A differential pressure transmitter does not touch the process directly; it senses pressure through impulse lines, the small-bore tubing that connects the process taps to the transmitter. When one of those lines plugs with hydrate, wax, scale, sediment, or ice, the transmitter is cut off from the real pressure and starts reporting a stale, damped, or dead value while every diagnostic still says the instrument is healthy. Because the number often looks plausible, a plugged line can masquerade as a calm process for a long time. This page explains how plugging corrupts the reading, why asymmetric plugging offsets rather than freezes it, and how statistical detection and SCADA flag a suspiciously quiet DP signal as a blockage.
Plugged Impulse Line in one line: A plugged DP transmitter impulse line is a blockage in the small-bore sensing tubing between a process tap and a differential pressure transmitter, caused by hydrate, wax, scale, sediment, or an ice plug. The blockage isolates the transmitter from the true process pressure, so a DP flow or level reading becomes stale, heavily damped, or frozen even though the transmitter itself reports as healthy. When only one of the two legs plugs, the reading is offset rather than simply frozen, and detecting the problem relies on noticing that the DP signal has stopped tracking and lost its normal noise rather than on any instrument fault alarm.
A DP transmitter measures the difference in pressure between two points, a high side and a low side, and it reaches those points through impulse lines filled with process fluid or a fill medium. Pressure changes at the tap travel down the line to the transmitter's sensing element, and the transmitter converts the differential into a flow or level output. The measurement is only as good as the hydraulic connection those lines provide: if a line is open and clear, the transmitter feels the real process pressure faithfully, and if a line is blocked, it does not.
Plugging chokes that connection. A hydrate plug in a wet gas line, wax or asphaltene deposition in a crude line, scale or sediment settling in the tubing, or an ice plug where the line is not adequately heat-traced can partially or fully block the bore. As the line narrows, pressure changes propagate more slowly and are damped, so the transmitter sees a sluggish, smoothed version of reality that lags the true process. When the line fully plugs, the transmitter is isolated and its reading freezes at whatever pressure was trapped, no longer responding to the process at all. In a DP flow application this looks like a flow that has gone flat; in a DP level application it looks like a level that has stopped moving.
What makes this dangerous is that the transmitter is not broken. Its electronics are fine, its self-diagnostics pass, and its output is a smooth, in-range value, so nothing in the ordinary instrument chain flags a problem. The failure is upstream of the transmitter, in the tubing, where no sensor is looking. Operators consequently see a healthy-looking number and have no direct indication that it is stale, which is exactly why a plugged impulse line can quietly deliver bad data for an extended period before anyone suspects the sensing line rather than the process.
Plugging does not always freeze the reading; when only one of the two legs blocks, the effect is a shift rather than a flatline, which is more insidious. A DP transmitter subtracts the low-side pressure from the high-side pressure, so both legs must track the process for the difference to be correct. If the high-side line partially plugs while the low side stays open, the high side responds sluggishly and the computed difference is biased; if a static head becomes trapped in one leg, the transmitter reads a constant offset because one side is stuck at a pressure that no longer reflects the process.
This asymmetric case produces a reading that is wrong but still moving, which defeats the simple intuition that a plugged line means a frozen number. A single-leg plug can make a DP level read consistently high or low, or a DP flow carry a persistent bias, while the signal still wiggles enough to look alive. Because the value is neither flat nor obviously erratic, it can pass casual inspection and even survive a comparison against expectations if the offset is modest. The trapped head in one leg effectively adds or removes a fixed pressure from the measurement, and unless you know the leg is compromised, you have no reason to subtract it back out.
Recognizing asymmetric plugging usually means comparing the DP reading against an independent expectation and noticing a bias that does not fit the process. A DP level that disagrees with a second level method, or a DP flow that no longer matches a downstream check, in a way that looks like a fixed offset rather than random error, is a candidate for a single-leg plug. The direction of the offset can even hint at which leg is affected, since blocking the high or low side biases the result in opposite directions. But the key mental shift is to treat a persistent, unexplained offset in a DP reading as a possible sensing-line problem, not automatically as a real process change or a transmitter that needs recalibration.
The most reliable tell of a plugging line is the disappearance of normal signal noise. A live process signal is never perfectly still, a real flow or level carries small fluctuations from turbulence, pump pulsation, and process dynamics, and a healthy DP reading shows a characteristic amount of that noise. As an impulse line begins to plug, the damping filters out those fluctuations, so the signal becomes unnaturally smooth before it fully freezes. Statistical plugged-line detection watches exactly this: it characterizes the normal noise and variability of the DP signal and raises a flag when that variability collapses, catching the blockage while the mean value still looks fine.
This is a case where a suspiciously quiet signal is the alarm. A DP reading that has gone flatter and steadier than the process should ever be is not evidence of a calm process, it is evidence that the transmitter has been hydraulically disconnected from the process. A monitoring system that understands this treats a sudden loss of signal noise, or a signal that stops tracking correlated variables it used to follow, as a plugged-line indication rather than as reassuring stability. Some transmitters even provide their own statistical process monitoring that computes these noise metrics internally and can report an impulse-line plugging alert.
Cloud SCADA is well placed to make this practical across many meters at once. When a platform like Merobix trends the DP signal and its statistical character over time, it can flag a transmitter whose signal has become abnormally quiet or has stopped moving with the variables it should track, distinguishing a genuinely stable process from a sensing line that is plugging. It can also cross-check a DP measurement against an independent one and alarm on a fixed offset consistent with single-leg plugging. Instead of an operator eventually noticing that a flow has been suspiciously steady for days, the system surfaces the loss of noise as a diagnostic, so maintenance can blow down or clear the impulse line, and add heat tracing or a purge where hydrate, wax, or ice is the recurring cause, before the frozen reading drives a bad decision.
The impulse line carries process pressure from the tap to the transmitter, so a blockage isolates the transmitter from the real process. As the line narrows, pressure changes are damped and the reading becomes sluggish and lags reality, and when the line fully plugs, the reading freezes at the trapped pressure and stops responding at all. Crucially, the transmitter itself remains healthy and in range, so nothing in the normal instrument chain flags that the value has gone stale.
A DP transmitter subtracts the low-side pressure from the high-side pressure, so both legs must track the process. If only one leg plugs, that side stops responding or traps a static head while the other side still follows the process, so the computed difference is biased rather than flat. The result is a reading that is wrong but still moving, which can look alive and pass casual inspection, making single-leg plugging harder to catch than a fully frozen signal.
The key indicator is the loss of the normal signal noise that a live process always carries. Statistical plugged-line detection characterizes the expected variability of a DP signal and flags when that variability collapses or the signal stops tracking variables it used to follow, catching the blockage while the mean value still looks plausible. A monitoring platform can trend this over many meters, treating a suspiciously quiet or frozen DP signal as a sensing-line problem rather than as a calm process.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.