How to Tell Whether a Dryer Is Doing Its Job

Why this matters

A compressed air dryer with a green light on the panel, a clean sight glass and a gauge reading in the middle of its band can still be handing the plant air that rusts a header, blisters a spray finish and freezes a valve in an outdoor line at 3 am in January. Every indicator on the dryer reports the dryer's health. Not one of them reports the dryness of the air leaving it. The gap between those two things is where callbacks live, and it is the single most sellable measurement in a compressed air survey, because almost no customer has ever been shown the number.

Isolate before you tap anything

Adding a sample point means opening a pressurised path. Close the upstream isolation valve, bleed the section through its own vent and confirm the section gauge reads zero with the vent still open before a wrench touches a fitting, then lock and tag the isolation under 29 CFR 1910.147, which is the standard that covers stored mechanical and pressure energy. A receiver or a dryer vessel holds its energy after the compressor stops, so a silent machine is not a safe one. It is also not a stopped one: a pressure switch will restart a packaged compressor with nobody standing at it, which is why the disconnect is locked and not merely switched. And on a twin-tower desiccant machine a verified zero is not stable, because the switching valve repressurises the offline tower on its own cycle timer. Isolate the tower from the switching valve as well as from the header, put the controller in a state that cannot switch, and re-verify zero at that tower's own gauge with its vent still open.

If the sample tee has to go in near the dryer's control panel, treat the electrical work as its own job: de-energize at the disconnect, lock and tag it under 29 CFR 1910.333(b)(2), which is where electrical isolation lives rather than 1910.147, and prove dead live-dead-live per NFPA 70E-2021, 120.5, in the edition your employer's electrical safety program has adopted. On a construction site the electrical counterpart is 29 CFR 1926.417.

Two more that apply while you are working: never point a bleed or a sample vent at skin, because compressed air driven under the skin can embolise, and never clear a plugged sample line with a blow gun. Cleaning with compressed air is limited under 29 CFR 1910.242(b) to under 30 psi with effective chip guarding and personal protective equipment, and a sample line is cleared by replacing it.

The gate: what the air has to be, not what the dryer has to be

There is one gate and everything else in this article feeds it.

Measured pressure dew point at the dryer outlet, taken at line pressure, under real load, at the worst hour, must sit below the lowest temperature any pressurised metal in the system reaches, with margin.

Pressure dew point (PDP) is the temperature at which water starts to condense out of the air at the pressure the air is actually at. It is not the same number as the dew point of that same air after it expands to atmosphere, and the difference is large.

A common design margin is 18 F below the lowest metal temperature the pressurised line reaches. Confirm it against the dryer manufacturer's stated requirement and against any purity class the customer's own specification calls out. Where a specification names ISO 8573-1:2010, that standard binds you through the customer's contract or the equipment listing, never on its own authority, and its humidity classes are stated as pressure dew points: the +3 C class, which is Class 4, is roughly what a refrigerated dryer targets, and the -40 C class, Class 2, is desiccant territory. Specifications are written as the class number, so confirm both the number and the edition before quoting against one.

Two consequences fall straight out of the gate. A dryer sized for an indoor header is the wrong dryer the day someone runs a branch out to a yard crane. And a dryer that meets its own nameplate rating can still fail the gate, because the nameplate says nothing about your pipe.

Step 1: Set the requirement before you measure anything

Walk the pressurised pipe, all of it, and find the coldest metal. Unheated loading docks, the last 20 feet through an exterior wall, a roof run, a line into a walk-in, the branch nobody remembers that feeds the compactor outside. Take the temperature at the coldest point in the coldest week you can get data for, or take it from the building's own recorded low if the pipe is exposed to ambient.

Skipping this step is what turns the whole exercise into theatre. Without a requirement, any measured number looks acceptable, and the tech ends up comparing the outlet reading to the dryer's nameplate, which only proves the machine is doing what it was sold to do.

Step 2: Measure at line pressure, never on an expanded sample

Expanding the sample to atmosphere lowers its dew point substantially, because the same water is now spread through a much larger volume. A hygrometer sniffing a vented sample therefore reads far drier than the line actually is, and the error runs in the flattering direction, so it never gets questioned.

Put the probe in a sample cell held at full line pressure with the flow set by a needle valve downstream of the cell, not upstream. If site conditions force an expanded sample, convert the reading back to line pressure using the conversion chart from the instrument manufacturer, and write on the report which pressure the stated number belongs to.

Use stainless or PTFE sample tubing. Rubber and most flexible plastics hold water in the wall and release it slowly, so a dry line reads wet for hours and a wet line reads dry after the hose has been sitting. Give the probe its full settling time from the data sheet before recording; on a dry measurement that can run to tens of minutes, and a reading taken at five minutes on a slow probe is the most common way a good instrument produces a bad number.

Step 3: Load it and time it

Dryer capacity is published at reference conditions, typically a stated inlet pressure, a stated inlet air temperature and a stated ambient, with correction factors for anything else. Three things push the outlet dew point up: more flow, hotter inlet air, hotter ambient around the dryer. All three peak in the same hour, which is a summer afternoon at shift change with every tool running.

So measure at that hour, at full plant load. At the other end of the range, a lightly loaded refrigerated dryer on a cool morning will read at or below its rating, which is why a morning reading proves nothing about August.

Step 4: Give the number its basis and its character

Before the reading enters any comparison, write down two things about the instrument's error.

Its basis. Percent of reading, percent of span, or a fixed figure in degrees. A dew point sensor's error is usually quoted as a fixed number of degrees over a stated range, which means it does not shrink as the reading gets drier.

Its character. A specification of the form "plus or minus 4 F over this range" from a data sheet is a worst-case bound, not a measured spread. Bounds are reported as bounds and, where two instruments are involved, they add linearly. Independent random spreads are the other case and they combine in quadrature, so two equal spreads multiply by the square root of two rather than by two.

One useful consequence: if you take inlet and outlet with the same probe, the probe's fixed offset is common to both readings and cancels in the difference, so the inlet-minus-outlet figure is far more trustworthy than either absolute number. The absolute outlet reading, which is what the gate needs, carries the full bound.

Step 5: One gate, two systems

System A. Refrigerated dryer, nameplate 38 F PDP at its reference conditions, panel light green. Measured at the dryer outlet at line pressure under full load on a hot afternoon: 41 F. Pipe walk finds the coldest pressurised metal at a loading-dock header, measured at 28 F on a winter morning. Required PDP is 28 minus the 18 F margin, so 10 F. Measured 41 F fails by 31 F. Apply the sensor's plus or minus 4 F bound at its most favourable edge and the reading becomes 37 F, still 27 F above the requirement. The verdict does not depend on the instrument at all, which is exactly what you want a gate to do.

The finding is not "the dryer is broken." The dryer is meeting its rating within the bound. The finding is that this plant needs either a dryer class that reaches the required dew point on the branches that leave conditioned space, or that dock branch fed from a small point-of-use dryer, or the branch heat-traced and insulated. Three options, and the customer picks on cost and layout, not on the tech's preference.

System B. Same class of dryer, same nameplate 38 F. The machine looks sick: condenser coil matted with lint, discharge air warm at the outlet, refrigerant suction reading high on the panel gauge. All of those point the same way, which is why they are consistent: a fouled condenser raises condensing pressure, raises evaporator temperature and therefore raises outlet dew point. Measured at the worst hour under full load: 44 F, six degrees above nameplate, which is the degradation showing up honestly.

Every pressurised line in this plant is inside a heated production space, and the coldest recorded metal temperature on the pipe is 66 F. Required PDP is 66 minus 18, so 48 F. Measured 44 F passes. Apply the same plus or minus 4 F bound at its unfavourable edge and the reading becomes 48 F, sitting exactly on the requirement with nothing left over.

So System B passes the gate today and has no margin, which is a different work order from System A's. Clean the condenser, restore the airflow path, re-measure at the worst hour, and expect the number to fall back toward the nameplate. Schedule it before summer rather than dispatching it tonight.

The two systems resolve opposite ways, and the machine that looks worse is the one that is currently delivering acceptable air. That is the whole argument for measuring the air instead of inspecting the box.

How to verify you got this right

Re-read your own report and check four things. That the stated dew point names the pressure it belongs to. That the reading was taken under load, with the load and the ambient written next to it. That the probe's settling time from its data sheet was actually allowed, and the reading was stable, not still falling when you wrote it down. And that the requirement on the page came from a measured pipe temperature, not from the dryer's nameplate.

The failure mode to catch in your own work: a number recorded before the probe settled, on an expanded sample, at 9 am, compared against the nameplate. Every one of those four errors pushes the verdict in the same direction, toward "the dryer is fine," which is why they stack up so reliably into a clean-looking report and a callback in six weeks.

References

• 29 CFR 1910.147, control of hazardous energy, for isolating and verifying depressurisation of a pressurised air path before opening it
• 29 CFR 1910.333(b)(2) for electrical isolation of dryer and compressor control equipment, with 29 CFR 1926.417 as the construction counterpart, and NFPA 70E-2021, 120.5, in the edition your employer's electrical safety program has adopted, for the live-dead-live proving sequence
• 29 CFR 1910.242(b), which limits compressed air used for cleaning to under 30 psi with effective chip guarding and personal protective equipment
• ISO 8573-1:2010 purity classes, which bind through a customer specification, a contract or an equipment listing rather than on their own authority; dryer manufacturer data for rating conditions, correction factors and sensor settling time
• See related: Why a Refrigerated Dryer Fails Quietly; What a Desiccant Dryer Costs You in Purge Air