What Happens When a Generator Stops During RECD Regeneration?

It happens on Indian sites every day. The grid fails, the DG set starts, and after half an hour the RECD begins a regeneration cycle. Twenty minutes later the grid returns, the ATS transfers the building back, and the set runs its cool-down and stops. The regeneration cycle was not finished.

Is that a problem? Usually not a serious one, provided the device and the site controls were designed with it in mind. Occasionally it is, if the filter was heavily loaded and the heat had nowhere to go.

This article walks through what happens inside a filter-based RECD when the engine stops mid-cycle, how the outcome differs between regeneration types, what operators should and should not do, and which design features make the event harmless. For the background on how regeneration works, start with RECD Regeneration Explained: Active vs Passive Regeneration.

Why Engines Stop Mid-Cycle

A DG set rarely stops because someone wants to interrupt regeneration. More often, the stop is driven by something else:

• Mains restoration followed by ATS retransfer and the set’s cool-down timer
• A scheduled test run ending after a fixed number of minutes
• A protective trip such as high coolant temperature, low oil pressure or overspeed
• An emergency stop pressed by staff
• Fuel starvation from an empty day tank or blocked filter
• Load-management logic shutting down one set in a multi-set system

On standby sets, the first two account for the large majority of interrupted cycles. That matters, because both are predictable and can be coordinated with the RECD controller.

What the Filter Is Doing at the Moment of Stopping

To understand the risk, picture the filter at three different points in an active cycle.

Point in the cycle	Filter state	What a stop does
Heat-up phase	Filter warming, soot not yet burning	Heating stops, filter cools, soot remains. Low risk.
Early burn phase	Soot igniting at the inlet, flame front starting to travel	Burning may continue on residual oxygen, then slows as gas stops flowing. Moderate risk if soot load is high.
Late burn phase	Most soot already burnt, ΔP near clean	Little fuel left to burn. Low risk.

The middle row is the one that matters. When exhaust flow stops, two things happen at once. The gas that was carrying heat out of the filter is gone, and the soot that has already ignited keeps releasing heat for a while. If there is a lot of soot, local temperatures can rise faster than they would during normal regeneration.

Vehicle engineers know a related problem as the “drop-to-idle” event: a filter regenerating under load, then suddenly given low flow and high oxygen, can see peak temperatures well above design. A full engine stop is a more extreme version of the same physics. DieselNet’s regeneration overview discusses this type of uncontrolled regeneration.

How the Outcome Differs by Regeneration Type

Passive, catalysed systems

There is no discrete cycle to interrupt. When the engine stops, exhaust stops, temperature falls and the slow NO₂ reaction fades out. The filter simply waits until the next run. The only concern is a filter that had already built a large soot load during previous light-load running and was burning it fast at the moment of stopping. Even then, passive systems seldom hold enough heat to cause damage.

Heater-assisted systems that regenerate with exhaust flowing

This is the case that needs most care. If the heater keeps running after the engine stops, it is heating a filter with no gas flow to spread or remove heat. Well-designed controllers cut the heater as soon as they receive an engine-stop or low-flow signal. If that signal is missing or slow, the heater can create a local hot zone.

Heater-assisted systems designed to regenerate with the engine off

Some devices are designed to regenerate when the set is stopped, using a controlled low air flow and a heater. For these, an engine stop is not an interruption at all; it may be the normal starting condition. The concern is the opposite: an engine start during such a cycle. See Parked vs In-Service RECD Regeneration: What Changes for Operators?.

Load-assisted regeneration

If the device relies on the engine running at higher load to regenerate, a stop simply ends the attempt. Soot remains and the request will return. The main issue is repeated failures, which build soot load over time; see RECD Regeneration Not Completing: What Operators Should Record.

A Minute-by-Minute View of a Typical Interrupted Cycle

The sequence below describes a heater-assisted device on a 320 kVA standby set, interrupted by mains return. Timings are illustrative; your device’s controller may differ.

Minute 0. Mains returns. The ATS waits for its retransfer delay, then switches the building back to the grid. The DG set unloads.

Minutes 0–3. The set runs unloaded on its cool-down timer. Exhaust flow drops sharply, and exhaust temperature starts falling. A good RECD controller sees the load drop and either pauses heating or ends the cycle in a controlled way.

Minute 3. The engine stops. Exhaust flow is zero. The controller should now have switched the heater off and recorded “Regen interrupted”.

Minutes 3–10. Any soot already burning continues briefly using oxygen in the filter, then slows. Outlet temperature may hold or rise slightly before it falls. The housing stays hot.

Minutes 10–60. The filter cools by conduction and natural convection through the stack. Soot that did not burn stays in the filter.

Next run. At the next start, the controller reads ΔP and its soot estimate. If the soot load is within its safe limit, it can request a new cycle once conditions are right. If not, it may inhibit regeneration and raise an alarm, which is covered in RECD Regeneration Inhibit Signals.

What Operators Should Do When It Happens

Most interrupted cycles need nothing more than a record. Follow your supplier’s instructions first, but the following is a sensible default.

Do

1. Note the time, the reason for the stop and the controller message. A photo of the panel is ideal.
2. Check the RECD panel for over-temperature or fault alarms.
3. Look at the stack and the housing from a safe distance. Light haze can be normal; dense smoke from joints, or a glowing housing, is not.
4. Let the device cool naturally.
5. Record ΔP at the next run, at a known load, so you can compare it with the value before the interrupted cycle.

Do not

• Do not open inspection covers, sensor fittings or clamps on a hot housing. Admitting fresh air to a filter with burning soot can intensify the burn, and the surfaces can cause severe burns.
• Do not restart and stop the engine repeatedly to “help” the cycle unless the supplier’s procedure says so.
• Do not spray water on the housing. Sudden cooling can crack ceramics and distort metal.
• Do not force a manual regeneration straight away on a filter you know was heavily loaded, unless the controller and supplier confirm it is safe.

When an Interrupted Cycle Becomes a Concern

Treat the event as a fault, not a routine note, if any of these occur:

• An over-temperature alarm on the RECD controller
• Smoke escaping from flanges, clamps or the housing rather than the stack outlet
• Discoloured or visibly glowing housing
• A burning smell from insulation or nearby materials
• ΔP after the next run that is much lower than expected and does not recover, which can indicate a cracked substrate
• Repeated interrupted cycles in a short period

In these cases, keep people clear of the housing, follow your site’s emergency procedure if there is any fire risk, and call your RECD service provider. A low ΔP that stays low is particularly important, because it may mean exhaust is bypassing a damaged filter; our article on separating a sensor fault from an RECD performance problem explains how to check.

Design and Control Features That Make Stops Harmless

A stop mid-cycle should be a designed-for condition, not a surprise. When you specify or review an RECD, look for these features.

Feature	Why it matters
Engine-running or exhaust-flow signal to the RECD controller	Lets the controller cut the heater instantly when the engine stops
Soot-load limit for starting active regeneration	Prevents a cycle from starting with so much soot that an interruption could overheat the filter
Over-temperature cut-out with an independent sensor	Stops heating if temperature rises unexpectedly, even if the main logic fails
“Regen interrupted” status and log entry	Gives operators and service engineers a clear record
Controlled pause or wind-down on load drop	Reduces heat input before the engine actually stops
Restart logic that re-checks ΔP and temperature before re-attempting	Avoids stacking cycles on a hot, loaded filter
Clear operator instructions on the panel or nearby	Reduces risky improvised actions

Many of these depend on good wiring between the DG controller and the RECD controller. Our guides on RECD communication failure with the DG controller and RECD alarm vs generator trip control response explain how those links should behave.

Site-Side Measures: Reducing Interruptions in the First Place

You cannot prevent the grid from returning, but you can reduce how often regeneration is cut short.

Coordinate the ATS retransfer delay

A slightly longer retransfer delay, agreed with the electrical team, can give an in-progress cycle time to finish. This must be balanced against fuel use and building needs. See RECD fitment and ATS testing coordination.

Plan test runs around regeneration

If weekly test runs are short and light, regeneration may start near the end and never finish. Extending some tests or running them on load, perhaps with a load bank, gives the device a fair chance. Our RECD commissioning load bank testing plan shows how load banks are used during commissioning; the same approach can support periodic regeneration runs.

Give the RECD a status input from the DG controller

A simple “DG running” and “DG loaded” signal allows the RECD to time its cycles sensibly. Where a BMS is present, it can also hold off non-urgent cycles when a stop is expected; see integrating RECD monitoring with a building management system.

Review protective trips

If stops are often caused by engine trips, fix the engine problem. Overheating, low oil pressure and fuel faults all have their own causes; see common causes of generator overheating and how to fix low oil pressure in a diesel generator.

Different Sites, Different Priorities

Hospitals and data centres. Availability comes first. The RECD must never delay a start, a transfer or a stop the electrical system needs. Regeneration logic should defer to the DG controller, and interrupted cycles should simply be logged. See RECDs for hospitals and RECDs for data centres.

Housing societies and offices. Outages are irregular and often short, so many cycles will be interrupted. A device that tolerates this gracefully, and a monthly loaded run to let it complete a full cycle, is usually the practical answer.

Telecom and remote sites. Sets start and stop many times a day and nobody is present. Remote alerts for interrupted cycles and over-temperature are worth having; see remote RECD alerts for unattended sites.

Frequently Asked Questions

Is it harmful if the generator stops during RECD regeneration?

Usually not, if the controller stops heating promptly and the filter was not heavily loaded. The risk rises when a large soot load is burning and gas flow suddenly stops.

Will the RECD finish the regeneration next time?

Normally the controller re-checks the filter at the next run and requests a new cycle when conditions allow. If the soot load is too high, it may block regeneration and ask for service.

Should I restart the generator to finish the cycle?

Only if your supplier’s procedure says so. Repeated starts and stops can add heat stress and make matters worse.

Can a stopped generator’s RECD catch fire?

The filter itself is designed for high temperatures, but soot burning without gas flow can overheat it. Keep combustible materials away from the housing and follow your site’s fire safety rules.

How can I reduce interrupted regeneration cycles?

Coordinate ATS retransfer delays and test-run lengths with the RECD, provide engine status signals to its controller, and schedule occasional loaded runs.

Plan for the Stop Before It Happens

Every standby DG set will eventually stop during a regeneration cycle. The difference between a non-event and a damaged filter lies in the controller logic, the signals it receives and the instructions operators follow.

Aceget supplies CPCB type-approved Retrofit Emission Control Devices and works with site teams on control interfaces, ATS coordination and operator procedures. See how Aceget approaches RECD installation or contact us to review your DG set’s run pattern.