- November 15, 2026
- Posted by: editor
- Category: RECD & Emission Control Devices
Two quotations land on your desk. One says “on-site RECD filter cleaning, same day”. The other says “off-site cleaning at our workshop, five working days, with flow test report”. The first is cheaper and faster. The second sounds more serious. Which one is right?
The honest answer is that the venue is not the point. A filter is not clean because it was cleaned in a plant room or because it was cleaned in a workshop. It is clean when a defined process has removed a measurable amount of deposit and an independent test shows that flow has been restored. RECD filter cleaning can fail in either place, and it can succeed in either place, depending on what equipment and procedure are involved.
This article therefore compares on-site and off-site service by breaking the job into its separate tasks, asking which venue can really perform each one, and then scoring both options against the factors that matter to a DG set owner: depth of cleaning, proof of result, downtime, dust and safety, waste control, risk to the element, cost and warranty. A scored worked example at the end shows how to turn two quotations into a decision you can defend.
A note on scope. RECD designs differ, and so do cleaning methods. Where this article says “element” it means the ceramic filter core or cartridge inside the device. Always follow the supplier’s approved procedure for your model, and read the process descriptions below as general explanation, not as instructions for your own device.
Start with the job, not the venue
A cleaning visit is not one task. It is a bundle of up to five, and they are not equally easy to do in a plant room.
| Job | What it achieves | Can it be done on site? | Can it be done off site? |
|---|---|---|---|
| 1. Inspect and photograph the inlet face | Shows soot cake, ash crust, oil wetting, cracks, melted areas | Yes, if the device can be opened | Yes, and usually under better light |
| 2. Remove loose surface soot and ash | Clears the face and the housing | Yes, with a HEPA-fitted vacuum and damp wiping | Yes |
| 3. Remove ash from inside the channels | Restores flow capacity | Only with enclosed, extracted equipment, which many site visits do not bring | Yes, in a cabinet or wet process designed for it |
| 4. Treat oily or hardened residue | Loosens material that air and vacuum cannot shift | Rarely | Yes, typically with a controlled thermal step |
| 5. Prove the result | Weighs the element and tests airflow or pressure drop before and after | Rarely, because a flow bench is workshop equipment | Yes, where the workshop has the bench |
Read the table from the bottom up. Jobs 3, 4 and 5 are what separate a real cleaning from a tidy-up. If a quotation for on-site service promises job 1 and job 2 only, you are buying an inspection and a vacuum, and the price should reflect that. If it promises job 3, ask what equipment is brought. If it promises job 5, ask which instrument measures it, how it is calibrated and what number counts as a pass.
What “on-site” can mean: three versions
Providers use the term loosely, so find out which version is being quoted.
Version A: inspection and vacuum. A technician opens the housing, photographs the inlet face, vacuums loose deposit with a HEPA-fitted unit, wipes down and closes up. It is useful for surface soot, for looking at the condition of the element and for checking sensor ports. It does little for ash packed in channels.
Version B: portable enclosed cleaning. The technician brings a small sealed cabinet or enclosure with its own dust extraction and a regulated air supply, and cleans the element beside the set. This can remove far more material than version A, but it is only as good as the enclosure, the extraction and the operator’s method. Ask for photographs of the equipment and a description of how exhausted air is filtered.
Version C: mobile workshop. A vehicle carries cleaning, and sometimes testing, equipment to the site. Its capability varies enormously. Some are close to a small workshop, while others are version B in a van.
For all three, what matters is whether dust is contained. Our guide to handling soot and ash removed during RECD servicing sets out the dust, PPE and waste expectations that apply wherever the work is done.
What “off-site” means
Off-site means the element is removed, sealed, transported and processed elsewhere. A well-run workshop usually follows a sequence something like the one an industry maintenance article describes for truck filters: weigh the filter and flow test it, clean it, bake it to oxidise any remaining soot, blow it out once more, then weigh and test it again (Fleet Maintenance). Your supplier’s sequence may differ, because coated cores and some substrates need gentler steps, and an inspection of the substrate and sealing surfaces belongs at the start and the end.
The benefits are control and evidence: a repeatable process, proper extraction and a bench that measures the outcome. The costs are time, handling and logistics. A filter that leaves the building has to be removed, packed, moved, received, processed, tested, packed again and refitted, and each of these steps can add damage or delay.
Eight factors, one at a time
Factor 1: Depth of cleaning
Soot burns during regeneration, but ash does not, so ash is what a cleaning service exists to remove. Ash is distributed through the channels, and the outer channels often collect more than the centre. One industry source cited in a trade article notes that pulse cleaning assumes every cell is plugged equally, which leads to inconsistent results (Overdrive). A venue that can combine more than one method, and can repeat steps until a test shows flow has recovered, will usually clean more deeply than one that makes a single pass and stops.
What to ask. Which steps will be used? Can they repeat a step if the first pass fails the test?
Factor 2: Verification
This is the factor that most often decides the question. A trade article quoting a truck-parts manufacturer put it bluntly: testing after cleaning is the most important part of the process, because without it nobody knows whether the filter is ready to go back into service (FleetOwner). A flow bench measures how easily air passes through the element at set conditions. Compare the before and after numbers and you have a measure of restoration. Weighing the element before and after gives a second, independent figure. A photograph of a clean inlet face gives neither.
What to ask. Is there a flow or pressure-drop test before and after? Is the element weighed? What is the pass criterion, and who set it? Our article on what an RECD cleaning report should show explains how to check the numbers once you have the report.
Factor 3: Downtime
On-site service is typically measured in hours. Off-site service is measured in days, including transport. If the DG set protects a hospital, a data centre or a process that cannot stop, that difference is the reason to plan the work carefully. The usual answers are an exchange cartridge (a pre-cleaned element swapped in on the day), a scheduled maintenance window or a temporary arrangement. Running the set without the approved device in place carries compliance risk, so do not treat that as a fallback.
What to ask. Is an exchange element available for my model? What is its condition and test record? How long does the swap take?
Factor 4: Dust and safety
Fine soot and ash are easily airborne. Doing work in a plant room means dust in a space people work in, and often in a space with air intakes. Compressed-air cleaning in the open turns a local deposit into a room-wide aerosol. A California Air Resources Board guide for filter ash handling states that blowing compressed air through a filter without controls would release ash to the atmosphere and that an acceptable alternative is to blow into a chamber fitted with a HEPA filter (CARB ash guide). Workshops are generally better placed to provide extraction, though a poor workshop can be worse than a good on-site team. The deciding question is whether the dust is captured at the source.
What to ask. Where does the extracted air go? Is there a HEPA stage? What PPE and washing arrangements are used?
Factor 5: Waste control
Removed ash and soot, used filters and contaminated wipes must be collected, labelled and sent for authorised disposal under your own rules. Off-site cleaning moves this responsibility to a place with a waste stream, but only if the contract says so. On-site cleaning leaves everything at your premises until someone collects it. Either way, the contract should name who owns the waste, and you should receive a record.
Factor 6: Risk to the element
Every time an element is moved, it can be dropped, knocked or sealed badly. Thermal shock during rapid heating and cooling, mishandled water or unsuitable chemicals can damage the substrate or any catalyst coating. On the other hand, a rushed on-site job in a cramped room can do equal harm. The mitigation is the same in both cases: trained handlers, a documented procedure approved for the element and a post-clean inspection.
Factor 7: Cost
Compare total cost, not the line price. Include the exchange element or hire, transport, waste handling, any temporary arrangement and the downtime cost. A cheap on-site visit that does not restore flow, and is followed by another visit three months later, is not cheap. Our article on tracking RECD service cost per operating hour shows how to put the numbers into one log.
Factor 8: Warranty and approved method
Check the warranty terms before booking anything. Some suppliers approve specific cleaning methods or service centres for specific filters and may not accept damage from other methods. See RECD warranty terms: parts, labour and exclusions. If you are unsure, ask the supplier in writing whether the proposed provider and method are acceptable.
A weighted decision matrix with worked numbers
Numbers help you stop arguing about impressions. The matrix below is an illustrative example for a hospital with a 320 kVA standby DG set. Replace the weights and scores with your own.
Step 1: Set weights that add to 100. The hospital cares most about proof of result and continuity of supply.
| Criterion | Weight |
|---|---|
| Verification (proof of result) | 25 |
| Downtime and continuity | 20 |
| Depth of cleaning | 20 |
| Dust and safety control | 15 |
| Waste traceability | 10 |
| Cost | 10 |
| Total | 100 |
Step 2: Score each option from 1 (poor) to 5 (strong). Four options are compared: on-site vacuum only, mobile on-site cabinet, off-site cleaning with an exchange element and off-site cleaning without one.
| Criterion (weight) | On-site vacuum only | Mobile on-site cabinet | Off-site with exchange | Off-site, no exchange |
|---|---|---|---|---|
| Verification (25) | 1 | 3 | 5 | 5 |
| Downtime (20) | 5 | 4 | 4 | 2 |
| Depth (20) | 1 | 3 | 5 | 5 |
| Dust and safety (15) | 3 | 3 | 4 | 4 |
| Waste traceability (10) | 3 | 3 | 4 | 4 |
| Cost (10) | 5 | 3 | 3 | 4 |
Step 3: Multiply, add and divide by 5 to get a score out of 100.
- On-site vacuum only: 25 + 100 + 20 + 45 + 30 + 50 = 270, and 270 ÷ 5 = 54
- Mobile on-site cabinet: 75 + 80 + 60 + 45 + 30 + 30 = 320, and 320 ÷ 5 = 64
- Off-site with exchange: 125 + 80 + 100 + 60 + 40 + 30 = 435, and 435 ÷ 5 = 87
- Off-site, no exchange: 125 + 40 + 100 + 60 + 40 + 40 = 405, and 405 ÷ 5 = 81
Reading the result. For this hospital, off-site cleaning with an exchange element wins by a clear margin, and the cheap on-site option comes last because it scores poorly on the two factors the hospital weights most. Notice what would change the outcome. A site with no continuity requirement could lower the downtime weight, which narrows the gap between the two off-site options. A provider that could show a flow bench on its mobile unit would lift the mobile option’s verification score and bring it closer. The matrix does not decide for you, but it shows which assumption the decision rests on.
A must-pass rule. Some criteria should be pass or fail, not weighted. For example, “dust is captured at source” and “waste has a named, authorised destination” should be conditions for any provider, not points to be traded against price.
Downtime arithmetic: exchange or wait?
If the choice is between paying for an exchange element and keeping the set out of service, put numbers on it. The figures below are assumptions for illustration.
- Off-site turnaround: five working days
- Cost of a temporary replacement arrangement (hired set or other cover): ₹12,000 per day
- Exchange element fee quoted by the supplier: to be asked
Waiting without an exchange element means five days of cover, so about 5 × 12,000 = ₹60,000. If the exchange fee is well below that, the exchange option wins on cost as well as on continuity. If the set is a standby unit at a site with a robust grid supply and the outage risk for a week is acceptable, the arithmetic may favour waiting. The point is to calculate it, not to guess.
An exchange element has its own questions. What is its history? Was it cleaned and tested, and is the test record available? Is it the right size and coating for your device? Our guide to refurbished RECD filter core reuse sets out the acceptance checks worth applying to any element that did not come straight from the factory. Larger fleets may find a shared spare economical, as described in the shared spare RECD for a generator fleet.
When on-site is genuinely the right call
On-site service earns its place in several situations.
- A visual and sensor check between cleanings, to decide whether a full cleaning is needed. Our RECD maintenance guide lists what to look at.
- Light surface deposits on the inlet face when pressure readings are normal.
- Sensor tapping lines and ports, which block with soot and can mislead the monitoring.
- Sites where the element cannot be removed safely without a lift or crane that is not available, in which case the best option may be a planned outage with proper equipment, not a compromised on-site job.
- Emergency stabilisation, understood as temporary, with a proper cleaning planned.
On-site service is a poor fit when ΔP at a reference load stays high after a successful regeneration, because that pattern points to ash or residue inside the channels. Our article on how to tell whether an RECD filter is ash-loaded shows the signs. It is also a poor fit when the filter has been wet with oil, when there is any sign of cracking, or when the supplier says the element needs a particular process.
When off-site is the better call
Off-site cleaning is the sensible default when ash loading is confirmed, when oily or hardened deposits are present, when a flow test and weight record are needed for compliance or warranty evidence and when the process the element needs, such as a controlled thermal step or a wet method, cannot be reproduced at your site. The questions to put to a provider about thermal steps are in our article on thermal cleaning of RECD filters. Some methods, including air, are unsuitable for certain filters, as explained in why compressed-air cleaning may be unsuitable.
A site-readiness checklist for either venue
Good preparation shortens the visit, protects the element and reduces dust. Complete these steps before the provider arrives.
- Agree the scope in writing. Which jobs from the five-job table are included?
- Isolate and cool down. The DG set is stopped, locked out and allowed to cool. Exhaust parts stay hot for a long time after shutdown.
- Clear the access route. The route from the plant room to the vehicle is free, with floor protection if needed.
- Check lifting arrangements. Weights, lifting points and equipment are confirmed in advance. Rooftop sites may need extra planning; see structural support and lifting access for rooftop RECDs.
- Provide waste containers or confirm who brings them. See the waste handling article above.
- Record the “before” state. Read and note ΔP at a reference load before removal, with date, load and run hours. Photograph the housing and sensors.
- Arrange witnessing. Name someone to see the element before it is packed or tested.
- Plan the cover. Exchange element, temporary arrangement or approved downtime.
- Plan the first runs after refit. Decide in advance when you will read ΔP at the reference load and who compares it with the “before” reading.
Questions to put to either provider
Our batch of service-provider questions in the post on handling soot and ash removed during RECD servicing covers dust, PPE and waste. For the cleaning itself, add these.
- Which cleaning steps will you use on my element, and why are they suitable for its substrate and coating?
- What is measured before and after, and with which instrument?
- What result counts as a pass, and what happens if the element fails?
- Can I see a sample report from a similar job?
- What guarantee applies if pressure rises again soon after refit?
- Who is responsible for damage during handling and transport?
- What happens to the removed material, and what record will I receive?
Mistakes to avoid
- Treating “same day” as “better”. Speed is a benefit only if the job has been done properly.
- Treating “workshop” as “better”. A poor workshop with no flow bench is not an upgrade.
- Skipping the “before” reading. Without it, the improvement cannot be measured.
- Accepting a photograph as proof. A clean-looking inlet face says little about the channels behind it.
- Booking cleaning as the answer to every high-pressure reading. Sometimes the problem is a sensor, a regeneration issue or an exhaust restriction. See RECD regeneration vs filter cleaning: which service is needed.
- Forgetting the follow-up. If pressure stays high after refit, you need a structured diagnosis, not a second guess. See why RECD pressure can stay high after a cleaning service.
FAQ
Is off-site RECD cleaning always better than on-site?
No. Off-site is usually better for removing ash from channels and proving the result, because the equipment and test bench are there. On-site is better for inspection, surface cleaning and keeping downtime short. The right choice depends on what your element needs, as shown by pressure data, not on the venue’s reputation.
How long does a filter cleaning take?
It varies by method and provider. A trade article quoting a truck-parts supplier gave about 20 to 60 minutes for a standard air-based cleaning on a vehicle filter and about nine hours for a thermal cleaning (FleetOwner). Those are machine times for truck filters, not turnaround times, which include removal, transport, queueing, cool-down, testing and refitting. Ask your provider for a turnaround commitment in days for your element.
Can I clean my own RECD filter to save money?
Only if the supplier’s procedure allows it and you have the training, containment, PPE and waste route. A wrong method can damage the ceramic or coating, and in some cases the warranty. If your team is capable, ask the supplier for an approved procedure and a way to verify the result.
Does the filter have to be removed to be cleaned?
For ash removal from the channels, in nearly all cases yes. For inspection, surface cleaning and sensor port cleaning, no. If a provider claims to clean ash from the channels without removing the element, ask exactly how and how they will prove it.
What should the cleaning report include?
Before and after measurements, element weights, photographs, the method used, who did the work and where, how waste was handled and any defects found. See how to verify that the cleaning was effective.
Bringing it together
On-site and off-site RECD filter cleaning are not rivals. They are different tools. Decide what the element needs by looking at the data, break the job into its five tasks, ask which of them each quotation really delivers, score the options against your own priorities and insist on proof of result whichever venue you pick. If continuity matters, arrange an exchange element before the old one leaves the site.
If you would like help judging a cleaning quotation, or want to know what the service routine looks like for your rating, you can browse our retrofit emission control devices, read our DG set emission regulations overview or contact the Aceget team.
[…] cleaning is chosen, use the questions in RECD filter cleaning: on-site vs off-site service compared, and take the “before” readings before the element […]