How to Leak-Check Transmitter Impulse Lines
The impulse lines between a process tap and a transmitter are part of the measurement, and a leak in them is both a measurement error and a process release. On DP services a small leak on one leg quietly biases flow or level; on gas service it wastes product and can create a hazard around the instrument. Leak-checking impulse lines is cheap, fast, and routinely skipped. This page covers the soap-and-visual pass, the block-in pressure-hold test, and how to read the results.
Leak-Check Impulse Lines in one line: To leak-check transmitter impulse lines, work the fittings on gas service with a soap solution or gas detector and watch for bubbling or response, then use the manifold to block in the transmitter and watch its own reading: a blocked-in line holding pressure reads steady, while a leaking one decays toward atmospheric. The transmitter itself is the most sensitive leak detector on the line. All isolation and venting steps on live process follow site procedures, and any fluid the line contains dictates the precautions.
Why Impulse-Line Leaks Corrupt the Measurement
An impulse line transmits process pressure to the transmitter; a leak means the pressure arriving at the cell is no longer the pressure at the tap. On a single line the reading droops toward atmospheric by an amount that depends on the leak size and the tap's ability to keep the line packed. On a DP pair the damage is subtler and worse: a leak on one leg shifts the differential, biasing indicated flow or level in a direction that depends on which leg leaks, while both individual pressures still look plausible. A leaking wet-leg or condensate-filled line also slowly loses the liquid column its calibration assumes.
Leaks are also the sibling failure to plugging, and the two get confused because both produce readings that stop tracking the process honestly. A plugged line freezes or slows the reading; a leaking line biases it. The diagnostic thinking for the plugged case is covered in the plugged impulse line guide; this page is the procedure for proving the lines tight.
What You Need
For the external pass: a soap solution rated for leak detection, or a portable gas detector on hydrocarbon or sour service where sniffing beats bubbles, plus a mirror and light for fittings that face the wrong way. For the hold test: nothing more than the manifold, the transmitter's own indication in pressure units via local display or a communicator, and patience. The loop drawing matters too, so you know every fitting, tee, and valve that belongs to the measurement and none get skipped.
The safety framing is not optional. Impulse lines contain live process fluid at line pressure, and every isolation, equalization, venting, or draining move is a site-procedure and permit matter. On sour or toxic service, the gas detector and the site's exposure rules govern how you approach the work at all; a leak-check is precisely the activity that puts your hands and face near escaping process.
The Soap-and-Visual Pass
With the line at operating pressure, work methodically from the tap valve to the transmitter's process connections: every compression fitting, adapter, manifold joint, plug, and vent. On gas service, apply soap solution and watch each joint long enough for a slow leak to build a bubble; a fine mist leak shows as foam rather than discrete bubbles. On liquid service soap is useless - look instead for weeping, staining, salt tracks from evaporated product, and drips under fittings, all of which testify to a leak that may only weep when pressure or temperature cycles.
Do not skip the transmitter end. Vent plugs left loose after a previous calibration, drain screws, and manifold packing are the most common leak points because they are the parts people touch. A gas detector passed slowly around the assembly catches what the eye misses, and on any service where the process is hazardous the detector goes first, before your face does.
The Block-In Pressure-Hold Test
The most sensitive leak detector available is the transmitter itself. Following site procedures, block in the line at the manifold so the transmitter and the tubing between the block valve and the cell hold trapped pressure, and watch the reading in pressure units. A tight system holds essentially steady once temperature settles; a leak shows as a steady decay toward atmospheric, and the decay rate scales with the leak. On a DP pair, block in each leg and watch both the differential and, where available, the static reading, which localizes the leak to one leg.
Interpret with care: trapped pressure changes as trapped fluid changes temperature, so a line blocked in while hot will sag as it cools without any leak. Give the system time to settle, and judge sustained decay rather than the first minutes. When the hold test says leak but the soap pass found nothing, suspect the joints you could not soap - a buried fitting, valve packing, or the manifold's internal seat leaking past to the vent.
Verifying the Result and Common Mistakes
The lines pass when the external pass finds nothing and the blocked-in reading holds steady after thermal settling. Return the manifold to its operating lineup per procedure - block-and-bleed states left wrong after a leak check are their own category of incident - and confirm the transmitter reads process again. Where the point is trended in a monitoring platform such as Merobix, the recorded data gives a bonus check: a slow unexplained bias that had been growing for weeks should stop growing once a found leak is repaired, and that before-and-after is worth noting in the work record.
The recurring mistakes: overtightening compression fittings to "fix" a leak, which ruins the ferrule and guarantees the next one; mixing fitting brands and expecting the ferrules to seal; leak-checking only at commissioning and never after the tubing has seen vibration and thermal cycling; blaming the transmitter for a reading bias that is really a leg leak; and ignoring the manifold internals, where a passing seat quietly connects a leg to the vent.
Frequently Asked Questions
How do I tell an impulse-line leak from a plugged line?
A plugged line makes the reading sluggish or frozen: it stops responding to process changes and flatlines convincingly. A leak biases the reading instead: it still moves with the process but sits offset, drooping toward atmospheric on a single line or shifting the differential on a DP pair. The block-in hold test separates them cleanly, because a plugged line holds trapped pressure beautifully while a leaking one decays.
Can I leak-test liquid-service impulse lines with soap solution?
No - soap bubbles need escaping gas. On liquid service look for weeping at joints, staining, dried salt tracks, and drips, and rely on the block-in pressure-hold test, which works for any fluid: trapped pressure that decays after thermal settling means a leak. Remember that a liquid-filled line losing fluid also loses the hydrostatic column its calibration assumes, so a leak shows up as a growing level or flow bias too.
Which way does a leak on one leg of a DP transmitter bias the reading?
It depends which leg leaks. The leaking leg's pressure sags toward atmospheric, so a leak on the low-pressure leg raises the measured differential and a leak on the high-pressure leg lowers it. On flow service that reads as flow too high or too low respectively; on level service the direction also depends on which side carries the reference leg. The useful point is that a stable, unexplained bias on a DP measurement should put both legs on the leak-check list.
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.