How to Handle Remote-Seal Temperature Effects

Back to Blog

Understand Where the Error Comes From

The capillary and the diaphragm cavity are packed with an incompressible fill fluid, and its job is to transmit process pressure faithfully to the sensor. The catch is that the fluid is only incompressible, not temperature-independent: as it warms it expands and its density drops, and as it cools it contracts. On the sensor side, a warmer, lighter column of fill fluid exerts a slightly different head than a cold one, so the transmitter reads a small pressure change that no process caused. This is the temperature effect, and it is inherent to the sealed system, not a defect.

The magnitude depends on the fill fluid's thermal expansion coefficient, the volume of fluid in the capillary, and how far the temperature swings, which is why long thin capillaries and large-swing environments show the effect most. The concept pages on the remote seal transmitter and its remote seal capillary cover the hardware; here the point is that every metre of fill fluid is a small thermometer feeding a false signal into your pressure reading.

Choose the Fill Fluid and Diaphragm for the Range

The single biggest lever is the fill fluid. Manufacturers offer several, trading thermal expansion against temperature range, viscosity, and process compatibility. A fluid with a lower expansion coefficient produces less temperature error, but the choice is bounded by the process: it must stay liquid across the full temperature range without boiling at the hot end or turning to sludge at the cold end. Selecting a fill fluid is a documented decision on the datasheet, and the tradeoffs are laid out in the note on the diaphragm seal fill fluid.

The diaphragm and mounting matter too. A larger, more flexible diaphragm needs less force to move, so it transmits pressure with less error, and a flush mount that sits at the vessel wall behaves differently from an extended one reaching into the process. The difference between flush and extended seal mounting changes both the process-temperature exposure and the response, so match the mount to whether the concern is a hot process or a swinging ambient.

Balance the Capillaries so Errors Cancel

On a two-seal differential arrangement, the temperature effect can be made to largely cancel if both legs are treated identically. Route the two capillaries together along the same path so they see the same ambient temperature, and specify them the same length and bore so equal temperature swings produce equal expansion on each leg. When both fill columns shift together, most of the effect subtracts out in the differential, leaving only the residual from any imbalance. Mismatched capillary lengths defeat this and leave a temperature-dependent zero shift.

For a single remote seal measuring gauge pressure or level against atmosphere, there is no second leg to cancel against, so the fill-fluid choice and keeping the capillary short carry the whole burden. Mount the transmitter close to the seal where practical, and avoid running the capillary through a spot that bakes in the sun on one side and sits in shade on the other. Where the process itself is hot, the process-side temperature drives its own contribution that the capillary routing cannot fix, so the fill-fluid selection has to cover it.

Verify Temperature Behaviour After Commissioning

You cannot bench out a temperature effect, but you can confirm it is within expectation once the seal system is in service. Note the transmitter reading at a stable known level early on a cold morning and again on a hot afternoon at the same physical level, and see how far the zero moves with ambient. A modest shift is inherent; a large one points to a mismatched capillary, a poorly chosen fill fluid, or a partially gassed seal. Compare against a hand dip at each temperature so you separate a real level change from a thermal one.

This is where a continuous trend is genuinely diagnostic. When the level tag feeds a monitoring history, a temperature effect shows up as a daily or seasonal sinusoid in the reading that correlates with ambient temperature rather than with any process event. Seeing that pattern lets you distinguish a seal temperature artifact from a real level swing, and quantify whether it is small enough to live with or large enough to justify a different fill fluid. The trend turns an invisible effect into a measurable one.

Avoid the Common Mistakes

The frequent errors are treating a temperature shift as a calibration problem and re-zeroing it away, which only moves the error to a different temperature, and specifying mismatched capillaries on a differential seal so the two legs never cancel. Running one capillary in sun and the other in shade, or leaving a long excess coil of capillary exposed, amplifies the effect. And ignoring the process-side contribution when the process is hot leaves a residual the ambient balancing cannot touch.

A partially filled or gassed seal makes everything worse, because a compressible bubble in the fill both slows response and exaggerates the thermal shift. If the temperature behaviour is worse than the datasheet predicts, suspect the fill before blaming the environment. Watching the trend against ambient temperature over a full daily cycle is the cheapest way to confirm whether the seal is behaving as designed or has developed a fault that a fresh fill or a re-terminated capillary would fix.

Frequently Asked Questions