Measuring Pressure and What the Reference Is

Why this matters

Every pressure number is a difference between two pressures. One of them is the thing you care about and the other is the reference, and the instrument almost never tells you which reference it used. Two techs can read the same system at the same instant with two correct instruments and write down two different numbers, both right.

Most of the time this costs nothing, because the reference difference is small against the reading. But the reference on the most common field instrument is the local atmosphere, and the local atmosphere moves. It falls with elevation and it drifts with weather. A number carried between sites, or compared against a table that assumed a different atmosphere, is a number whose zero changed underneath it. Where the work is vacuum, saturation, or any ratio of pressures, that shift is not a rounding error, it is the whole answer.

Before a gauge goes on anything

Breaking into a pressurized system to fit a gauge is stored-energy work. Isolate under 29 CFR 1910.147, apply your lock and tag, relieve the pressure through a controlled path, and verify the system is at zero at the point you are about to open before a wrench turns. Verify at the point of work, not at a gauge somewhere else on the system, because a closed valve between the two makes the remote gauge a lie.

Do not vent refrigerant to relieve pressure. Knowingly venting most refrigerants during service, maintenance or disposal is prohibited under 40 CFR Part 82 Subpart F, which also carries the technician certification and recovery-equipment requirements. Recover it into certified equipment.

Match the gauge to the service before it touches a port. Three checks, all fast. Working pressure: a gauge taken past its rated range can fail, and a Bourdon-tube gauge failing under pressure throws its window and its movement outward, which is why gauges on hazardous service carry blowout backs and why you stand out of the plane of the dial when pressurizing. Fluid compatibility: the wetted parts have to suit what is in the line, and the gauge manufacturer's data sheet owns that. Oxygen service: a gauge for oxygen must be cleaned and marked for oxygen service, and a gauge that has ever seen hydrocarbon oil never goes on oxygen, because hydrocarbon in an oxygen-enriched stream ignites.

Three zeros

The reference is the zero. Naming the zero is the whole discipline.

Absolute. Zero is a perfect vacuum. It never moves, anywhere on earth, in any weather. Absolute pressure is what physics uses, so any relationship that comes out of physics rather than out of a habit is an absolute relationship.

Gauge. Zero is the local atmosphere at the moment of reading. A gauge reading of zero does not mean nothing is there, it means what is there matches what is outside. This is the reference on nearly every dial gauge, gauge manifold and pressure transducer a field tech touches, and it is the one that moves.

Differential. Zero is the pressure at a second point that you choose, usually elsewhere in the same system. This is what a manometer across a filter, a coil, or a heat exchanger is doing, and the number is only as defined as the second port's location. The sibling card on why a gauge reads differently than a manometer owns that instrument comparison.

The conversion between the first two is one line: absolute equals gauge plus local atmospheric pressure. Standard atmosphere at sea level is 14.696 psi, which is 29.92 inches of mercury.

The zero that moves, and by how much

Elevation. Atmospheric pressure falls with height. At about 1,000 feet, standard atmospheric pressure is roughly 14.2 psia; at about 5,000 feet, roughly 12.2 psia. Those are standard-atmosphere values and the actual local pressure on a given day sits around them.

Weather. At one fixed location, day-to-day barometric variation is commonly within about 1 inch of mercury, which is roughly 0.5 psi, with storm systems pushing wider than that.

So the gauge zero at your job can differ from the gauge zero at the job you did last month by a couple of psi from elevation and by a fraction of a psi from weather, without any instrument being wrong. Elevation is the larger and more predictable of the two, and it is the one that follows your trucks between valleys and ridges in the same service area.

When the moving zero does not matter

Most of the time. If you are reading 200 psig and comparing it against a nameplate or a manufacturer's operating range that is also expressed in psig, a 2.5 psi shift in the underlying atmosphere is about 1 percent of the reading and it affects your number and the specification the same way. Do not add complexity where the physics does not ask for it.

The rule for when to bother is simple: convert to absolute whenever the calculation divides two pressures, whenever the quantity is defined against vacuum, or whenever you are using a relationship that came from a property table rather than from a nameplate. Those three classes are where the moving zero decides the answer, and they get the rest of this card.

Worked example: one compressor, two elevations

The same model of equipment, in the same condition, running the same duty, installed at a job near sea level and at a job at about 5,000 feet. The question is compression ratio, which is the discharge absolute pressure divided by the suction absolute pressure, and which is used to judge whether a compressor is working within a sensible range. All pressures below are illustrative and stand in for readings on your own equipment.

Both machines read the same on the gauges: suction 5 psig, discharge 200 psig.

At the sea-level job, taking local atmospheric as 14.7 psia:

• Absolute suction: 5 plus 14.7 is 19.7 psia.
• Absolute discharge: 200 plus 14.7 is 214.7 psia.
• Compression ratio: 214.7 divided by 19.7 is about 10.9.

At the 5,000-foot job, taking local atmospheric as 12.2 psia:

• Absolute suction: 5 plus 12.2 is 17.2 psia.
• Absolute discharge: 200 plus 12.2 is 212.2 psia.
• Compression ratio: 212.2 divided by 17.2 is about 12.3.

Identical gauge readings, and the compression ratio at the higher site is about 13 percent greater. The suction side is where the elevation change bites, because 2.5 psi is a large share of a 19.7 psia suction and a trivial share of a 214.7 psia discharge, so the denominator shrinks much more than the numerator does. That is the mechanism, and it runs the same way in reverse: take the same machine down to sea level and the ratio falls, because the denominator grows.

What the tech who skips the conversion gets. Dividing the gauge numbers, 200 divided by 5, gives 40. That is not a compression ratio, it is a number with no physical meaning at all, and it is more than three times the true value at either site. It is also identical at both sites, so it hides the very difference the calculation exists to find.

Why the difference matters rather than being a curiosity. Compression ratio drives discharge temperature and volumetric efficiency, so the higher-elevation machine is working harder for the same gauge readings, running hotter, and moving less mass per revolution. A shop that judges both sites against one gauge-pressure habit will keep finding the high-elevation machines mysteriously hot.

What would change the answer. If the acceptance range you are comparing against is itself published in gauge pressure for a specific installation altitude, then converting only your reading and not the range is worse than not converting at all, because you have corrected one side of a comparison and left the other uncorrected. Convert both, or compare both uncorrected and say so. And if the equipment manufacturer publishes an altitude-corrected operating range for that model, that range governs and this calculation is a cross-check on it rather than a replacement.

The vacuum case, where the gauge cannot express the question

Deep evacuation is defined against a perfect vacuum, so it is measured in microns of mercury absolute. A compound gauge on a manifold is referenced to the local atmosphere, and that is not a scaling problem, it is a range problem.

At sea level, atmospheric is about 29.92 inches of mercury, so the deepest vacuum a compound gauge can indicate is about 29.92 inches. A target of 500 microns absolute is 500 divided by 25,400, or about 0.02 inches of mercury absolute, which on that compound gauge is about 29.90 inches of vacuum. The needle sits hard against the end of its travel, indistinguishable from a system that is at 5,000 microns, or 20,000, or still full of moisture.

At 5,000 feet the same gauge cannot go past about 24.9 inches of vacuum, because that is all the atmosphere there is to work against. Local atmospheric of about 12.2 psia, scaled by 29.92 over 14.696, is about 24.9 inches of mercury. The needle pegs there and stays pegged while the absolute pressure in the system continues to fall through everything you care about.

This is why a micron gauge is not an upgrade, it is a different instrument answering a different question. It reads absolute, so its number means the same thing at both sites and no conversion is required. A compound gauge is fine for confirming that a system is under vacuum at all, and it is structurally incapable of confirming that a system is dry.

The same trap in the second class. A pressure-temperature relationship is a saturation property, which is a physics relationship, so it is fundamentally absolute even when the chart prints gauge pressures for convenience. A chart that printed gauge values assuming sea-level atmosphere reads slightly off at elevation. The size of that error is small in the middle of most charts and grows in the low-pressure region where the absolute pressure is small, which is exactly the region where the conversion is worth doing.

How to verify you got this right

• Say the zero out loud before you write the number. Absolute, gauge, or differential against what. If the answer is "the gauge said so", you have not named a reference.
• Check the units for the giveaway. psia, torr, microns and bar absolute name their reference. psig and inches of water column relative to room name theirs. Bare "psi" and bare "bar" name nothing and need chasing.
• Before dividing two pressures, convert both. A ratio of gauge pressures is not a ratio of anything physical.
• Compare like with like. If you corrected your reading to absolute, the specification you are judging it against gets the same treatment or you state in the same breath that it is uncorrected.
• Ask whether the instrument can even reach the answer. A compound gauge asked for a micron-level result is not inaccurate, it is out of range, and it will show you a confident needle position anyway.

References

• 29 CFR 1910.147 - lockout and the relief of residual stored energy before a pressurized system is opened
• 40 CFR Part 82 Subpart F - the refrigerant venting prohibition, technician certification and recovery-equipment requirements
• Gauge manufacturer's data sheet for pressure rating, wetted-material compatibility and oxygen-service cleaning
• Equipment manufacturer's documentation for altitude-corrected operating ranges where published
• See related: Relative vs Absolute Measurement; The Reference Point Every Measurement Needs; Why a Gauge Reads Differently Than a Manometer