On this page6 sections
Many pressure transmitter problems in the field turn out to be selection problems rather than accuracy problems: a sensor ranged too wide to resolve the operating point, a diaphragm that corrodes, a fill fluid that vaporises under vacuum, an enclosure that lets water in, or an output the control system does not expect. Each is avoidable if the right questions are asked before the order is placed.
Start with the process data
Collect the following on an instrument data sheet before comparing products. Every item either rules options out or changes the model code.
- Medium and phase — liquid, gas, steam or slurry; corrosive components; solids; any tendency to crystallise, polymerise or coat.
- Pressure — minimum, normal and maximum operating pressure; design pressure; vacuum; pulsation; and events such as hydrostatic testing, water hammer or pump dead-head.
- Reference — gauge, absolute or differential.
- Temperatures — process temperature including cleaning or steaming cycles, and the ambient range at the installation point.
- Purpose — indication, control, alarm, safety function or accounting. This decides how much accuracy, stability and diagnostic coverage are worth paying for.
- Interface — output signal and protocol, loop power, and the control-system input it connects to.
- Location — hazardous-area classification, exposure to water and dust, vibration, access and mounting.
- Process connection — thread, flange or hygienic fitting, and the size and rating of the tapping.
Range, turndown and the operating point
Three terms define the range. The upper range limit (URL) is the highest pressure the sensor module can be set to measure. The calibrated span runs from the lower range value (LRV, 4 mA) to the upper range value (URV, 20 mA). Turndown is the ratio of URL to calibrated span: a sensor with a 40 bar URL calibrated 0–10 bar is at 4:1.
Turndown is useful — one sensor module can cover several duties — but accuracy is usually stated as a percentage of calibrated span only up to a given turndown, with an additional term beyond it. Check the figure at the turndown you will use. A common practice is to place the normal operating point in the middle portion of the calibrated range, keeping headroom above it for upsets and alarms and resolution below it for start-up and low-load conditions.
| Condition | Pressure | Signal | Consequence |
|---|---|---|---|
| Normal operation | 6.5 bar g | 14.4 mA (65 %) | Operating point in the middle of the range, with resolution and headroom. |
| Maximum operating | 9.0 bar g | 18.4 mA (90 %) | Still inside the range, so high-pressure alarms remain meaningful. |
| Pump dead-head | 12 bar g | Saturates at 20.5 mA (≈ 10.3 bar g) with NE 43 limits | Not measurable. Widen the range if the event must be seen. |
| Hydrostatic test | 15 bar g | Not a measurement condition | Must be within the overpressure limit, or the transmitter isolated during the test. |
Widening the range to 0–16 bar g puts normal operation at about 41 % of span — still well resolved — and makes the dead-head event visible. Which choice is right depends on whether that event needs to be seen, which is a process question rather than an instrument one.
Overpressure, pulsation and vacuum
Datasheets separate the overpressure limit — the pressure the transmitter survives without damage or permanent shift — from the burst pressure, which only describes containment. Neither is a working limit. Compare the overpressure limit with every event the process can produce, not only the normal maximum.
- Pulsation and water hammer from reciprocating pumps, compressors or fast-closing valves can briefly exceed the steady maximum pressure. Mechanical protection — a snubber, restrictor or pulsation dampener — protects the sensor. Electronic damping only smooths the output signal; it does nothing for the diaphragm.
- Vacuum needs a sensor and fill fluid rated for it, particularly with diaphragm seals at elevated temperature, where the fill fluid can outgas or vaporise and the seal stops transmitting pressure correctly.
- Differential transmitters also need a static pressure rating above the line pressure and a one-sided overpressure rating that covers commissioning and blocked-impulse-line scenarios.
Wetted materials, fill fluids and diaphragm seals
The wetted parts — process connection and isolating diaphragm — must resist the medium at process temperature. Stainless steel 316L is the common default. Nickel alloys such as alloy C-276, tantalum and titanium are used for aggressive media, and gold-plated diaphragms where hydrogen could permeate a thin metal diaphragm. Chemical-resistance tables are a starting point; for unusual chemistry, confirm the material with the process owner.
The fill fluid behind the diaphragm matters as well. Silicone oil suits most duties; inert fluorinated fills are used for oxygen and chlorine service, together with cleaning for oxygen service; food-grade fills are chosen for hygienic processes. For steam, the usual protection is not a seal but a condensate-filled impulse line or siphon that keeps live steam temperature away from the sensor.
Diaphragm seals (remote seals) isolate the transmitter from the process. They are the usual answer when:
- the process temperature exceeds what the transmitter can tolerate;
- the medium is viscous, crystallising or a slurry, or would plug an impulse line;
- the process is hygienic and needs a flush, cleanable connection;
- the medium is corrosive enough that only the seal diaphragm should be exposed to it.
Caution
Seals change the error budget
The fill fluid expands with temperature, so seal systems add temperature error; long capillaries slow the response; and any height difference between seal and transmitter adds a fixed head that must be accounted for in the range. Use the largest practical diaphragm, the shortest capillary and direct mounting wherever the temperature allows.
Accuracy in service, not on the datasheet
Reference accuracy is measured under constant, ideal conditions. Installed performance adds the ambient temperature effect, the static pressure effect on DP transmitters, long-term stability between calibrations, mounting-position effects and the error of the input card that reads the signal. Combined as a total probable error — usually by root-sum-square — these give a realistic basis for comparing transmitters and for setting calibration intervals.
Check how each figure is expressed. Transmitters state accuracy as a percentage of calibrated span or of URL; pressure gauges are specified by accuracy class, as a percentage of the scale range. A specification that looks tighter may simply be referenced to a different base.
Match the accuracy to the purpose: an indication-only point does not need the performance of a control or accounting measurement. A transmitter in a safety instrumented function must also satisfy that function’s safety requirements specification under IEC 61511 — a separate selection step from accuracy.
Output, hazardous areas, enclosure and mounting
- Output
- 4–20 mA with HART is the common default: it suits most PLC and DCS analog inputs and adds remote configuration and diagnostics. Fieldbus or wireless variants make sense where the architecture already uses them. Confirm how the loop is powered and that enough voltage remains at the transmitter terminals.
- Hazardous area
- The zone or division, gas group and temperature class decide the protection concept — intrinsic safety (Ex ia or Ex ib) with barriers or isolators, or a flameproof (Ex d) enclosure — and the certification required, such as ATEX or IECEx. National approvals, such as PESO in India, may also apply.
- Ingress protection
- IP ratings to IEC 60529 describe specific tests: IP66 covers powerful water jets, IP67 temporary immersion, and one does not automatically include the other. Match the rating to the real exposure — washdown, rain, flooding — and close unused cable entries with rated plugs.
- Process connection
- Threaded (for example ½ NPT or G ½), flanged to the piping class, or a hygienic clamp or fitting for sanitary processes. It must suit the tapping, the pressure rating and the isolation and vent valves needed to remove the transmitter safely.
- Mounting
- Direct or bracket mounting, orientation, access to the display, and exposure to sun and vibration. For DP transmitters, specify the manifold together with the transmitter.
Engineering questions
- What information is required to select a pressure transmitter?
- The medium and its phase; minimum, normal and maximum operating pressure, design pressure and any vacuum, pulsation or test-pressure events; the pressure reference; process and ambient temperatures; the purpose of the measurement; the output and control-system interface; the hazardous-area classification; and the process connection and mounting.
- What is turndown on a pressure transmitter?
- Turndown is the ratio of the sensor’s upper range limit to the calibrated span. A sensor with a 40 bar upper range limit calibrated 0–10 bar is at 4:1. Accuracy is usually specified up to a certain turndown with an additional term beyond it, so check the figure at the turndown you intend to use.
- When is a diaphragm seal needed on a pressure transmitter?
- When the process is too hot for the transmitter, when the medium is viscous, crystallising, a slurry or would plug an impulse line, when a hygienic flush connection is required, or when the medium is corrosive enough that only the seal diaphragm should be wetted. Seals add temperature error and response time, so keep capillaries short.
- Does electronic damping protect a transmitter from pressure pulsation?
- No. Damping filters the output signal only. The sensor still experiences every pressure peak, so pulsation and water hammer need mechanical protection such as a snubber, restrictor or pulsation dampener, and an overpressure rating that covers the peaks.