RECD Restriction After Exhaust Lagging Repairs: What to Inspect

Exhaust lagging is the kind of job that looks too simple to cause trouble. A crew removes old, tired insulation from the hot pipework around the RECD, fits new blankets and jackets, tidies up and leaves. The set restarts. A day later the controller reports a higher pressure than before, or a temperature that does not match the earlier records, or a faint burning smell in the plant room. Somebody asks the obvious question: did the lagging job do this?

Often it did, although rarely in the way people first imagine. Insulation is not a flow obstruction in the usual sense. It lives outside the pipe. But a lagging repair touches almost everything around an exhaust system. To reach the surface the crew removes covers, moves cables, disturbs sensor tubes and sometimes undoes flanges. To refit insulation they wrap, strap, clip and cut around fittings. Any of these steps can leave behind something that looks, to the controller, like a restriction: a crushed sensing tube, a covered pressure tap, a gasket pushed into the bore, a bellows bound by a tight jacket, loose fibre in the pipe, a changed heat balance or a small leak that was not there before.

This guide is a peel-the-layers inspection. It works from the inside of the exhaust path outward, because the cause of an apparent restriction can sit at any depth. It also covers two things that are easy to forget: heat is data, since insulation changes the temperatures your sensors report, and fire safety, because lagging that meets oil is a known hazard. We finish with a sign-off sheet you can attach to the work order next time. The advice is general. Your supplier’s drawings and the insulation maker’s instructions come first, and any inspection of a hot system needs cooling, isolation and permits.

Why lagging work can change what the RECD reports

Start with the physics, because it explains why one small job can touch so many readings.

Insulation changes heat flow. Bare hot metal loses heat to the room by radiation and convection. Wrapped metal loses far less. After a repair, the surface temperature of the pipe and canister can change, and so can the temperature of the gas inside, particularly at light load where a long run of pipe would normally shave off heat. A temperature sensor mounted on the surface, or in a pocket close to the wall, may read differently even though nothing else has changed. Our guide to RECD temperature sensor placement explains why sensor position matters.

Insulation changes mechanical freedom. Thick blankets and rigid jackets can restrain parts that were meant to move. A flexible bellows that was free now carries the stiffness of a jacket. A pipe that grew freely now presses on a clamp. See flexible bellows in an RECD exhaust system and thermal expansion planning for RECD exhaust connections.

Insulation hides things. A jacket covers leaks, soot stains and cracked welds that you would otherwise see on a walk-round, and it can trap leaking gas, oil or water against hot metal.

The work itself disturbs hardware. Whatever the insulation does, the crew has handled the sensing taps, the tubes, the cables and the flanges. That disturbance is often the real cause.

So there are two broad kinds of “restriction” after lagging work. A real restriction is something in the gas path or on the pressure path that physically reduces flow. An apparent one is a reading that has changed because a tap, a tube or a sensor has been affected, even though the exhaust itself flows as before. The inspection below tells them apart.

Layer 1: inside the exhaust path

Real restrictions live here. They are less common after an external lagging job but not unknown, especially if flanges were opened.

Gaskets pushed or fitted into the bore. When a flange is opened and closed, a new gasket that is slightly too large, or a reused gasket that is crushed and extruded, can protrude into the pipe. It narrows the passage and creates an edge that can whistle. See whistling after RECD installation.

Insulation or packing material inside the pipe. If a pipe was opened, loose fibre, offcuts, tape or a protective plug can end up in the flow. It is rare but it has been found, and it is worth a torch and a camera at any opened joint.

Bellows liners and internal sleeves. A bellows disturbed during the job can be refitted with its internal liner turned the wrong way, which can create a restriction and a noise. Check the flow arrow.

Dampers and diverters. If a diverter or bypass damper was moved or its actuator linkage disturbed, check that it is in the position it should be and reaches full travel.

How to check: Where a flange was opened, look inside with a torch and a borescope (cold and isolated), compare pipe bore at the joint against the drawing and measure pressure drop across the section. Our guide to high back-pressure after RECD installation explains the pressure-budget method, which finds the section holding the excess pressure.

Layer 2: the joints, gaskets and fasteners

Many lagging jobs involve removing old insulation from flanges so that the joints can be inspected or re-torqued. This is the layer where small errors create leaks that mimic a change in pressure.

  • Uneven bolt tightening after a flange was opened.
  • Reused gaskets that no longer seal.
  • Missing or substituted fasteners.
  • Misaligned faces pulled together by bolts.
  • Disturbed door or inspection covers on the canister.

A leak upstream of the filter lowers the flow through it and tends to make ΔP read lower than expected for the load. A leak near a pressure tap changes the reading directly. Our guide to exhaust leaks around an RECD walks the whole route station by station, and the guide on updating drawings and records after exhaust modifications explains why any change to a joint should be recorded.

How to check: Cold inspection of every joint disturbed, torque check against the supplier’s values, then a loaded run and a second inspection when cool, looking for new soot streaks.

Layer 3: sensing taps, thermowells and sensor tubes

This is the layer where most “restrictions after lagging” turn out to live, and most of them are apparent restrictions.

Pressure taps covered or blocked. A tap is a small port on the pipe or canister. Wrapped over, packed with insulation or sealed with adhesive tape or sealant, it may no longer sense the gas pressure properly.

Sensing tubes crushed, kinked or pinched. The thin tubes that carry pressure to the transmitter are easily crushed by a jacket strap, a clip or an over-tight blanket. A crushed tube reads wrong, lags or sticks. Tubes laid against a hot surface and now wrapped in insulation can also overheat and soften.

Tubes rerouted into a sag. If the tubes were moved aside during the work and put back by a different route, a new low point can collect condensate. See our article on RECD faults that appear only in wet weather.

Thermowells and temperature probes covered or displaced. An insulating blanket over the head of a probe, or a probe pushed out of its pocket, changes the temperature reading. A probe that is insulated at the head reads closer to the gas temperature, which may be higher than before.

Cables dislodged. Cables that were tied to the pipe or moved out of the way may now be touching hot surfaces. See RECD cable routing near hot exhaust lines.

How to check: Compare photographs from before the job with the job as left. Trace each sensing tube from tap to transmitter, check that it is clear, correctly connected, sloped and free from pinch points, and compare the controller reading with an independent gauge at a steady load. Our article on separating a sensor fault from an RECD performance problem offers a method.

Layer 4: the insulation itself

Now the lagging. The material, its thickness, its fit and the way it was fastened all matter.

Wrong material. Not all insulation is the same. Some is porous and soaks up liquid. Some is closed-cell or encapsulated and does not. A mineral-wool blanket that is fine on a dry pipe can become a hazard on a pipe that leaks oil or fuel. The rating of the material and its jacket must suit the surface temperature, which on a regenerating RECD can be much higher than on a plain exhaust. Ask the supplier for the maximum surface temperature during regeneration and check that the material is rated well above it.

Too thick, too tight or in the wrong place. Over-wrapped flanges and bellows can bind movement. Blanket stuffed into a gap around a sensor can change readings. Insulation must not cover items that the supplier specifies as exposed: for example drain points, inspection ports, pressure taps, certain heater housings and cooling surfaces.

Gaps and cold spots. A missing patch on a hot flange is a burn risk and a fire risk. A bridge of metal strap between canister and jacket can act as a thermal short.

Loose pieces and fibre. Loose fibres can enter air intakes or settle on hot surfaces. Check around the area and clear debris.

Mixed materials. Old and new insulation of different type butted together may behave differently when hot.

Our guide to RECD insulation and heat shields sets out the planning side, including what to insulate and what not to.

Layer 5: jackets, straps and supports

The outer skin and its fixings complete the picture.

  • Jacket clamps and straps that bear on tubes, cables or the bellows.
  • Support brackets and hangers moved or removed to reach the pipe and not put back.
  • Weight added to the pipe by new jackets, now carried by a flange.
  • Rattles. A loose jacket can buzz or rattle. See rattling inside an RECD for how to separate external hardware from internal sounds.

A system that was balanced in weight and movement before the job may now hang unevenly, and the joints take the difference. See wall-mounted versus floor-supported RECD support layouts.

Heat is data: reading the temperature change

If the readings changed after the lagging work, ask what the insulation did to each one. A simple comparison table helps. Take readings at the same load and the same length of time since start, before and after the job.

Sensor or readingPossible effect of new insulationWhat to ask
Surface-mounted temperature sensorReads higher because heat no longer escapesIs the sensor calibrated for a bare or an insulated surface?
Gas temperature at RECD inletMay rise slightly at light load because less heat is lostDoes the rise match the change in pipe loss?
Outlet temperatureRises more than inlet if heat is retained in the canisterIs the canister now hotter than the supplier’s design?
Regeneration durationCan shorten if the canister retains heat, or lengthen if a sensor reads wrongDid the cycle profile change?
Plant-room temperatureFalls because surfaces radiate lessIs cooling air flow now over-provided?
Differential pressureShould not change from insulation aloneIf it did, look at Layers 1 to 3

That last row matters. Insulation alone should not change the differential pressure. If ΔP changed after the lagging job, the cause is almost certainly in Layers 1 to 3: a tap, a tube, a gasket, a leak or an obstruction. This is a useful way to focus an investigation. If a temperature changed but ΔP did not, the likely cause is the insulation’s heat effect or a covered probe, not a flow restriction. Our guide to what regeneration temperature means explains the numbers, and RECD regeneration not completing covers what a changed temperature profile can do to a cycle.

Fire safety: why lagging deserves respect

Lagging is not only an insulation problem. It is a fire-safety problem, and the best evidence comes from ships’ engine rooms, where the consequences of a fire are severe and the records are detailed.

A marine safety flash notes that oil or hydraulic fluid spraying onto an uninsulated exhaust flange was the probable cause of an engine-room fire, that soft or absorbent lagging can accumulate leaking oil until it ignites, and that attention should be paid to satisfactory re-installation of insulation after maintenance (IMCA safety flash). Another IMCA note reports that oil-resistant laggings are available and offer some protection, while stressing that covering every hot surface without suitable material can increase the risk (IMCA SF 11/05). A shipowners’ club describes a generator exhaust-lagging fire that traced back to lagging soaked with oil during earlier repairs (The American Club), and a P&I club bulletin explains the “lagging fire” mechanism, in which flammable liquid seeping into porous insulation on a hot surface can heat up by oxidation until it ignites, and warns that smoking lagging should not be pulled apart until water is at hand (UK P&I Club). An insurer’s survey note finds that missing or dislodged exhaust-manifold insulation after maintenance is a leading cause of the deficiencies seen on engines (IUMI).

These are marine sources and marine rules, and a stationary plant room is a different setting. The lessons carry over:

  • Check for fuel, oil or coolant leaks before and after any lagging job. Fix them first.
  • Inspect lagging for soaking and replace any that is oil-stained.
  • Check for smoke or smell from lagging in the first runs after the job.
  • Do not disturb smouldering lagging without a plan and water or the right extinguisher on hand.
  • Avoid porous insulation near leak-prone joints, unless it is protected by a jacket that is leak-tight and oil-resistant, as your fire-safety adviser and the supplier approve.
  • Make sure insulation is refitted completely. Bare flanges and fittings are a heat and fire risk.

A smell after lagging work may simply be binder burn-off, which is common with new insulation and fades within hours. Our article on why an RECD smells different after a service visit shows how to tell burn-off from a problem.

An inside-out inspection sequence

Use this sequence when something changed after a lagging repair. Everything is done cold, isolated and with permits where needed.

  1. Gather the before-and-after evidence. Photographs from before the job, the work order, the drawing, the controller log and any independent readings.
  2. Compare readings at the same load. ΔP, back-pressure, inlet and outlet temperatures, regeneration profile, plant-room temperature. Decide which changed.
  3. If ΔP changed, go to Layers 1 to 3. If only temperatures changed, go to Layers 3 and 4.
  4. Check the sensing chain first (Layer 3): taps, tubes, probes, cables. It is the cheapest layer to inspect and the most likely culprit.
  5. Check disturbed joints (Layer 2): torque, gaskets, soot streaks.
  6. Check inside the path (Layer 1) where a flange was opened.
  7. Inspect the insulation and jacket (Layers 4 and 5): material, fit, covered items, binding, loose pieces, oil staining.
  8. Run a loaded test under observation and repeat the same readings.
  9. Document the finding and the fix, and update the drawings.

A worked illustration: the pressure that was a pinch

The following is an illustrative scenario, not a real customer. A textile plant has its 400 kVA set’s exhaust re-lagged during a shutdown, with the contractor replacing worn blankets on the pipework and the RECD canister. After restart, the controller reports ΔP about 25% above the earlier figure at the same 60% load, and an exhaust smell is noticed on the first day.

Step 1. The team has pre-job photographs and a log of ΔP at 60% load from the last four months, around 31 mbar. The new value is about 39 mbar.

Step 2. A portable gauge at the same taps reads 32 mbar. The controller and gauge disagree, so the exhaust is probably not the problem. The issue sits in the sensing chain.

Step 3. On a cold inspection, the high-pressure sensing tube is found pinched under a new jacket strap, a few centimetres from the tap. The photographs show that the tube was previously routed on the other side of the pipe, and the crew moved it to fit the jacket. A pressure tap on the low-pressure side is also half-covered by a blanket.

Step 4. The installer frees the tube, re-routes it on the original path with a continuous slope, adds a protective sleeve and trims the blanket back from the tap. A reading taken at 60% load after the next run shows 32 mbar on both the controller and the gauge.

Step 5. The smell, tracked on the second day, is binder burn-off from the new blankets and has faded. During the inspection the team also notices a small oil weep from a nearby pipe flange that the old lagging had hidden. They fix it before it can soak the new insulation. The “restriction” was a pinched tube, and the more serious find was the weep.

Writing the next lagging job so it does not cause this

The best fix is a better work order. Before the next lagging job, agree these points.

  • Scope and exclusions. Which items must stay uncovered: taps, drains, inspection ports, probes, heater housings and any hot-surface warning labels.
  • Material and rating. Insulation type, maximum temperature rating above the supplier’s regeneration figure, jacket type and whether it suits oil-prone locations.
  • Before-job photographs of taps, tubes, cables and probes, taken by the contractor and the owner.
  • Disturbance list. Which flanges, tubes and cables will be moved, and who will check them afterwards.
  • Leak check before refit. Fuel, oil and coolant leaks found and fixed before insulation goes on.
  • Torque and gasket specification for any joint opened.
  • After-job test. A loaded run, a comparison of ΔP and temperatures with the pre-job baseline and an independent gauge reading.
  • Sign-off. A named person signs that taps are clear, tubes are free, cables are routed, joints are tight and the drawing is updated.

Sign-off sheet for lagging work around an RECD

CheckDone (initials, date)Notes
Pre-job photographs filed
Pressure taps visible and unobstructed
Sensing tubes traced, free, sloped, unpinched
Temperature probes in place and not insulated unless specified
Cables routed away from hot surfaces
Opened joints re-torqued, new gaskets, no soot streaks after loaded run
Bellows free to move, liner direction correct
No leaks of fuel, oil, coolant before refit
Insulation type and rating recorded
Jacket straps clear of tubes and cables
Supports and hangers restored
Loaded test: ΔP and temperatures compared with baseline
Independent gauge agrees with controller
Drawing and asset record updated

Questions people ask

Can lagging alone raise the differential pressure across an RECD?

Not through its insulating effect. If ΔP rose after lagging work, look for a pinched tube, a covered tap, a leak or a physical obstruction created during the job.

Is a burning smell after new lagging normal?

A mild smell as binders burn off is common and fades within hours of running. A smell that persists, or smoke from the insulation, deserves immediate attention.

Should I insulate the RECD canister at all?

That is a design question for the supplier. Some designs call for insulation, some call for exposed surfaces, and some need specific areas left bare. Follow the supplier’s drawing.

What if I cannot tell whether the change is real or apparent?

Compare the controller against an independent gauge at the same load. Agreement points to a real change, and disagreement points to the sensing chain.

Who is responsible if the lagging contractor caused it?

It depends on your contract and the scope of work. Good before-and-after records settle most disagreements. See customer and installer responsibilities in an RECD project.

Final thought

A lagging repair around an RECD touches the pipe, the joints, the taps, the tubes and the heat balance, and any of these can change what the controller reports. Peel the layers from the inside out, remember that insulation alone does not change ΔP, treat oil-soaked lagging as a fire hazard and use a sign-off sheet so the next job leaves the system as it found it. If you would like help reviewing a work order or comparing before-and-after readings, you can browse our retrofit emission control devices, read the FAQ or contact the Aceget team.

Related Reading

Sources



Leave a Reply