- October 20, 2026
- Posted by: editor
- Category: RECD & Emission Control Devices
When a Retrofit Emission Control Device is specified, the monitoring section often contains a line like “pressure monitoring included.” That can mean very different things. It might be a continuous pressure transmitter that reports the exact differential pressure every second. It might be a pressure switch that does nothing until the pressure crosses a fixed setpoint, then closes a contact. It might be a simple mechanical gauge that someone has to walk over and read.
All three measure pressure, but they give very different information, fail in different ways and suit different jobs. This article explains how pressure sensors and pressure switches work in RECD monitoring, what each can and can’t tell you, and how to decide which you need.
What Pressure Is Being Measured, and Why
Two pressure measurements matter most on an RECD-fitted DG set:
- Differential pressure (ΔP) across the RECD: the pressure difference between the inlet and outlet of the device. It rises as soot and ash accumulate, falls during regeneration, and drops unexpectedly if the substrate cracks or a seal fails.
- Engine exhaust back-pressure: the pressure at the engine’s exhaust outlet, compared with the engine maker’s limit. Excessive back-pressure harms engine performance and can damage components. See how exhaust back-pressure affects a DG set.
Both can be measured by a switch, a sensor or a gauge.
The Three Options at a Glance
| Mechanical gauge | Pressure switch | Pressure sensor (transmitter) | |
|---|---|---|---|
| What it gives | A local needle reading | An on/off signal at a setpoint | A continuous value (e.g. 4–20 mA, voltage or digital) |
| Remote visibility | No | Alarm only | Full value and trend |
| Trend analysis | No (unless manually logged) | No | Yes |
| Detects slow change | Only if someone reads it | Only when setpoint is crossed | Yes |
| Detects falling ΔP (crack) | If someone notices | Only with a separate low setpoint switch | Yes |
| Self-diagnosis | No | Limited | Often (e.g. out-of-range signal) |
| Cost | Lowest | Low | Moderate |
| Typical use | Local indication; quick checks | Simple alarms; interlocks | Monitoring, alarms, regeneration control |
How a Pressure Switch Works
A pressure switch contains a diaphragm or piston that moves with pressure. When pressure crosses a set value (the setpoint), the movement operates an electrical contact, opening or closing it. When pressure falls back by a certain amount (the deadband or hysteresis), the contact returns.
Strengths
- Simple and robust. No electronics in many designs, and no scaling or software.
- Direct. The contact can drive an alarm, a relay or a controller input without configuration.
- Inexpensive, and easy for any electrician to wire.
Limitations
- It tells you only one thing: above or below the setpoint. It doesn’t tell you how far above or below.
- No warning of approach. ΔP could be 1 mbar below the setpoint for weeks and the switch would say nothing.
- One setpoint per switch. Warning and critical levels, or a low ΔP alarm for cracks, each need another switch.
- Hidden failures. A switch that has seized, has a blocked impulse line or has a broken contact gives no signal, which looks exactly like “normal.”
- Setpoint drift. Mechanical setpoints can shift over time, and you won’t know without testing.
- Load sensitivity. ΔP and back-pressure rise with exhaust flow, so a fixed setpoint suitable at full load won’t trigger at light load, and one suitable at light load may trigger at full load. See how to compare RECD pressure readings at different generator loads.
How a Pressure Sensor (Transmitter) Works
A pressure transmitter converts pressure into a continuous electrical signal using a sensing element, often a strain gauge or capacitive cell. Common outputs are:
- 4–20 mA current loop, widely used in industry, robust over long cables
- 0–10 V or similar voltage, simpler but more sensitive to cable losses and noise
- Digital output (such as Modbus), sometimes from transmitters with built-in electronics
The RECD controller, BMS or data logger reads the signal, scales it to engineering units, and uses it for display, trends, alarms and, in some systems, regeneration control.
Strengths
- Full information. You see the actual value at all times.
- Trends. You can see slow build-up, sudden drops and changes after cleaning. See can RECD efficiency change between new and aged conditions.
- Multiple alarms from one sensor. Warning, critical and low-ΔP alarms can all be set in software.
- Load-aware alarms, where the controller supports them, comparing ΔP with an expected value for the current load.
- Fault detection. With a 4–20 mA signal, the “live zero” of 4 mA means a broken wire or failed sensor typically produces a signal below the normal range. Many systems follow the NAMUR NE 43 convention, which treats signals at or below about 3.6 mA, or at or above about 21 mA, as sensor faults.
Limitations
- More to configure. Range, scaling and alarm limits must be set correctly. See RECD Modbus data: units, scaling and register mapping explained.
- Needs a receiving system, such as a controller, BMS or logger, and power.
- Drift is possible, so calibration checks are still needed. See installed RECD sensor calibration: what should facility teams document.
- Higher cost than a switch, though modest compared with the RECD itself.
Failure Modes Compared
How each device fails matters as much as how it works.
| Failure | Pressure switch | Pressure transmitter (4–20 mA) |
|---|---|---|
| Wire broken | Depends on wiring. With a normally open contact, no alarm: looks normal | Signal drops below 4 mA: detectable as fault |
| Sensor element seized or failed | No alarm: looks normal | Signal often frozen or out of range: detectable with flatline or range checks |
| Impulse line blocked with soot | No alarm: pressure can’t reach the switch | Value frozen or slow to respond: detectable with flatline checks |
| Setpoint drift | Alarm triggers at wrong pressure: undetectable without testing | Not applicable (drift shows in calibration checks) |
| Condensate in impulse line | Offset at switch: alarm early or late | Offset in value: visible as a zero error when DG stopped |
The key point: a pressure switch tends to fail silently. A transmitter’s failures are usually visible, if the receiving system is set up to look for them. See detecting missing data in RECD monitoring records.
Making switches fail safer
If you use pressure switches, reduce the silent-failure risk by:
- Wiring the alarm “fail-safe”. Use a contact that is closed when healthy and opens on alarm, so a broken wire also produces an alarm.
- Testing regularly by applying pressure or using a test port, and recording the actual trip point.
- Inspecting and clearing impulse lines as part of routine maintenance.
Where Mechanical Gauges Still Fit
A simple dial gauge across the RECD is still useful:
- Local indication for technicians and operators during walk-rounds
- A cross-check for transmitter readings
- A backup if electronic monitoring fails
But a gauge only helps if someone reads it and records the reading. On its own, it isn’t monitoring. If you rely on gauges, log readings at a standard condition (same load, fully warm) every time. See why RECD test results may differ from day-to-day site readings.
Getting more from a gauge
If a gauge is your main pressure indication, make it count:
- Choose a range that suits normal readings. A 0–100 mbar gauge for a device that normally reads 40 mbar is far more readable than a 0–600 mbar one
- Mark the normal band and the action level on the dial face, for example with coloured tape
- Log the reading at every DG run, with the kW load and minutes since start
- Check the gauge against a calibrated reference at least annually
- Keep the impulse line clear, since gauges suffer from blocked lines too
Common Combinations
Many installations use more than one type:
| Configuration | How it works | Suits |
|---|---|---|
| Gauge only | Manual readings | Very small sets with frequent attendance |
| Gauge + high back-pressure switch | Local reading plus one hardwired alarm | Small passive RECDs; budget installations |
| Transmitter only | Continuous monitoring, software alarms | Sites with a controller and BMS, where fault detection is configured |
| Transmitter + independent switch | Continuous monitoring plus a hardwired backup alarm on back-pressure | Critical sites; engine protection independent of software |
| Transmitter + gauge | Monitoring plus local visual cross-check | Most controller-equipped installations |
For engine back-pressure protection on critical sites, a combination of a transmitter for monitoring and an independent, fail-safe switch for the hardwired alarm gives both visibility and redundancy. If software, communication or the controller fails, the switch still works. See how to design an RECD alarm response matrix for operators.
Signal Types Compared
If you choose a transmitter, the output signal type affects reliability and integration:
| Signal | How it works | Strengths | Weaknesses | Typical use |
|---|---|---|---|---|
| 4–20 mA | Current proportional to pressure | Robust over long cables; resistant to noise; broken-wire detection via live zero | Needs an analogue input and scaling | Most industrial and BMS installations |
| 0–10 V (or 0–5 V) | Voltage proportional to pressure | Simple; inexpensive inputs | Sensitive to cable length and noise; 0 V can’t be told apart from a broken wire | Short cable runs inside panels |
| Digital (e.g. Modbus) | Value transmitted as data | Many values on one cable; diagnostic data available | Needs correct register mapping; communication monitoring | Smart transmitters; controller-integrated sensors |
For DG rooms, where cables often run past hot and electrically noisy equipment, 4–20 mA is usually the safest analogue choice.
How to Test a Pressure Switch Properly
A switch that has never been tested is an assumption, not a protection. A practical test, following the supplier’s instructions:
- Isolate the switch from the exhaust using its test valve or by disconnecting the impulse line safely, with the DG stopped.
- Connect a calibrated pressure source, such as a hand pump with a reference gauge.
- Raise the pressure slowly and record the pressure at which the contact operates. This is the as-found trip point.
- Lower the pressure slowly and record where the contact resets. The difference is the deadband.
- Confirm the alarm reaches the BMS or alarm panel with the correct text.
- Adjust the setpoint if needed, and record the as-left trip point.
- Inspect and clear the impulse line, then reconnect and check for leaks.
- Record everything. See installed RECD sensor calibration: what should facility teams document.
Quarterly testing is common for switches used in engine protection. Annual testing may suit lower-consequence alarms.
Choosing Setpoints for Switches
If you use a switch, the setpoint deserves care:
- For back-pressure protection, base it on the engine maker’s maximum back-pressure, with a margin so the alarm comes before the limit is reached.
- For high ΔP, base it on the RECD supplier’s recommended cleaning or regeneration threshold.
- Consider load. Since pressure rises with exhaust flow, a setpoint chosen at full load won’t warn of a partly blocked filter during light-load running. If the set often runs light, a transmitter with load-aware alarms is better.
- Set and record the deadband, so the alarm doesn’t chatter on and off around the setpoint.
- Record the as-found trip point at each test, so drift can be tracked.
Installation Points That Apply to Both
Whichever device you use, the connection to the exhaust is often the weak point.
Tapping locations
- ΔP tappings should be on the RECD’s inlet and outlet sections, as specified by the supplier, away from turbulent zones right at cones and flanges
- Back-pressure tapping should be close to the engine’s exhaust outlet, where the engine maker specifies its back-pressure limit
Impulse lines
- Use suitable heat-resistant tubing for hot exhaust, as the supplier specifies
- Slope impulse lines continuously back towards the exhaust, so condensate drains away instead of collecting in low points
- Keep them as short as practical
- Mount the sensor or switch away from direct radiant heat, with the impulse lines providing some cooling distance
- Provide isolation or test valves where practical, so the device can be checked without dismantling
Vibration and heat
Mount devices on brackets isolated from engine and exhaust vibration, and protect cables from hot surfaces. See RECD cable routing near hot exhaust lines.
A Worked Example: The Switch That Said Nothing
The scenario is illustrative. A commercial building’s 200 kVA DG set had a passive RECD with a mechanical ΔP gauge and a single high back-pressure switch set at the engine’s limit, wired as a normally open contact to the BMS.
Over several months, the DG ran mostly at light load during short outages. Soot built up. The gauge, which nobody was logging, crept from 35 to 70 mbar. Then a long outage ran the set at high load for six hours. The engine’s turbocharger later showed signs of overheating.
Investigation found:
- The back-pressure switch’s impulse line was blocked with soot, so the switch never saw the high pressure
- Because the contact was normally open, the blockage gave no signal: the BMS showed “normal” throughout
- The gauge readings would have shown the build-up, but nobody was recording them
The fix:
- A differential pressure transmitter across the RECD and a back-pressure transmitter, both connected to the BMS with trends, load-aware warnings and flatline detection
- The back-pressure switch retained as an independent hardwired alarm, but rewired fail-safe and added to a quarterly test schedule
- Impulse line inspection added to routine maintenance
- Gauge readings logged at the monthly test run
Specifying Pressure Monitoring in Your RECD Order
Instead of “pressure monitoring included,” specify:
- What is measured: ΔP across the RECD, engine back-pressure, or both
- Device type: transmitter, switch or gauge for each, and the output signal
- Range and accuracy
- Alarm functions required, and whether they’re in the controller, BMS or hardwired
- Fail-safe wiring for any hardwired alarm
- Impulse line material and routing requirements
- Calibration certificates and recommended check intervals
See how to write a technical RFQ for an RECD project and integrating RECD monitoring with a building management system.
Questions to Ask Your RECD Supplier
- Which pressures does the device monitor as standard, and with what device types?
- What range and accuracy do the sensors or switches have?
- Are switches wired fail-safe, and can they be tested in place?
- What output signals are available for the BMS?
- How are impulse lines specified and routed to avoid condensate and soot blockage?
- Does the controller provide load-compensated ΔP alarms?
- What calibration or test intervals do you recommend?
Related Reading
- Installed RECD sensor calibration: what should facility teams document?
- How to compare RECD pressure readings at different generator loads
- RECD temperature sensor placement: what should be measured?
- How to separate a sensor fault from an RECD performance problem
Frequently Asked Questions
Is a pressure switch enough for RECD monitoring?
For a small, simple installation, a fail-safe back-pressure switch plus a regularly logged gauge may be adequate. For most sites, a pressure transmitter gives far more useful information, especially trends and early warning, and makes failures visible.
Why does my pressure switch never alarm even though the gauge reads high?
Common causes are a blocked impulse line, a setpoint drifted or set too high, a failed contact, or a broken wire on a normally open circuit. Test the switch by applying pressure and confirm the alarm reaches the BMS.
Can one transmitter replace several switches?
Yes. A single ΔP transmitter can provide warning, critical and low-ΔP alarms in software. For engine back-pressure protection on critical sites, an independent switch is still worth keeping as a hardwired backup.
What does 4–20 mA “live zero” mean?
Zero pressure is represented by 4 mA, not 0 mA. A reading near 0 mA therefore means a broken wire or failed sensor, not zero pressure, so faults can be detected.
Can a pressure switch be used for regeneration control?
Some simple systems use a ΔP switch to request regeneration when the filter reaches a set loading. It works, but because ΔP depends on load, a fixed switch may request regeneration too early at high load and too late at light load. Transmitter-based control with load compensation is more precise. Follow the supplier’s design.
What range should a ΔP transmitter have?
Choose a range that covers the highest ΔP you expect, including heavy soot loading at full load, with some margin, but not so wide that normal readings sit in the bottom few percent of the range, where accuracy is poorest. Your RECD supplier can recommend a range based on the device’s design limits.
Do pressure transmitters need calibration?
Yes. They can drift, especially in hot, vibrating environments. Check zero with the DG stopped, compare with a reference periodically, and keep calibration records.
Choose the Device for the Decision
A switch answers “is it above the line?” A transmitter answers “where is it, and where is it heading?” Most RECD installations need the second answer, and some benefit from both. Aceget supplies CPCB type-approved Retrofit Emission Control Devices for DG sets from 25 kVA to 2500 kVA, and we can advise on the pressure monitoring that suits your device and site. See our certifications, or contact our team.