How to Field-Check a Pressure Gauge Against a Test Gauge
Dial pressure gauges do real work: operators make lineup decisions from them, and on plenty of skids they are the only local indication there is. A gauge that reads wrong gets people hurt or equipment misoperated, and gauges drift, take overpressure hits, and wear their movements. A field check against a known-good test gauge is the quick, honest way to decide whether a working gauge still deserves trust. This page covers picking the reference, running the comparison, and judging the result.
Field-Check a Pressure Gauge in one line: To field-check a pressure gauge against a test gauge, connect both to the same pressure at the same elevation, walk the pressure up through several points and back down, and compare readings at each point, tapping the working gauge lightly to reveal movement friction. The test gauge must have a current calibration and a suitable range so the check points land in the middle portion of its scale. Any isolation or venting on an in-service connection follows site procedures. Judge pass or fail against the tolerance the site assigns to that gauge's duty.
What You Need
The reference is everything: a test gauge with a current calibration record and an accuracy grade meaningfully better than the gauge being checked. Gauge accuracy grades are defined by the ASME B40.100 gauge standard, and test gauges are built and graded specifically for this duty. Range matters as much as grade - pick a test gauge whose range puts your comparison points in the middle portion of its scale, where a bourdon-tube instrument performs best, rather than pinned near either end.
For an offline check you also need a pressure source, typically a hand pump with a fitting tee so both gauges see the same pressure, or a comparator stand in the shop. For an in-service spot check, you need a test connection on the same line as the working gauge. In both cases, breaking into or venting any connection that has seen process pressure is governed by site isolation and venting procedures, and on liquid service the two gauges should sit at the same elevation so a column of liquid between them does not masquerade as gauge error. The basics of the instrument itself are covered in what a pressure gauge is.
Run the Comparison Up and Down
Start at zero: with both gauges vented, the working gauge's pointer should rest at the zero mark or within its stop tolerance. A gauge that no longer returns to zero has likely taken an overpressure set or has a damaged movement, and there is little point checking further before noting that. Then raise pressure smoothly to a series of points spread across the working gauge's scale, comparing readings at each point on the way up, and repeat the same points coming back down.
Tap the working gauge case lightly with a finger before reading each point. A healthy movement barely reacts; a worn or dirty movement jumps noticeably as friction releases, and the size of that jump is itself diagnostic. The difference between the rising and falling readings at the same point is the gauge's hysteresis, and a gauge that reads acceptably on the way up but hangs badly on the way down has a worn movement even if every individual reading squeaks inside tolerance.
Judging the Result
Compare each pair of readings and judge against the tolerance the site assigns for that gauge's duty, which follows from its accuracy grade and what decisions are made from it. A utility-air gauge and the gauge an operator uses to confirm isolation before a line break do not deserve the same acceptance band. Errors concentrated at one end of the scale, a consistent offset across all points, growing hysteresis, or a large friction jump each tell you something different: a set from overpressure, a slipped pointer or linkage, and a worn movement respectively.
Record what you found, even for a passing gauge - the as-found record is what turns the next check into a trend rather than a coin flip. A gauge that fails belongs out of service, not "noted": a slipped pointer can sometimes be reset by a qualified shop, but a gauge with a set bourdon or a jumpy movement is a replacement, and dial gauges are cheap compared to the decisions made from them.
Common Mistakes
The classic errors are reference errors. Checking a gauge against another working gauge of unknown history proves nothing except that two gauges agree or disagree. Using a test gauge whose range is far too large puts your comparison points down in the worst part of its scale, where its own error can swallow the error you are looking for. And forgetting elevation on liquid-filled lines introduces a fixed head offset that gets blamed on the gauge.
The other family is technique: reading at an angle and taking a parallax error, comparing during pulsating pressure where the two gauges' dynamics differ, checking only rising points and missing hysteresis, and skipping the zero check that would have condemned the gauge in seconds. On pulsating service, check with the process settled or through a snubbed connection, and remember that a gauge that flutters constantly in service is telling you it needs a snubber or a liquid fill regardless of its calibration.
Frequently Asked Questions
What makes a gauge a test gauge?
A test gauge is built and graded to a tighter accuracy class than ordinary process gauges, per the ASME B40.100 gauge standard, and it carries a current calibration traceable to a reference. It typically has a larger dial with a finer scale and often a knife-edge pointer or mirror band to kill parallax. The point is that its error is small enough, relative to the gauge being checked, that a disagreement can be pinned on the working gauge.
How many points should a field check cover?
Enough to see the shape of the error, not just its size at one spot: zero, plus several points spread across the scale, read both rising and falling. A single mid-scale comparison misses zero set, span error concentrated at the top of the range, and hysteresis entirely. If the gauge guards one critical pressure, make sure one comparison point lands at or near that value, since that is the reading decisions are made from.
When should a working pressure gauge be replaced instead of adjusted?
Replace it when the pointer no longer returns to zero when vented, when it shows a friction jump on tapping, when hysteresis between rising and falling readings has grown noticeable, or when the error is nonlinear across the scale. Those symptoms indicate a set bourdon tube or a worn movement, which adjustment does not cure. A simple slipped pointer can sometimes be corrected by a qualified shop, but for most process-grade gauges replacement is cheaper than repair.
Automation services
Need help turning this into a working system?
Merobix integrates SCADA, programs Allen-Bradley and Siemens PLCs, and designs and fabricates industrial control panels.
Meeting requests are reviewed before confirmation.