Verify Air Quality Against What the Process Needs

Why this matters

"The air is dry, we have a dryer" is the sentence that precedes a ruined paint job, a rusted air motor, and a customer who now believes the shop sold them a machine that does not work. A dryer carries a nameplate rating, and a nameplate is a statement about the machine at the machine's rated inlet conditions, not a measurement of the air arriving at the process. Air quality is three separate specifications, each set at the point of use, each verified by a different reading. A shop that can walk in and put numbers on all three is selling the customer something they have almost certainly never been shown.

Before you fit a gauge, open a bowl or pull a coupler

Every reading below is taken on a pressurised system, so each access carries its own isolation. To fit a test port or open a filter bowl, close the upstream isolation valve, open the vent, confirm the section gauge reads zero with the vent still open, and lock and tag the compressor disconnect open under 29 CFR 1910.147, which covers stored pressure and mechanical energy, before a wrench touches the fitting. A pressure switch will restart a compressor with no warning and no one standing at it, which is why the disconnect is locked and not merely switched.

Never uncouple a pressurised hose to reach a tool inlet: close the drop shutoff, bleed the hose through the tool's own trigger with the tool pointed into a clear area and hands and face out of the discharge path, then pull the coupler. Never clear a port or a bench with a blow gun; 29 CFR 1910.242(b) limits compressed air used for cleaning to under 30 psi with effective chip guarding and personal protective equipment, because air driven into broken skin can enter the bloodstream. Compressor rooms commonly sit at or above the 85 dBA eight-hour time-weighted average at which 29 CFR 1910.95 attaches, so hearing protection goes on at the door. Condensate drawn from a separator or a bowl is an oil and water mixture: catch it in a sealed container with nitrile gloves on, and do not put it down a floor drain until you have confirmed the disposal route with the local sewer authority, because in most jurisdictions it is a regulated waste.

Step 1: write the requirement down before you measure anything

You cannot verify air quality without a target, and the target belongs to the process, not to the air room. Ask what the air touches at each drop: a rotary screw's own controls, an impact wrench, a bag house, a paint booth, an instrument-air actuator, a food contact surface, a supplied-air respirator. Each of those has a different tolerance for the same three contaminants.

Those three are solid particulate, water, and oil, and they are independent. ISO 8573-1:2010 expresses them as three digits in that order, one class each, and it binds you only where a contract, an equipment specification or a warranty calls it out; it is not law on its own. Write the requirement as three separate lines even if the customer only names one, because the most common error in this work is treating "clean dry air" as a single quantity that one box fixes.

Step 2: separate what a dryer does from what a filter does

This is the step that decides whether the rest of the visit finds anything.

• A dryer is the only device in the room that lowers dew point. Nothing else does.
• A water separator removes liquid water and coarse droplets that have already condensed. It removes no vapour, so it cannot change dew point.
• A coalescing filter merges oil and water aerosol into drops and drains them, and catches fine particulate. It does not remove oil vapour and it does not remove water vapour.
• An activated carbon adsorber removes oil vapour and odour. It removes no water vapour, and it is destroyed by liquid carryover, which is why it always sits downstream of the coalescer.

A plant that answers a wet-air complaint by adding a filter has bought pressure drop and no dew point. Say that sentence out loud on site; it is usually the moment the conversation changes.

Step 3: read pressure dew point at the point of use, under load

Pressure dew point is the temperature at which water begins to condense out of the air at line pressure. It is the specification that matters, because that is the state the air is actually in inside the pipe. The same air expanded to atmosphere has a lower dew point, since water's partial pressure falls with total pressure, so an atmospheric dew point figure quoted for a pressurised line is not comparable and the conversion belongs to the dryer manufacturer's chart.

The rule that decides pass or fail: pressure dew point must sit below the coldest temperature that air will see anywhere downstream, with margin. The margin the trade works to is 18 F, which is 10 C carried over from the selection guidance dryer manufacturers publish; take it from the manufacturer's data for the machine you are specifying and record where you got it rather than repeating it. Not below the compressor room temperature. Not below the average shop temperature. Below the coldest metal the line runs against, which is usually an unheated bay, a roof run, or an outside wall in January.

Take the reading with a dew point transmitter at the drop, with the plant running. Note the instrument's specification honestly: a dew point transmitter is normally specified as a fixed accuracy in dew point degrees over a stated range, not as a percentage of reading, so it does not shrink as the reading falls, and it is a bound rather than a spread. Bounds add linearly and are reported as bounds. Use the same instrument at the dryer outlet and at the drop, because a fixed instrument offset is common to both readings and cancels in the difference, so the difference carries only the probe's repeatability while each absolute reading still carries the full bound. The gate is an absolute reading, so the gate carries the full bound.

Step 4: test for oil where the oil ends up

Oil arrives three ways: aerosol carried over from an oil-flooded compressor, vapour that passes straight through a coalescer, and oil already in the pipe from years before anyone cared. A field indicator tube or a sampling kit at the drop tells you total oil in the air stream; a wipe on the inside of an opened, isolated and vented drop tells you what has accumulated on the pipe wall. Both matter and they answer different questions. Fresh carryover is an air-room fault. A greasy pipe wall with clean current air is a legacy problem the process will keep meeting every time flow surges.

An oil-free compressor is oil-free at the compression element, not necessarily at the drop, because the pipe still holds whatever the last machine put there.

Step 5: particulate is a differential-pressure reading, not a visual one

You will not see the particulate that matters. What you can read is the differential across each filter element, and that is the only honest health indicator a filter has. Change elements on measured differential against the element's own rated clean drop, from the filter manufacturer's documentation, not on a calendar. A filter changed early is bought pressure drop; a filter changed late is bypass or collapse, and a collapsed element puts its own media downstream of itself.

Step 6: where the process is breathing air, ISO 8573 is not the specification

If any drop feeds a supplied-air respirator, the governing requirement is federal, not consensus. Compressed breathing air must meet at least Grade D as described in ANSI/CGA G-7.1, and under 29 CFR 1910.134(i) an oil-lubricated compressor supplying breathing air requires a high-temperature or carbon monoxide alarm, or both, and where only a high-temperature alarm is fitted the same paragraph requires the air to be tested frequently for carbon monoxide to confirm it stays inside Grade D's 10 ppm ceiling, with the intake sited so exhaust and other contaminants cannot be drawn in. Do not verify a breathing-air drop against a dew point number and call it done. If the shop has quietly teed a respirator line off the shop header, that is the finding of the visit and it goes in writing before anything else does.

The worked example

A cabinet shop runs a spray booth and a bank of impact wrenches off one header. Complaint: water spotting in the finish, intermittent, worse in winter.

The requirement. The booth's coldest downstream run passes through an unheated bay that reads 48 F on a winter morning. Applying the step 3 rule with the 18 F margin, the booth needs a pressure dew point at or below 30 F.

The nameplate. The refrigerated dryer is rated 38 F pressure dew point at its rated inlet conditions of 100 psig, 100 F inlet air and 100 F ambient. That does not meet a 30 F requirement, and no refrigerated machine can: its evaporator cannot be driven much below freezing without icing, which is a mechanism floor rather than a sizing question. So the nameplate is already telling you this branch is on the wrong class of dryer, before a single reading is taken.

The readings. Compressor room air on the day of the visit: 104 F. Dryer inlet air: 118 F. An air-cooled aftercooler at design airflow is normally expected to bring discharge air to within roughly 10 to 20 F of the cooling medium, so against 104 F room air an inlet of 114 to 124 F is what a healthy aftercooler produces. The aftercooler is doing its job. The room is the fault.

Pressure dew point at the booth drop under load, plant running: 62 F. The same transmitter at the dryer outlet: 60 F.

The reasoning. 62 F against a requirement of 30 F is a 32 F miss, and against the 48 F coldest metal it is 14 F on the wrong side, so water condenses in the bay run and arrives at the gun as liquid. The transmitter carries a fixed accuracy bound in dew point degrees, so take the worst case in the direction that would rescue the plant: even at a 4 F bound the reading is 58 F, still 10 F above the coldest metal. The conclusion survives the instrument. The 2 F rise from dryer outlet to drop is a different quantity and is not covered by that bound at all. A fixed offset is common to both readings and cancels when you subtract, so what remains on the difference is the probe's repeatability from its data sheet, typically well under a degree. Take the 2 F as real: something between the dryer and that drop is adding moisture, or the run is picking up condensate. It is small, and it is not why the plant is failing.

The call. The dryer is not failed. It is being fed inlet air 18 F above its rated condition in a room 4 F above its rated ambient, and a refrigerated dryer loses dew point capability as inlet temperature rises. The fix is compressor room ventilation, and only after the room is fixed is it fair to re-measure and ask whether the dryer itself has anything left. Be honest about where that lands: the room work recovers this dryer to its own nameplate and still leaves the booth branch short of 30 F, because 30 F is below the class floor. The booth needs a point-of-use desiccant dryer on that branch, or the bay run heat-traced and insulated so the coldest metal stops being 48 F.

The failure mode of getting it wrong. The default repair here is a bigger dryer. It gets installed into the same 104 F room, delivers better than 62 F but not enough better, and the shop has bought a machine and kept the spotting. The tell that this happened is a dryer nameplate newer than the compressor's in a room with no louvre.

How to verify you got this right

Re-read the booth line at the same drop, same instrument, with the plant at the same running state you first recorded, after the room work is done. Same state or the comparison is worthless. Confirm the new pressure dew point sits below the requirement of 30 F, not below the 48 F coldest metal, and below it by more than the instrument's bound rather than merely below it. Then walk the coldest run with a surface thermometer on the actual winter morning rather than trusting the 48 F you were told, because that number was the one your whole requirement was built on.

Finally, re-read your own three-line requirement from step 1 and confirm you closed all three. A visit that fixes water and never sampled oil has answered one of three questions and reported it as the answer.

References

• 29 CFR 1910.134(i), breathing air quality and use, including the Grade D reference to ANSI/CGA G-7.1 and the alarm requirement for oil-lubricated compressors
• 29 CFR 1910.242(b), compressed air used for cleaning
• 29 CFR 1910.147, control of hazardous energy, for isolation of stored pressure energy before opening a component
• ISO 8573-1:2010, compressed air purity classes, binding where a contract, specification or warranty adopts it
• See related: Why Moisture Is the Central Problem in Compressed Air; The Dryer Types and What Each One Can Actually Deliver; Why a Compressor Room Needs More Ventilation Than It Has