Pressure Measurement, References & Transmitters

Pressure is normal force per area,

$$p=\frac{F_\perp}{A},$$

but a transmitter never measures “pressure” without a reference. It measures the deflection or stress caused by a difference between process pressure and a sealed, atmospheric, or second process reference. The reference choice determines both meaning and installation failure modes.

Name the reference #

The relation is

$$p_g=p_{abs}-p_{atm}, \qquad \Delta p=p_H-p_L.$$

A sealed-gauge device uses a trapped nominal atmospheric reference, so it does not follow current atmosphere exactly. A compound gauge spans vacuum and positive gauge pressure. Always state whether a range such as “0–10 bar” is absolute, gauge, sealed gauge, or differential.

The sensing chain #

Process pressure acts on a diaphragm, bellows, capsule, or tube. Displacement or stress is converted electrically, commonly by piezoresistive strain elements or capacitance. Signal conditioning excites the element, linearizes it, compensates temperature, digitizes the result, and maps it to an output such as 4–20 mA or a fieldbus value.

flowchart TB
  P["Process pressure"] --> I["Connection, impulse line or seal"]
  I --> D["Sensing diaphragm"]
  D --> T["Electrical transduction"]
  T --> C["Compensation and calibration"]
  C --> O["Current loop or digital value"]

The installed mechanical path is part of the sensor. It can add static head, trapped gas or liquid, temperature gradients, damping, resonance, leakage, plugging, and phase-state errors.

Select range from all pressures #

Distinguish calibrated span from survival limits. A DP transmitter may measure 10 kPa while both ports sit at 10 MPa static pressure. It must tolerate common-mode static pressure, one-sided overpressure, reverse differential pressure, pressure cycling, temperature, and process compatibility.

Rangeability lets one sensor be configured over smaller spans, but it cannot remove absolute zero error, noise, thermal effects, installation uncertainty, or limited resolution. The Emerson overview emphasizes range, accuracy, stability, and reliability; each must be read under its stated reference conditions and time interval.

Check wetted materials, fill fluid, gasket, process connection, diaphragm thickness, response time, hazardous-area approval, ingress, vibration, and cleaning. Remote seals isolate corrosive, sanitary, hot, viscous, or plugging media, but capillary length and temperature create extra dynamics and zero shift.

Impulse-line installation #

For liquids, keep both legs liquid-filled and avoid high-point gas pockets. For gases, avoid low-point liquid traps. Steam service commonly uses condensate legs so hot vapor does not reach the transmitter; the two legs must have equal, stable head. Slope lines continuously, minimize unnecessary volume, provide isolation/equalization/drain/vent functions, and mount so bubbles or sediment can be removed.

A three- or five-valve manifold supports isolation, equalization, venting, and calibration. Operating sequence matters: applying full process pressure to only one side can overload a sensitive DP cell.

DP as an indirect measurement #

Across a restriction, idealized incompressible flow scales approximately as

$$Q\propto\sqrt{\Delta p}.$$

Therefore a DP transmitter used for flow needs the correct square-root extraction, density treatment, low-flow cutoff policy, and primary-element coefficients. A linear 4–20 mA signal proportional to DP is not linearly proportional to flow.

For hydrostatic level in an open vessel,

$$h=\frac{p_g}{\rho g}.$$

Density change becomes level error. In a closed vessel, a DP connection to the vapor space cancels vessel pressure, while wet or dry reference-leg behavior must be modeled explicitly.

Calibrate the installed measurement #

Calibration compares indications with traceable reference pressures over the declared range, direction, environment, and configuration. NIST pressure and vacuum calibrations illustrate the traceability chain from standards to instruments. A useful sequence includes zero, multiple upscale and downscale points, repeats, and post-test zero.

Bench calibration proves the transmitter, not every installed effect. After mounting, verify zero with manifold state understood, leak-test connections, inspect impulse-line fill, confirm range and engineering-unit mapping, and run a loop check through the receiving system. Preserve as-found and as-left data; otherwise maintenance erases evidence of drift.

Diagnose by mechanism #

Pressure troubleshooting begins by separating process behavior, mechanical transmission, sensor conversion, configuration, and receiving-loop errors. Replacing the transmitter addresses only one of those layers.