In an open tank, the pressure at the bottom tells you how much liquid sits above the tap, and a simple gauge pressure transmitter reads level directly. Close the tank, or let it hold vapor above the liquid, and that trick stops working, because the vapor pushes down on the surface and adds its own pressure to the reading. Differential-pressure level measurement solves this by measuring the pressure at the bottom of the vessel and subtracting the pressure in the vapor space above it. What is left is the pressure due to the liquid column alone, which converts straight into level. The subtlety is in how you get the vapor-space pressure to the transmitter, which is where wet legs, dry legs, and density corrections all come from.
Differential-Pressure Level (Wet/Dry Leg) in one line: Differential-pressure (DP) level measurement determines liquid level in a closed or pressurized vessel by measuring the pressure at the bottom (the high side) and subtracting the vapor-space pressure at the top (the low side). The difference is the hydrostatic head of the liquid alone, which divided by the fluid's density gives the height. It relies on the reference leg connection and a known fluid density to stay accurate.
A DP level transmitter has two pressure ports, a high side and a low side, and it reports the difference between them. On a closed vessel, the high side connects to a tap near the bottom, where it feels the vapor pressure plus the head of liquid standing above the tap. The low side connects to a tap in the vapor space near the top, where it feels the vapor pressure alone. Because the transmitter outputs high minus low, the vapor pressure appears in both readings and cancels, leaving only the hydrostatic head of the liquid column. That is the whole idea: measure at both ends and let the shared vapor pressure subtract itself out.
Once you have the head, converting to level is basic hydrostatics. Head pressure equals density times gravitational acceleration times height, so height equals head divided by density and gravity. In practice the transmitter is scaled so that a given differential pressure maps to a given level, using the design fluid density and the geometry of the taps. If the fluid is water-like, the numbers are easy; if it is a hydrocarbon with a specific gravity well below one, the same head corresponds to a taller column, and the scaling reflects that.
In an open tank you do not need the low side at all, because atmospheric pressure sits above the liquid and vents to atmosphere on the transmitter reference too. The DP method really earns its place on closed and pressurized vessels: separators, treaters, boiler drums, and any tank running under a gas blanket. Anywhere the space above the liquid carries a pressure of its own, subtracting that pressure is the only way a head measurement can tell you level rather than a mixture of level and vessel pressure.
The low-side connection to the vapor space is called the reference leg, and how you handle it defines the whole installation. If the vapor is dry and non-condensing, you can run an empty impulse line from the top tap to the low side and let it sit full of gas. This is a dry leg, and it is simple, but it assumes the line stays empty. If the vapor is wet and will condense, that empty line slowly fills with liquid on its own, and the changing column corrupts the reading. The answer is a wet leg: you deliberately fill the reference line with a known liquid and account for its constant, known head in the calibration.
A wet leg adds a fixed pressure to the low side, which offsets the transmitter's zero. Calibration handles this with a zero elevation or suppression so that an empty vessel still reads zero level and a full vessel reads full scale. The catch is that the wet leg must stay full and its fluid density must stay constant. If the leg loses fluid, boils off, or freezes, the reference pressure changes and the level reading shifts even though the actual level has not moved. A great many DP level troubleshooting calls trace back to a reference leg that is no longer what the calibration assumed.
The high-side impulse line has its own failure modes worth naming. Plugging with wax, hydrates, or solids freezes the head signal so the level appears stuck. Trapped gas in a liquid-filled high line makes the reading read low and sluggish. These are mechanical problems in the tubing, not faults in the transmitter, and they are the reason remote seals exist for the harder services: they replace the vulnerable impulse lines with sealed, fluid-filled capillaries so there is nothing to plug or freeze.
The quiet weakness of every DP level measurement is that it assumes a fixed fluid density. The transmitter measures head, and it can only convert head to level using the density it was scaled for. When density actually changes, the level reading drifts even though the transmitter is working perfectly. Heat a tank of oil and it expands and lightens, so the same real level produces less head and the DP method reads low. Change the composition, let water settle under oil, or let temperature swing overnight, and the assumed density is wrong. This is not a calibration error; it is the physics of using pressure to infer height when the thing linking them is variable.
Because the failure is systematic rather than random, it is exactly the kind of thing a monitoring system can catch by watching trends. In a cloud SCADA platform such as Merobix, a DP level value is polled from the field alongside vessel temperature and pressure, timestamped, and stored. A level that drifts smoothly with the daily temperature cycle, while the operator swears nothing was pumped, is a density signature, and seeing that pattern on a trend beats chasing a phantom leak. Where the application demands it, a multi-transmitter hydrostatic setup measures density directly and corrects the level, and those corrected values flow into the same historian.
Remote and unmanned sites make historized DP level indispensable, because there is often no one standing at the vessel to notice a frozen reading or a lost wet leg. Alarming on a level that goes suspiciously flat, on a rate of change that is impossible for the process, or on a DP that swings with ambient temperature lets a cloud layer flag a sick measurement before it drives a bad control action or a false inventory number. The transmitter reports head; the monitoring system supplies the judgment about whether that head still means what the calibration says it does.
In a closed or pressurized vessel, the vapor above the liquid adds its own pressure to the reading at the bottom, so a single gauge transmitter reports level plus vessel pressure mixed together. A DP transmitter fixes this by also measuring the vapor-space pressure on its low side and subtracting it, leaving only the liquid head. That subtraction is why differential pressure, not simple gauge pressure, is the standard for closed tanks.
Both are the reference-leg connection from the vapor space to the transmitter's low side. A dry leg is an empty impulse line used when the vapor will not condense, while a wet leg is deliberately filled with liquid when the vapor is condensing so a known, constant head sits on the reference. The wet leg's fixed head is compensated in calibration, and problems arise if that leg loses fluid, boils, or freezes.
A DP transmitter measures hydrostatic head and converts it to level using a fixed assumed density. If the actual density changes, from temperature, composition, or water settling out, the same real level produces a different head, so the indicated level drifts even though the instrument is fine. Where density varies significantly, a density-compensated or multi-transmitter hydrostatic scheme is used to correct the reading.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.