RECD Temperature Sensor Placement: What Should Be Measured?

Temperature tells you more about a Retrofit Emission Control Device than almost any other measurement. It shows whether the catalyst is hot enough to work, whether passive regeneration is likely, whether an active regeneration is running as it should, and whether something dangerous, such as an uncontrolled soot burn, is happening inside the device.

But a temperature reading is only as useful as the sensor’s position. A thermocouple placed too close to a cone, poking only a few millimetres into the gas, or mounted where heat conducts away along the pipe wall, can read tens of degrees away from the real gas temperature. Placed in the wrong spot entirely, it measures something nobody needs.

This guide covers which temperatures to measure on an RECD-fitted DG set and why, the sensor types commonly used, and how to place and install them so the readings can be trusted.

What Each Temperature Tells You

1. RECD inlet temperature

Where: in the exhaust pipe just upstream of the RECD’s inlet cone, on a straight section.

Why it matters:

• Shows whether exhaust reaching the device is hot enough for the catalyst and for passive regeneration
• Explains RECD behaviour at different loads. Light-load running with low inlet temperature often explains soot build-up. See choosing an RECD for a generator running at low load
• Provides the reference for the temperature rise across the device

Priority: essential for any RECD with a controller.

2. RECD outlet temperature

Where: just downstream of the RECD’s outlet cone, on a straight section.

Why it matters:

• Shows the device’s thermal state and when it has stabilised after start-up. See how to account for warm-up time when testing an RECD
• Detects regeneration: outlet temperature rises above inlet when soot is burning
• Detects overheating or uncontrolled burns, a critical safety alarm
• Protects downstream components such as bellows, silencers and insulation

Priority: essential.

3. Temperature rise across the RECD (outlet − inlet)

This isn’t a separate sensor but a calculated value from sensors 1 and 2. It’s one of the most informative numbers you can have:

ΔT (outlet − inlet)	Typical interpretation
Slightly negative (a few degrees)	Normal; heat loss through the device housing, no regeneration
Near zero	Steady state
Moderately positive	Catalytic oxidation of CO/HC, or passive soot burning
Strongly positive, rising quickly	Active regeneration, or an uncontrolled soot burn. Check controller status immediately

The actual values depend on the device, so record typical ΔT from your own installation during normal operation and during regenerations. See RECD testing during regeneration: what should be recorded.

4. Engine exhaust temperature (turbo outlet or manifold)

Where: at the engine’s exhaust outlet, often already measured by the DG controller.

Why it matters:

• Confirms engine-side conditions independently of the RECD
• Helps distinguish engine problems from RECD problems. If exhaust temperature at the engine is abnormal, the cause is upstream. See how to separate a sensor fault from an RECD performance problem
• Shows heat loss along a long exhaust run. See where to position an RECD along a long exhaust run

Priority: useful; often available from the DG controller without a new sensor.

5. Internal or mid-bed temperature (some designs)

Where: inside the device, between stages or within the substrate, if the design includes a port.

Why it matters: gives direct insight into regeneration progress and peak temperatures. Some active-regeneration designs use it for control.

Priority: only where the supplier designs for it. Don’t add ports to a substrate housing without the supplier’s approval.

6. Heater temperature (heater-assisted designs)

Where: on or near the heater element, as designed by the supplier.

Why it matters: controls and protects the heater, and confirms it’s working. See RECD controller and heater power supply requirements.

7. DG room ambient and controller enclosure temperature

Where: in the DG room, and inside the RECD controller’s enclosure if the supplier recommends it.

Why it matters:

• Explains seasonal behaviour and warm-up time
• Protects electronics. Controllers and transmitters have maximum operating temperatures, and DG rooms can get very hot. See assessing an RECD for very hot ambient conditions

Priority: useful, especially in hot or poorly ventilated rooms.

Summary: The Measurement Map

Measurement	Typical location	Essential?	Main uses
RECD inlet	Straight pipe before inlet cone	Yes	Catalyst and regeneration conditions; ΔT reference
RECD outlet	Straight pipe after outlet cone	Yes	Stability, regeneration, overheating protection
ΔT across RECD	Calculated	Yes (calculated)	Regeneration detection, abnormal burns
Engine exhaust	Turbo outlet / manifold	Useful	Engine vs RECD diagnosis; heat loss
Mid-bed / internal	Supplier-designed port	Design-dependent	Regeneration control and peak temperature
Heater	Supplier-designed	If heater fitted	Heater control and protection
Ambient / enclosure	DG room; controller box	Useful	Context; electronics protection

Choosing the Sensor Type

Two sensor types are common for exhaust temperatures.

Thermocouples

A thermocouple produces a small voltage from the junction of two different metals. Type K thermocouples are widely used for exhaust measurements because they cover the temperature range comfortably and are robust and inexpensive. Standard reference tables for thermocouple types are published in the NIST temperature scale and thermocouple database.

• Strengths: wide range, fast response (especially with thin sheaths), rugged
• Weaknesses: lower accuracy than RTDs; need the correct extension cable type and a cold-junction compensation at the receiving instrument; can degrade over long periods at high temperature

Resistance temperature detectors (RTDs), such as Pt100

An RTD’s electrical resistance changes predictably with temperature.

• Strengths: good accuracy and stability
• Weaknesses: usually slower to respond; more sensitive to vibration damage; upper temperature limits depend on construction, so check suitability for exhaust temperatures, especially during regeneration

Practical choice

For most RECD inlet and outlet measurements, mineral-insulated Type K thermocouples in a suitable sheath are a sensible, common choice. RTDs suit lower-temperature points, such as ambient or enclosure, or applications where accuracy matters more than speed. Follow the RECD supplier’s specification, since the controller must be designed for the sensor type.

Placement Rules for Trustworthy Readings

1. Put the sensor where the gas is well mixed

Avoid positions immediately after bends, cones, flanges or the RECD body itself, where hot and cool streams may not be mixed and the flow is uneven. A short straight section before and after each sensor improves readings. For layout guidance, see RECD exhaust bends and reducers: a layout review.

2. Insert deep enough

A sensor that only just pokes through the pipe wall measures something between the gas temperature and the (cooler) wall temperature. As a rule of thumb:

• Aim for the sensing tip to reach towards the central part of the pipe, often around one-third of the diameter or more into the flow
• Make sure the immersed length is many times the sheath diameter, commonly at least ten times, to limit heat conducted away along the sheath

Follow the supplier’s specified insertion depth. In small pipes, there’s a trade-off between depth and blocking the flow.

3. Minimise conduction and radiation errors

• Insulate the pipe around the sensor, so the wall and fitting are close to gas temperature. See RECD insulation and heat shields: site plan
• Don’t let the sensor’s head and fitting act as a heat sink on an uninsulated, wind-cooled section
• Be aware that a sensor that can “see” a much hotter or cooler surface nearby can read slightly wrong from radiation

4. Make inlet and outlet sensors comparable

Because ΔT is calculated from two sensors, small errors in each matter. Use the same sensor type, similar insertion depth and similar insulation at inlet and outlet. Otherwise a few degrees of systematic difference appear as false “temperature rise.” Check both against each other periodically, with the DG running steadily and no regeneration, when ΔT should be small.

5. Use proper fittings

• Compression fittings or threaded thermowells rated for the temperature and vibration
• Sealed to prevent exhaust leaks, which would also affect pressure and emission measurements
• Arranged so the sensor can be removed for checking or replacement

Thermowell or direct insertion?

A thermowell is a protective tube that the sensor slides into. It allows the sensor to be removed without opening the exhaust, and protects it from soot and vibration, but it slows the response. Direct insertion responds faster, but means the exhaust is opened when the sensor is changed. For regeneration and overheating detection, response speed matters. Follow the supplier’s design.

6. Protect cables and connection heads

Thermocouple extension cables and connection heads have temperature limits lower than the sensor tip. Route cables away from hot surfaces, use the correct cable type, and secure them against vibration. See RECD cable routing near hot exhaust lines.

7. Keep sampling ports and sensor ports separate

Emission test sampling ports and temperature sensor ports have different requirements. Keep them apart, so neither disturbs the other. See how exhaust sampling location affects RECD test results.

Response Time: Why Speed Matters

Every temperature sensor takes time to follow a change in gas temperature. This is often described by a time constant: the time taken to reach about 63% of a sudden step change. Sensors in thick thermowells, or with heavy sheaths, can have time constants of tens of seconds or more. Thin, directly inserted thermocouples respond much faster.

Speed matters for RECD monitoring because:

• Uncontrolled soot burns can raise temperatures quickly. A slow sensor detects them late.
• Regeneration control in active systems relies on timely feedback, or temperatures can overshoot.
• ΔT across the device can show false transients if inlet and outlet sensors respond at different speeds. For example, during a load step, a fast inlet sensor and a slow outlet sensor briefly suggest a large temperature difference that doesn’t exist.

Use matching sensors at inlet and outlet. For protection functions, follow the supplier’s specification on response time.

Where Not to Put Temperature Sensors

• Directly at a flange or cone, where flow is disturbed and the metal is thicker
• On an uninsulated section exposed to drafts, where the fitting and sensor are cooled
• Inside a low point where condensate collects, which causes corrosion and false readings during warm-up
• Where the sensor head sits against a hot surface, which overheats the connection head and cable
• In a position that blocks a sampling port or interferes with emission testing access
• In the RECD housing, unless the supplier has designed a port for it

Specifying Temperature Measurement in Your RECD Order

To avoid surprises, specify:

• Which temperatures are measured (inlet, outlet, mid-bed, heater, ambient)
• Sensor type (e.g. Type K thermocouple, mineral-insulated) and sheath diameter
• Insertion depth and fitting type for each
• Insulation requirements around sensor locations
• Extension cable type and routing
• Signal path to the controller and BMS, including scaling
• Alarm functions based on temperature, and their recommended thresholds
• Check and calibration intervals

See how to write a technical RFQ for an RECD project and RECD general arrangement drawing: what to include.

Worked Example: A Misleading Outlet Temperature

The scenario is illustrative. A site noticed that its RECD’s outlet temperature consistently read about 40 °C lower than its inlet temperature during steady running with no regeneration, much more than the supplier expected. The controller interpreted this as normal, but regeneration alarms behaved oddly.

Investigation found:

• The inlet sensor was inserted to about the pipe centre, on an insulated section
• The outlet sensor had been installed with only a short insertion length, in a fitting on an uninsulated section, near a flange where the pipe was exposed to a ventilation draught

The outlet sensor was effectively reading closer to the pipe wall temperature than the gas temperature. Because the two sensors had been installed differently, the calculated ΔT across the device was wrong by tens of degrees, which confused the controller’s regeneration logic.

Fix: the outlet sensor was replaced with one of the correct insertion length, and the section was insulated. Afterwards, steady-state ΔT was within a few degrees, and regeneration detection worked as designed. The event was logged and the controller’s reference values were updated. See what should an RECD event log record.

Checking Temperature Sensors in Service

Routine checks for temperature sensors:

• Steady-state comparison: at steady load with no regeneration, inlet and outlet should be close. A growing difference suggests a sensor or installation problem.
• Cold check: when the DG has been off for many hours, all exhaust sensors should read close to ambient.
• Comparison with a reference probe: periodically, with a calibrated portable instrument through a spare port, or by removing the sensor for a check in a temperature calibrator.
• Response check: after a DG start, temperatures should rise promptly and in sequence: engine exhaust, then RECD inlet, then RECD outlet.

Record checks as described in installed RECD sensor calibration: what should facility teams document.

Keeping Placement Right Over Time

Sensor positions that were correct at commissioning can become wrong later:

• Insulation is removed during maintenance and not replaced, changing the conditions around a sensor
• A sensor is replaced with one of a different length or sheath diameter
• Exhaust pipework is modified, moving bends or reducers close to a sensor. See when should an RECD be retested after changes to the DG set
• Fittings loosen with vibration and thermal cycling, letting the sensor slide back

Add a quick placement check to routine maintenance: confirm insulation is intact, fittings are tight, and the steady-state inlet–outlet comparison is still within its normal band. Record replacements with the new sensor’s length and type.

Temperature Data in Investigations

When something goes wrong, such as engine damage, a cracked substrate or a fire alarm in the DG room, temperature records are often the most valuable evidence. A good set of records lets investigators answer:

• Did outlet temperature rise gradually, which suggests a soot burn, or suddenly, which suggests a sensor fault?
• Was a regeneration active at the time?
• Did the inlet temperature show the engine running abnormally hot?
• How long did any high temperature last?

For this to work, temperatures need to be logged at short intervals during running, with synchronised timestamps. See how to align RECD sensor timestamps with generator runtime logs and what should an RECD event log record.

Using Temperatures in Alarms and Dashboards

Common temperature-based alarms (thresholds set with your supplier):

• High outlet temperature, a critical alarm for possible uncontrolled burn or overheating
• High ΔT outside a regeneration, a warning of unexpected oxidation or a sensor fault
• Low inlet temperature for extended periods, an advisory that regeneration conditions are poor
• Sensor fault, for an open circuit or out-of-range reading

See how to design an RECD alarm response matrix for operators and what should a useful RECD operating dashboard show.

Related Reading

• Installed RECD sensor calibration: what should facility teams document?
• Pressure sensors vs pressure switches in RECD monitoring
• How to compare RECD pressure readings at different generator loads
• How to separate a sensor fault from an RECD performance problem

Frequently Asked Questions

Do I need both inlet and outlet temperature sensors?

For any RECD with a controller or regeneration system, yes. Together they show the device’s thermal state, regeneration activity and any abnormal heating, none of which a single sensor can do reliably.

Why is my outlet temperature lower than my inlet when the DG is running?

A small drop is normal, because the device loses heat through its housing. A large drop often points to a sensor problem: too short an insertion, an uninsulated mounting, or a damaged thermocouple. Check both sensors’ installation and compare them during steady running.

Which is better for exhaust, a thermocouple or a Pt100?

Thermocouples, commonly Type K, are the usual choice for exhaust temperatures, because of their range, robustness and speed. RTDs offer better accuracy at lower temperatures. Use what the RECD controller is designed for.

Can I add an extra temperature sensor to the RECD myself?

Only on exhaust pipework, with suitable fittings and without compromising sealing. Never drill into the RECD housing or substrate area without the supplier’s approval, as that can damage the device and affect its warranty.

Should temperature sensors be on the BMS as well as the RECD controller?

Yes, if the RECD controller can share them. Bringing inlet, outlet and ΔT values into the BMS gives operators context alongside DG load and status, enables trends and supports alarm routing. See integrating RECD monitoring with a building management system.

How often should temperature sensors be replaced?

There’s no fixed life. Thermocouples exposed to high exhaust temperatures for long periods gradually drift, and junctions can fail. Replace them when checks show drift beyond tolerance, when readings become erratic, or at the interval your supplier recommends. Record each replacement with the new sensor’s type, length and serial number, so later trend shifts can be explained.

How fast should temperature sensors respond?

Fast enough to detect regeneration and overheating promptly, typically seconds to tens of seconds. Thin-sheath thermocouples without thick thermowells respond fastest. Follow the supplier’s specification for protection functions.

Measure the Gas, Not the Pipe

Good temperature data starts with good placement: the right points, deep enough, comparable at inlet and outlet, and protected from heat loss. Aceget supplies CPCB type-approved Retrofit Emission Control Devices for DG sets from 25 kVA to 2500 kVA, and we specify sensor positions and installation details for each installation. See our certifications, or contact our team.