Pressure Measurement
Almost every higher-level decision in a process plant depends on the integrity of a pressure signal. A transmitter’s real job is not to report pressure — it is to make physical process behaviour trustworthy enough for machines and people to act on.
Key points
A transmitter is a boundary, not a sensor. It is where physical conditions become information for control, optimisation, maintenance and safety.
The reference type decides what the number means. Gauge, absolute and differential answer genuinely different questions.
Accuracy is a system property. Datasheet reference accuracy and real installed uncertainty are not the same figure.
Installation can overwhelm sensor quality. A blocked impulse line produces a stable, believable, wrong reading.
A SIL-capable transmitter is not a safe loop. Certification is a component property; safety is a property of the whole function.
From process pressure to an automation decision
A pressure transmitter contains a sensing element plus electronics that condition the measurement and produce an electrical output. Compared with a basic transducer, a transmitter compensates, linearises and amplifies the signal before sending it to a control or monitoring system. Common outputs across our pressure sensors and switches range include 4–20 mA, HART and RS‑485 digital variants such as the PTS 310D RS.
In a conventional feedback loop the transmitter is the primary measurement element. The PLC or DCS compares its process variable against a setpoint, applies control logic, and commands a valve, drive or damper. The same value may also feed an operator display, an alarm system, a historian, a performance calculation or a maintenance application.
| Automation role | What the transmitter enables | Engineering priority |
|---|---|---|
| Monitoring and reporting | Remote indication, trends, historian records, inventory or energy calculations | Stability, useful range, data quality, maintainability |
| Feedback control | Continuous correction of valves, pumps, compressors or dampers | Response time, noise, damping, accuracy under operating conditions |
| Alarm and equipment protection | Detection of high pressure, low suction, filter plugging, loss of draft | Reliable thresholds, failure signalling, independent validation |
| Safety instrumented function | Initiation of a defined trip that moves the process to a safe state | IEC 61511 lifecycle, proof testing, diagnostics, independence |
| Derived measurement | Inference of flow, level, density or equipment condition | Correct physical model, compensation, primary-element condition |
This classification matters because the same hardware can be perfectly adequate for trending and unsuitable for fast control or risk reduction. HART’s digital channel, for example, provides device variables and diagnostics, but its ordinary two to three updates per second are generally too slow for closed-loop control from a central host. The analog channel therefore controls the process while digital communication supports maintenance.
Decision-ready insight: classify the consequence and response-time requirement of every loop before selecting the transmitter. A measurement used only for trending should not inherit the cost of a safety loop — and a safety-critical measurement must never be treated as an ordinary historian tag.
Gauge, absolute and differential solve different problems
The reference pressure defines the meaning of the number. A gauge transmitter compares the process with local atmospheric pressure. An absolute transmitter compares it with a vacuum reference. A differential transmitter measures the difference between its high- and low-pressure ports.
| Design | Reference and mechanism | Best-fit applications | Main limitation |
|---|---|---|---|
| Gauge pressure | Process pressure relative to atmosphere | Pump discharge, utility headers, compressed air, vented tanks | Atmospheric change matters when the true requirement is absolute |
| Absolute pressure | Process pressure relative to vacuum | Vacuum systems, gas calculations, barometric-independent processes | Usually unnecessary where only pressure above atmosphere is needed |
| Differential pressure | High-side minus low-side | Orifice flow, pressurised-vessel level, filter condition, draft, density | Two ports and two impulse paths add plugging, leakage and static-pressure effects |
| Flush diaphragm | Crevice-free wetted surface at the process connection | Viscous, pasty, crystallising or hygienic media | Process temperature and mechanical protection act directly on the diaphragm |
| Digital / precision | Digital sensor electronics with temperature compensation | Wide rangedown, tighter uncertainty, remote configuration | More configuration to control and document |
Differential pressure
The most versatile of the three. It infers liquid level from hydrostatic head, and flow from the pressure loss across a primary element. Our differential pressure transmitters cover both roles — the TDP 310 for general differential service and the TDP 340 where space is tight.
Straight pressure
Divides along process compatibility rather than principle. The PTC 310 and PTC 350 handle general industrial duty; the PTC 350F adds a flush hygienic connection for CIP and SIP cleaning; the PTS 330 covers high-pressure service; and the low pressure range, including the PTS 310, handles the small spans that draft and ventilation work demands.
Decision-ready insight: choose gauge, absolute or differential from the physical reference the process requires — never from price or availability. Use flush or remote-seal designs only when the reduction in plugging outweighs the added complexity.
Not sure which reference type fits your application?
Our engineers work through pressure and level selection with customers every day, including the awkward cases involving high temperature, aggressive media and very small spans.
Flow, level, pumps, filters and safety
Pressure earns its place because it is both a direct process variable and an indirect indicator of several others.
| Application | Arrangement | Automation value | Caveat |
|---|---|---|---|
| Pipe or vessel pressure | Gauge or absolute transmitter | Regulates pressure, detects overpressure or loss of supply | Select the correct reference and overpressure capability |
| Differential-pressure flow | DP transmitter across an orifice, nozzle or venturi | Flow input for control, allocation and energy balance | The relationship is nonlinear; primary-element condition matters |
| Hydrostatic level | Submersible probe in a vented tank, or DP on a pressurised vessel | Inventory, feed control, high and low alarms, shutdown logic | Density, vapour-space pressure and wet legs can bias inferred level |
| Filter or strainer condition | DP across inlet and outlet | Detects fouling; enables condition-based cleaning | Flow changes also change DP, so context is required |
| Pump and compressor protection | Suction, discharge or differential pressure | Detects loss of suction, abnormal head, blocked discharge | Pressure alone may not identify root cause |
| Furnace and ventilation draft | Low-range DP measurement | Stabilises combustion, protects against positive furnace pressure | Very small spans make routing and zero shift critical |
| Safety trip | Independent pressure input to an SIS | Initiates a defined action before pressure becomes unsafe | A certified transmitter is only one part of the function |
Choosing the level arrangement
A suspended probe such as the HLS 310 or HLC 310 measures hydrostatic head directly in the medium, which suits open tanks and sumps. A screw-in flush transmitter such as the HLS 330 mounts through the vessel wall instead, avoiding a cable in the process entirely. Aggressive or marine service pushes the choice again — toward the HLC 370 or the HLC 400. The full level sensor range sets these out side by side.
When a switch beats a transmitter
Where the requirement is a trip rather than a continuous value, an electronic pressure switch is usually the more direct answer. The EPS 200, EPS 205 and EPS 210 provide configurable switching outputs and set points with a local display. For spot checks and commissioning, a battery-powered gauge such as the DPG 200 or DPG 500 needs no loop at all.
A strong implementation separates control, alarm and safety objectives even when they observe the same physical variable. A control transmitter may be tuned for smooth response, while an independent trip device is configured and proof-tested for risk reduction. Reusing one connection or one device cuts hardware cost and creates a common point of failure at the same time.
Decision-ready insight: write a measurement purpose statement for every tag — physical quantity, inferred variable, expected range, process consequence, required response, failure action, and how the reading will be independently validated.
Accuracy beyond the datasheet
Selection begins with range but cannot end there. The upper and lower range limits describe what the sensor cell was built to measure; the configured range values define the calibrated operating range. Span is the difference between them, and turndown describes the relationship between the sensor’s capacity and the smallest usable calibrated span.
A worked example
Consider a cell rated to 10 bar configured to read 0–1 bar. Every fixed error term — zero uncertainty, temperature effect, static-pressure effect, calibration uncertainty — is now judged against a span one tenth the size, so each occupies ten times the share of the reading it did at full span. The device has not become worse; the measurement has become harder.
Reference accuracy quoted under laboratory conditions and application-level uncertainty in a plant answer two different questions. A transmitter capable of wide turndown can cover many ranges — but wide turndown does not mean constant percentage-of-reading accuracy.
| Selection dimension | Questions to answer | Common mistake |
|---|---|---|
| Reference and range | Gauge, absolute or DP? What are normal, startup, upset, vacuum and overpressure values? | Sizing only around normal operation |
| Total performance | What error results from accuracy, temperature, static pressure, turndown, mounting and calibration combined? | Comparing suppliers on one headline number |
| Dynamic behaviour | What response time and damping preserve stability without hiding a rapid upset? | Adding damping to suppress a mechanical problem |
| Wetted materials | Are diaphragm, seals, gaskets and fill fluid compatible with process and cleaning chemicals? | Checking only the housing material |
| Environment | What ambient temperature, vibration, ingress, EMC and hazardous-area conditions apply? | Treating an enclosure rating as universal suitability |
| Process connection | Direct mount, impulse line, manifold, seal or hygienic fitting? | Selecting the sensor before designing the interface |
| Lifecycle support | Are configuration tools, spares, certificates and diagnostics available? | Optimising purchase price rather than maintained cost |
Decision-ready insight: require a total-error calculation at the expected operating point, not at full span under reference conditions. Then select the narrowest sensor range that safely covers startup and upset without sacrificing overpressure margin.
Installation and calibration decide whether the number is true
Impulse lines are part of the measuring system. An excellent transmitter connected through badly routed tubing will produce a stable, believable and wrong value.
Keep impulse lines as short as the installation allows. Gas-filled lines should slope upward toward the transmitter and liquid-filled lines downward, with a self-draining or self-venting slope of at least 8%. An internal diameter of 6 mm is a practical minimum, and 10 mm is preferable wherever condensation or liberated gas is likely. The objective throughout is to prevent liquid pockets in gas service, gas pockets in liquid service, unequal heating, leaks, vibration damage and plugging.
Mounting position follows the same physical logic. For liquid service, mount beside or below the side taps. For gas, mount beside or above top or side taps. For steam, mount below the side taps and fill the lines with water. Manifolds should allow isolation, equalisation, venting, draining and safe calibration without exposing personnel or equipment to process pressure.
The failure calibration cannot fix
When an impulse line plugs, the normal small fluctuations in static and differential pressure fall away — which is precisely why fluctuation monitoring is a useful blockage diagnostic. It is also the clearest illustration of the problem: the sensor electronics can remain perfectly within calibration while the process interface has stopped delivering valid information.
Calibration alone is therefore not a maintenance strategy. You also need electronic diagnostics, line inspection or flushing where appropriate, comparison against process context, and confirmation that isolation and equalisation valves are correctly positioned.
A workable field routine
Review the service, range, alarm direction, isolation plan and acceptable tolerance.
Record an as-found test before making any adjustment.
Apply several pressure points across the operating span, increasing and decreasing.
Compare reference, transmitter output, control-system indication and alarm behaviour.
Adjust only where required, then repeat and record as-left results.
Restore valves, manifolds, damping, write protection and process connections.
Set the next interval from drift history, criticality, diagnostics and regulatory requirements.
Calibration itself should compare the device under test against a more accurate reference and document the result — a test accuracy ratio of at least 4:1 where practical, with total calibration uncertainty evaluated whenever that ratio cannot be achieved. The reference needs its own regular calibration and traceability to national standards.
Decision-ready insight: keep two records, not one — sensor calibration and complete-loop verification. The first tests measurement accuracy; only the second exposes wiring, scaling, input-card, logic, alarm and final-element errors.
Smart connectivity without over-reach
Digital communication turns a transmitter from a single-variable signal source into a configurable, diagnosable field asset. It does not turn every measurement into a candidate for it.
| Architecture | Mechanism | Best use | Limitation to manage |
|---|---|---|---|
| 4–20 mA | Continuous point-to-point current signal | Deterministic control and brownfield compatibility | Little contextual data without an added digital channel |
| HART over 4–20 mA | Frequency-shift keying rides on the analog loop without changing its average | Preserve analog control while adding configuration, status and diagnostics | About two to three digital updates per second; minimum 230 Ω loop impedance |
| RS‑485 digital | Multi-drop digital bus carrying the process value directly | Multiple devices on one cable run; digital precision end to end | Bus design, addressing, termination and host support |
| Wireless mesh | Self-organising network with managed scheduling | Retrofitting remote or stranded measurements | Batteries, radio survey, update rate, gateway dependency |
HART remains the lowest-disruption modernisation path: the analog signal carries the primary variable while digital communication carries secondary variables, configuration and status. Where the loop is HART-based, level instruments such as the HLC 400H integrate without changing the wiring approach. Where a genuinely digital bus is wanted, the RS‑485 variants such as the PTS 320D RS carry the value digitally from the sensor. Where the signal has to change form to reach an existing system, our signal converters handle the conversion, and the instrumentation range covers the displays, controllers and data loggers at the other end of the loop.
Connectivity also creates obligations. Operational technology security guidance treats monitoring and measurement devices as assets whose protection must preserve performance, reliability and safety. A smart-transmitter programme should inventory device identity and firmware, restrict configuration privileges, segment control networks, authenticate remote access, monitor unexpected changes, and plan patches around process risk.
Decision-ready insight: do not connect diagnostics simply because they exist. Define who will act on each alert before you enable it.
Functional safety and failure cases
IEC 61511 frames functional safety as a lifecycle covering specification, design, installation, operation and maintenance of a safety instrumented system. A transmitter’s SIL capability is a component property — not proof that the complete safety instrumented function achieves its required risk reduction.
Failure signalling has to be coordinated as well. Transmitters typically drive the current to a configured alarm direction on a detected fault, with standard and NAMUR NE 43 alarm and saturation thresholds differing from each other. If transmitter and analog input card use inconsistent thresholds, a genuine failure current can be read as a legitimate process value or as simple overrange.
| Safety question | Required decision |
|---|---|
| What hazardous event is being prevented? | Write a clear cause-and-effect specification and define the safe state |
| Is the transmitter independent of basic process control? | Remove shared taps, power, logic or maintenance dependencies where risk analysis requires it |
| What failures are detectable? | Map internal diagnostics, impulse-line failures, frozen values, wiring faults and input-card behaviour |
| What remains undetected? | Define proof-test procedures and intervals for dangerous hidden failures |
| Is redundancy effective? | Address common process connections, environment, configuration and maintenance error |
| Is the area classified? | Select equipment and installation methods appropriate to the hazardous-area classification |
The industry’s hardest lesson on this point came from a 2005 refinery explosion and fire in the United States that killed fifteen people and injured many more. Its value for measurement engineering is systemic rather than instrument-specific: catastrophic risk cannot be controlled by trusting a single indicated variable without independent safeguards, sound procedures, rationalised alarms, mechanical protection and an effective safety culture. The lesson is not that a transmitter caused a complex accident. It is to design so that one inaccurate, frozen, misinterpreted or unavailable measurement cannot defeat every layer of protection.
Decision-ready insight: never equate “SIL-capable transmitter” with “safe process”. Validate the entire function — process connection, failure direction, input thresholds, logic, final element, test coverage, bypass governance and human response.
Building the business case
The economic case comes from avoided installation work, fewer unnecessary field interventions, less bad-quality production, more stable control, earlier recognition of equipment problems and reduced probability of high-consequence events. These benefits are specific to each application, which is why a loop-level baseline is far more credible than a generic return percentage.
| Phase | Main work | Success measure |
|---|---|---|
| 1. Classify | Assign monitoring, control, protection or SIS role; document reference and consequence | Every tag has an approved measurement-purpose statement |
| 2. Baseline | Record bad actors, calibration history, trips, manual rounds, wiring cost, lost production | Benefits can be compared against a pre-project baseline |
| 3. Standardise | Limit approved ranges, materials, protocols, manifolds, seals and spares where practical | Lower engineering variation without forcing poor application fit |
| 4. Pilot | Pick a bounded, valuable problem — a remote pressure point, or recurring plugged lines | Demonstrated data quality and a named owner for each alert |
| 5. Integrate | Configure scaling, alarms, historian tags, diagnostics and security controls | End-to-end loop and failure-response tests pass |
| 6. Optimise | Use as-found history and diagnostic evidence to revise maintenance intervals | Less reactive work with no increase in failures or overdue proof tests |
| 7. Scale | Replicate only after documenting technical and economic results | Auditable lifecycle savings and sustained measurement availability |
A practical prioritisation score combines consequence of bad data, frequency of failure, accessibility, maintenance burden, control opportunity and ease of integration. High-consequence bad actors come before broad replacement of healthy instruments. For volume applications where the transmitter is built into a machine rather than a plant, our OEM pressure transmitters are designed around that different cost and integration profile.
Bringing it together
Three tensions define good decisions here. Integration increases information but also increases configuration effort and cyber exposure. Consolidation into multivariable devices reduces penetrations and hardware while increasing common dependency. Extended calibration intervals reduce maintenance, but only when as-found history, diagnostics, criticality and proof-test obligations all support the change.
The non-obvious conclusion is that the most advanced transmitter is not automatically the best transmitter. A correctly ranged, well-installed analog device on a verified loop can be worth more than a sophisticated digital instrument connected through a plugging impulse line. The reverse is also true: a smart transmitter’s diagnostics can expose process-interface degradation that routine calibration never catches.
A layered approach works in most plants
Get the pressure reference, range, material and process connection right.
Validate installed performance rather than relying on reference accuracy.
Separate control, alarm and safety responsibilities according to consequence.
Add digital diagnostics only alongside a process that assigns and closes alerts.
Treat calibration, proof testing, security and configuration control as lifecycle activities.
Scale modernisation according to measured loop-level value.
Talk through your application with our engineers
Browse the full pressure sensors and switches and level sensors ranges, or send us your application details and we will work through the selection with you.