Calculating Pressure Sensor Error From a Trapped Bubble

The panel shows a pressure value even though an air bubble covers the sensing face. At steady state, the trapped air transmits the liquid pressure, so the reading is nearly the pressure at the gas-liquid interface—not the bubble's original 1 atm pressure. Any remaining difference comes from elevation across the bubble, surface tension, or the sensor installation; during pressure changes, gas compressibility can produce a much larger transient effect.

Read the Symptom Before Changing the Sensor

Start here. Decide whether the questionable value is a stable offset or a response problem. A trapped bubble behaves differently in those two cases.

Observed symptom	Likely mechanism	First check
Stable reading close to the expected pressure	The compressed gas is transmitting static pressure normally.	Compare the sensor and reference elevations.
Small stable offset that changes with sensor orientation	The bubble displaces a vertical height of liquid, or interface curvature changes.	Measure the vertical distance between the sensing face and gas-liquid interface.
Slow response during a pressure ramp	Gas compliance combines with a restricted pressure path or diaphragm displacement.	Compare response before and after venting the bubble.
More piston travel than expected before pressure rises	The piston first compresses trapped gas.	Track pressure against displaced liquid volume.
Pressure-volume behavior changes most at low pressure	A gas undergoes a larger fractional volume change at lower absolute pressure.	Plot volume change against absolute pressure, not gauge pressure.
Reading remains wrong after the bubble is removed	The fault lies in the sensor, reference, mounting, wiring, scaling, or process connection.	Perform a controlled pressure comparison.

A bubble covering the entire sensing surface does not shield that surface from pressure. Pressure acts through the gas. Replacing or recalibrating the sensor before purging and repeating a controlled comparison usually wastes time.

Follow the Static Pressure Through the Bubble

At mechanical equilibrium, the liquid applies pressure to the gas-liquid interface. The trapped gas compresses until its pressure balances the interface pressure, subject to surface tension and the small hydrostatic pressure change within the gas. The gas then applies that pressure to the sensing face.

If capillary forces, surface tension, wetting forces, and material-attraction forces are neglected, pressure is continuous across the interface at the same elevation. Across a bubble of vertical height h, the surrounding liquid pressure changes by:

ΔP = ρgh

Here, ρ is liquid density and g is gravitational acceleration. That elevation term—not the fact that the medium touching the diaphragm is air—sets the basic static difference.

Suppose the gas-liquid interface is above the sensing face. The gas pressure changes very little from the interface to the face because gas density is much lower than liquid density. An undisturbed liquid at the sensing-face elevation would gain approximately ρgh below the interface. The bubble-covered sensor can therefore read lower than that same-elevation liquid pressure by approximately ρgh, before surface-tension effects are included.

A very small curved interface can also sustain a pressure jump caused by surface tension. The jump depends on interface curvature, liquid surface tension, contact angle, and the geometry of the sensing cavity. Read those properties from the liquid data and sensor construction; bubble volume alone does not define the correction.

Calculate the Bubble Compression With Absolute Pressure

Use absolute pressure for gas compression. For a fixed quantity of gas compressed slowly enough to remain approximately isothermal, the first calculation is Boyle's law:

P1V1 = P2V2

V2 = V1 × P1 / P2

For the stated example, take an initial air volume of 1 mm³ at approximately 15 psia. Two interpretations of the final 1000 psi produce slightly different results:

• If 1000 psi means 1000 psia: V2 = 1 mm³ × 15/1000 = 0.015 mm³.
• If 1000 psi means 1000 psig: using the same approximate atmospheric pressure gives P2 ≈ 1015 psia, so V2 ≈ 1 mm³ × 15/1015 = 0.0148 mm³.

The bubble becomes much smaller, but its pressure rises with the surrounding liquid. It does not remain at 1 atm. Temperature change, gas dissolution, leakage, nonideal gas behavior, and cavity geometry can move the actual volume away from this simple result.

Do not use the same calculation for a water-vapor cavity without checking phase equilibrium. The stated approximate vapor pressure is 0.45 psia at normal temperatures. When local absolute pressure rises above the applicable saturation pressure, water vapor tends to condense and the cavity can collapse; when local pressure falls sufficiently, vapor formation can occur. Use actual liquid temperature to obtain the applicable vapor pressure.

Separate Static Accuracy From Dynamic Response

A trapped air pocket can transmit the correct final pressure and still disrupt the way the system reaches it. Compressed gas stores volume elastically. That added compliance changes the pressure-volume relationship of the sensor cavity or hydraulic circuit.

In a piston-driven hydraulic system, piston displacement initially compresses the bubble as well as pressurizing the liquid and expanding other compliant components. More displacement may be required to achieve a given pressure rise. The effect is proportionally greater at low absolute pressure because the gas experiences a larger fractional volume change there. At high pressure, the remaining gas volume can become very small, although its stored energy still matters during a release.

Compliance alone does not define a time delay. A measurable lag requires interaction with flow resistance, such as a narrow pressure port, capillary, valve restriction, viscous passage, or another limited flow path. Together, resistance and gas compliance can slow response, filter pulsations, or alter oscillation. Read the sensor response specification and compare it with an air-free test; do not assign a time constant without the cavity volume and flow-path characteristics.

Dissolved air is different from a visible bubble. At constant temperature and stable pressure it may remain dissolved, but pressure reduction can bring gas out of solution. Repeated pressure cycling can therefore make a bubble disappear at high pressure and return after depressurization.

Diagnose the Installation in the Right Order

1. Classify the symptom. Record whether the error is static, appears only during ramps, follows pressure cycling, or changes with sensor orientation.
2. Identify the pressure basis. Mark every value as absolute or gauge. Convert to absolute pressure before calculating gas compression.
3. Match elevations. Locate the sensing face, gas-liquid interface, and reference gauge. Correct liquid-column differences with ΔP = ρgh.
4. Inspect the pressure path. Look for cavities, downward-facing ports, restrictions, dead legs, and high points that can retain gas. Estimate the bubble's vertical height as well as its apparent area.
5. Establish a reference. Connect a suitable reference instrument to the same pressure source and, where practical, at the same elevation. A remote gauge at another height includes its own liquid-head term.
6. Record an as-found test. Hold several stable pressures and record both instruments. Then apply controlled increasing and decreasing pressure ramps to expose lag or hysteresis.
7. Vent or purge the cavity. Use the equipment's approved filling and venting method. Orient the cavity so gas travels toward the vent instead of remaining against the sensing face.
8. Repeat the identical test. Use the same reference, pressure points, ramp direction, temperature, and mounting orientation. The before-and-after change isolates the bubble's contribution.

Do not diagnose from a single reading taken while pressure is moving. That mixes static offset, dynamic lag, reference response, and pressure-source behavior into one number.

Verify the Correction Under Static and Changing Pressure

Verify static performance first. Hold each test point until both instruments stop changing, correct for elevation, and compare the result with the sensor manufacturer's accuracy specification and the process tolerance. A stable agreement after purging shows that the pressure path and scaling are functioning at those points.

Next, repeat the operating pressure ramp. Compare rise time, settling behavior, and increasing-versus-decreasing traces before and after purging. A large dynamic improvement with little change in final readings identifies gas compliance as the main mechanism.

For a piston-driven system, plot applied pressure against piston displacement or admitted liquid volume. Trapped gas appears as added volume demand, especially at lower pressure. After effective venting, less displacement should be required for the same pressure rise, subject to hose expansion, seal motion, liquid compressibility, and structural elasticity.

Depressurize and repeat the test if the process normally cycles. A bubble that returns points to retained gas elsewhere, air ingress, or gas coming out of solution. Inspect the fill route and vent locations before disturbing the calibration.

Avoid the Recurring Pressure-Test Pitfalls

• Using gauge pressure in Boyle's law: Gas volume ratios require absolute pressure. A calculation that uses zero gauge pressure as zero absolute pressure is invalid.
• Treating bubble area as the pressure error: Covering the entire diaphragm changes the contacting medium, not the equilibrium pressure. Vertical height and interface forces control the static difference.
• Equating a small compressed volume with zero risk: Gas stores recoverable energy. A high-pressure failure can release that energy violently even when the visible bubble is small.
• Using gas when a hydrostatic test is intended: Remove trapped air as far as the approved procedure permits. A pressure hold alone does not show how much compressible gas remains.
• Ignoring the low-pressure portion of a test: Relative gas-volume change is greatest there. Pressure-volume traces often reveal trapped air more clearly at the beginning of pressurization.
• Confusing air with water vapor: Air compresses and may dissolve; vapor can condense or form according to local absolute pressure and temperature.
• Changing several variables at once: Keep reference location, orientation, temperature, pressure points, and ramp method fixed when comparing the vented and unvented states.

Frequently Asked Questions