When a pressure transmitter cannot sit right at the tapping, a remote seal moves the sensing diaphragm out to the process and pipes the pressure back to the transmitter through a slender oil-filled tube. That tube, the capillary, looks like a minor accessory but it shapes the measurement more than almost anything else in the assembly. Its length and bore set how fast the transmitter responds, its fill fluid reacts to ambient temperature and shifts the zero, and the height difference it spans creates a fixed offset that has to be corrected. This page focuses on the capillary itself: what it does, why temperature swings move the reading, and how matched capillaries keep a differential measurement honest.
Remote Seal Capillary in one line: A remote seal capillary is the thin, oil-filled tube that hydraulically connects a remote diaphragm seal to the transmitter body, transmitting process pressure through an incompressible fill fluid. Its length and bore affect response time, and because the fill fluid expands and contracts with ambient temperature, a long or exposed capillary causes a zero shift as conditions change. The height difference between the seal and the transmitter also creates a fixed head offset that is corrected during calibration, and on a differential install the two capillaries are matched so their temperature effects cancel.
A remote seal assembly is a closed hydraulic system. The process pushes on a flexible diaphragm at the seal, that push is transmitted through the fill fluid inside the capillary to the transmitter's sensing element, and the transmitter reads it. The fill fluid is chosen to be nearly incompressible so that a change in process pressure moves through the system faithfully, and the whole path from seal diaphragm to sensor must be completely filled and free of any trapped gas, because a bubble is compressible and would soften and delay the response. The capillary is simply the pipe carrying that hydraulic signal between the seal out at the process and the transmitter mounted somewhere more convenient.
The capillary's dimensions directly govern how quickly the transmitter reacts. A long capillary, and especially a narrow-bore one, offers more resistance to the small fluid movements that carry a pressure change, so the measurement responds more sluggishly. In viscous fill fluids or cold conditions the fill thickens and slows things further. This is why capillary runs are kept as short as the installation allows and why the bore and fill fluid are chosen with response requirements in mind: a needlessly long or thin capillary buys you a slow measurement that lags the process, which matters wherever the loop has to catch a fast change.
There is a related dynamic penalty in the diaphragm and fill volume the capillary has to move. Every pressure change has to flex the seal diaphragm and shuttle a tiny volume of fill fluid through the capillary, and the stiffer the diaphragm and the more restrictive the capillary, the more the assembly damps and delays the signal. The design goal is a system stiff enough to transmit pressure accurately but responsive enough to follow the process, and the capillary length and bore are central levers in that balance. A well-chosen capillary is the difference between a remote seal that reads promptly and one that always seems a step behind reality.
The capillary's biggest accuracy weakness is temperature. The fill fluid inside it expands when it warms and contracts when it cools, and because the system is a closed volume pushing on a diaphragm and a sensor, that thermal expansion changes the pressure the transmitter sees even when the process pressure has not moved at all. The result is a zero shift: the measurement drifts up or down as the ambient temperature around the capillary changes through the day and the seasons. The longer the capillary and the more of it exposed to swinging ambient conditions, the larger this thermal zero shift, which is why long, sun-exposed capillary runs are a known source of wandering pressure readings.
Separate from temperature, the geometry of the install creates a fixed head error. The fill fluid has weight, so if the seal sits at a different height than the transmitter, the column of fill fluid in the capillary adds or subtracts a constant pressure from what the transmitter reads, proportional to the height difference and the fill fluid's density. This head offset is not a fault, it is physics, and it is handled by correcting the transmitter's zero during calibration to account for the elevation between seal and transmitter. The important thing is to calibrate with the assembly in its installed orientation, because the head offset depends on exactly how the seal and transmitter end up mounted relative to each other.
The two effects behave differently and are managed differently. The head error is constant, so it is simply calibrated out once the geometry is fixed. The temperature-driven zero shift is variable, so it cannot be calibrated away with a single number; it is instead minimized by design, keeping capillaries short, choosing an appropriate fill fluid, and shading or insulating exposed runs so the fill fluid does not swing as widely. On tight low-pressure measurements the temperature effect can dominate the error budget, which is why capillary length and routing are treated as accuracy decisions and not just plumbing convenience.
On a differential-pressure measurement with remote seals on both the high and low sides, the temperature problem gets a clever solution: match the two capillaries. If both seals use capillaries of the same length, bore, and fill fluid, routed so they see the same ambient temperature, then when the fill fluid in both warms or cools, both sides shift by the same amount and the effect largely cancels in the difference the transmitter actually measures. This is why balanced, matched capillary systems are specified for differential remote-seal installs: the goal is not to eliminate the thermal expansion of each capillary but to make the two sides move together so their errors subtract to nearly nothing. Mismatched capillaries, or two capillaries routed through very different thermal environments, defeat that cancellation and reintroduce a temperature-dependent error into the differential.
Physically, the capillary is a delicate stainless tube that has to survive a field environment, so it is usually protected. Armored capillary, sheathed in a flexible metal conduit, guards the thin tube against crushing, kinking, and abrasion, because a kinked capillary restricts the fill fluid and a crushed one can be ruined. The routing also matters: capillaries are supported and coiled to avoid sharp bends and mechanical stress, kept away from heat sources that would exaggerate the temperature effect, and dressed so vibration does not fatigue them. A capillary is easy to damage and impossible to repair in place, so the installation practices around it, armoring, support, and gentle routing, are what keep the sealed hydraulic system intact for its service life.
In field operations, the capillary's temperature sensitivity produces a signature that monitoring can recognize. A remote-seal transmitter with a long exposed capillary tends to show a zero that breathes with the daily temperature cycle and drifts with the seasons, a slow oscillation on top of the real process signal. When these measurements feed a SCADA or cloud monitoring layer, that daily and seasonal pattern is visible against the history, so a platform such as Merobix can reveal a pressure reading that correlates with ambient temperature rather than with the process, pointing at the capillary's thermal effect rather than a genuine process change. Recognizing that pattern keeps the team from chasing a process ghost when the real story is a sun-baked capillary, and it flags installs where shading, insulation, or a matched-capillary correction would tighten the measurement.
The capillary carries the pressure signal as small movements of fill fluid between the seal diaphragm and the transmitter's sensor. A longer or narrower-bore capillary offers more resistance to that fluid movement, so a pressure change takes longer to reach the sensor and the measurement responds more slowly, an effect made worse by viscous fill fluid or cold temperatures. That is why capillary runs are kept as short as the installation allows when a fast, responsive measurement is needed.
The fill fluid inside the capillary expands when it warms and contracts when it cools, and because the system is a sealed volume pushing on a diaphragm and a sensor, that thermal expansion changes the pressure the transmitter reads even when the process pressure has not changed. The result is a zero that drifts with ambient temperature, and the effect is larger for long capillaries and for runs exposed to wide temperature swings. It is minimized by keeping capillaries short and shading or insulating exposed runs.
On a differential measurement with a remote seal on each side, matching the two capillaries in length, bore, and fill fluid, and routing them through the same temperature environment, makes both sides shift by the same amount when the temperature changes. Because the transmitter measures the difference between the two sides, equal shifts largely cancel, so the temperature error nearly disappears from the differential. Mismatched or differently routed capillaries break that cancellation and let a temperature-dependent error back into the measurement.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.