- October 25, 2026
- Posted by: editor
- Category: RECD & Emission Control Devices
Ask a facility manager with generators at six sites which DG sets have a Retrofit Emission Control Device, and when each was last tested, and the answer often takes two days and a dozen phone calls. The information exists, but it is scattered across purchase files, vendor emails, service reports and the memory of one engineer who may have moved on.
An RECD asset register fixes that. It is a single, structured list of every DG set in your fleet and the emission control equipment on it, with the dates, references and readings that matter for compliance, maintenance and budgeting. Done well, it lets you answer an SPCB inspector, an auditor or your CFO in minutes.
This guide builds one in five phases: define, collect, structure, maintain and use. A worked example with a sample register extract across three sites, and the KPIs calculated from it, sits near the end.
Why a Multi-Site Fleet Needs a Dedicated RECD Register
Most organisations already have a fixed asset register, but it was built for accountants. It records purchase value and location, not the differential pressure baseline of an RECD, the firmware on its controller, or the consent reference that ties the device to a particular DG set.
A dedicated RECD register (or an RECD section within a wider DG register) matters for four reasons:
- Compliance is tied to specific sets. The CPCB type approval, the RECD certificate and the state consent relate to an identified device on an identified generator. If you cannot show which device sits on which set, your paperwork loses value during a pollution control board inspection.
- Maintenance depends on history. Whether a rising back-pressure reading is normal loading or an early warning depends on the commissioning baseline and the date of the last cleaning.
- Restrictions arrive at short notice. When GRAP stages tighten in Delhi NCR, you need to know at once which sets at which sites can run.
- Budgets need a fleet view. Planning retrofits, cleaning contracts or warranty renewals requires the whole picture, not site-by-site fragments.
The register is also the input for ranking DG sets for a phased RECD rollout.
Phase 1: Define What the Register Must Do
A register that tries to capture everything rarely survives its first year. One that captures what people actually use does.
Decide the scope
Include every DG set the organisation owns or operates, whether or not it has an RECD yet. Unfitted sets are your future workload and your compliance gap. Also decide how to treat:
- Rented sets. Include them if they run on your premises under your consent, recording the rental company as owner and your site as operator.
- Sets operated by others. Where an O&M contractor or landlord runs the set, include it and record who holds responsibility.
- Decommissioned sets. Mark them “retired” rather than deleting the row. Auditors may ask about a set that appeared in an old consent.
Decide the questions it must answer
Write down five to ten questions the register must answer without anyone opening another file. For example:
- Which sets at each site have an RECD, and which do not?
- Which RECDs are due for cleaning or testing in the next 60 days?
- Which consents, warranties or service contracts expire in the next six months?
- Which sets show a back-pressure rise above the agreed trigger?
- Which records are incomplete?
Every field should help answer at least one of these. If a field answers none, leave it out.
Phase 2: Collect the Data
The field list, in groups
The table lists the fields most fleets need. Mark each as mandatory or optional for your organisation; the grouping also helps when you assign owners later.
| Group | Fields |
|---|---|
| Identity | Asset ID, site code and address, location within site (basement DG room, rooftop), status (active, standby, retired) |
| DG data | Make, model, serial, kVA and kW rating, engine model and serial, year of manufacture, duty (standby or prime), original emission norm |
| RECD data | RECD make, model and serial, technology type, installation date, installer, exhaust connection and mounting details |
| Compliance | RECD certificate or type approval reference, consent reference and expiry, DG registration, last emission test date and result, next test due |
| Monitoring | Sensors fitted with serials and calibration dates, controller make, model, serial and firmware version, communication method, BMS point list reference |
| Maintenance | Run hours (total and last 12 months), ΔP baseline with load, latest ΔP with load and date, last cleaning date, hours since cleaning, open defects |
| Commercial | Purchase order, supplier, warranty start and end, AMC or service contract reference and expiry, spares held |
| Documents | Links to nameplate photos, GA drawing, commissioning report, test reports, certificates, consent copy, O&M manual, training record |
Two fields deserve special attention. The ΔP baseline must be recorded with the load at which it was taken, because back-pressure varies strongly with load; see comparing RECD pressure readings at different generator loads. The firmware version matters because a controller replacement or update can change scaling, thresholds and logged data, and that is easy to miss unless the register prompts for it.
Data sources
Expect to pull data from purchase orders and delivery challans (model, serial, supplier), commissioning and site acceptance reports (installation date, ΔP baseline), third-party emission test reports (PM reduction, test dates), consent documents held by the EHS team, AMC service reports (cleanings, run hours, defects), and the DG and RECD controller displays (run hours, firmware, live readings).
Verify fields against the physical asset
Paper and reality drift apart. A set may have moved between sites, an engine may have been replaced during overhaul, or a controller swapped under warranty. The reliable fix is a site walk with a camera. For every set, photograph the DG and engine nameplates, the RECD serial plate, the controller label and firmware screen, each sensor tag, and the general installation.
Name each photo with the asset ID and date, link it from the register, and compare every serial number against the photo. The guide to DG nameplate photos for an RECD enquiry describes a usable photo; the same standard applies here.
A simple rule works well: a field is not “confirmed” until a document or photo backs it. Unverified values can go in, but flag them so nobody treats them as fact.
Phase 3: Structure the Register
An asset ID naming convention
A good asset ID is short, unique, readable and permanent. A convention that suits most Indian multi-site fleets is [Organisation]-[Site]-[Asset type][Number]:
ACM-GGN-DG01: the first DG set at the Gurugram plant.ACM-GGN-DG01-RECD: the RECD fitted to that set.ACM-NDA-DG02: the second DG set at the Noida warehouse.
Some practical rules:
- Use fixed-length site codes and two-digit numbers, so lists sort cleanly.
- Never reuse a number. If DG02 is retired, the replacement becomes DG04. Otherwise old reports become ambiguous.
- Give the RECD its own ID linked to the DG. If an RECD is ever moved (see transferring an RECD to another diesel generator), its history follows its serial while the register records the new pairing and date.
- Decide how relocations work. Many fleets keep a permanent ID and record the current site in a separate field, which keeps history intact.
- Use ISO dates. Write dates as YYYY-MM-DD following ISO 8601. A date like 03/04/2026 reads as 3 April at one site and 4 March at another.
Spreadsheet or CMMS?
| Factor | Shared spreadsheet | CMMS or asset software |
|---|---|---|
| Set-up time | Hours to days | Weeks to months |
| Cost | Very low | Licence plus configuration |
| Best fleet size | Up to roughly 30–40 sets | Larger fleets or many users |
| Reminders and work orders | Manual or basic formulas | Built in |
| Audit trail | Depends on version history | Usually strong |
| Main risk | Copies multiply, formulas break | Poor adoption if too complex |
For most organisations, start with a well-structured spreadsheet on a shared drive with version history, prove which fields people use, then migrate when the fleet or user count grows. If you already run a CMMS for other plant, add the RECD fields there instead of building a parallel file.
Either way, keep one row per asset and one column per field, use drop-down lists for status, duty and site code, and keep a separate event log (cleanings, tests, readings) so the main sheet shows only the latest values without losing history.
Phase 4: Maintain the Register
Most registers fail here. They are accurate on the day they are finished and then quietly decay. Clear ownership and defined update triggers prevent that.
Data ownership
Assign an owner to each field group, not just to the register as a whole.
| Field group | Typical owner |
|---|---|
| Identity and DG data | Site facility manager |
| RECD data | Project engineer who managed the retrofit |
| Compliance | EHS lead |
| Monitoring | Site facility manager, with the BMS or controls contractor |
| Maintenance | Site maintenance lead or AMC provider |
| Commercial | Procurement |
| Documents | Whoever creates the document |
Also name one register custodian at head office who runs the monthly checks, chases gaps and controls the structure.
Update triggers
Tie updates to events rather than relying only on periodic reviews. When one of these happens, the owner updates the register within an agreed period, say five working days.
| Trigger event | Fields to update |
|---|---|
| RECD installation or commissioning | All RECD data, ΔP baseline and load, certificate, documents, warranty start |
| Routine service or cleaning | Last cleaning, run hours, hours since cleaning, latest ΔP, defects |
| Emission test | Test date, result, report link, next test due |
| Relocation of a DG set | Site, location, consent references, reinstallation records |
| Controller swap or firmware update | Controller serial, firmware, configuration backup, sensor scaling check |
| Engine overhaul or major DG change | Engine data, new ΔP baseline, retest requirement |
| Consent renewal or AMC change | Reference numbers and expiry dates |
The controller swap trigger is the one most often missed. A new controller can arrive with default settings and an empty log, and nobody tells the register. See preserving RECD data during controller replacement. Engine work may also justify a fresh baseline or a retest; see when to retest an RECD after DG set changes.
A simple enforcement method: make the register update part of job closure. The AMC provider’s service report is not accepted, and the invoice not processed, until the row is updated.
Data quality checks
Run these monthly. In a spreadsheet, most need only conditional formatting or a few formulas:
- Blank mandatory fields, counted per row and per site.
- Stale data, such as a latest ΔP or run-hour figure older than 90 days.
- Impossible values: run hours lower than the last entry, future installation dates, ΔP with no load recorded.
- Duplicates: the same RECD serial on two rows, or one DG serial at two sites.
- Unverified serials without a linked photo, and broken document links.
Share the results as a short monthly note to site managers. Visible scores across sites tend to improve data quality faster than reminders.
Worked Example: A Sample Register Extract Across Three Sites (Illustrative)
This example is illustrative. The company, sites, ratings and readings are invented to show how the register works. The reference date for all calculations is 2026-10-16.
“ACM” has six DG sets across a Gurugram plant (GGN), a Noida warehouse (NDA) and a Faridabad office (FBD). Four have RECDs; two are planned for Phase 2.
Extract A: identity, DG and RECD data
| Asset ID | Site | kVA | Duty | RECD status | RECD installed | Run hours (last 12 months) |
|---|---|---|---|---|---|---|
| ACM-GGN-DG01 | Gurugram plant | 1010 | Standby | Fitted | 2025-03-18 | 410 |
| ACM-GGN-DG02 | Gurugram plant | 1010 | Standby | Fitted | 2025-03-25 | 380 |
| ACM-GGN-DG03 | Gurugram plant | 500 | Standby | Planned | n/a | 95 |
| ACM-NDA-DG01 | Noida warehouse | 380 | Prime (part-time) | Fitted | 2025-09-10 | 620 |
| ACM-NDA-DG02 | Noida warehouse | 160 | Standby | Fitted | 2026-01-22 | 140 |
| ACM-FBD-DG01 | Faridabad office | 125 | Standby | Planned | n/a | 60 |
Extract B: maintenance, compliance and documents (fitted sets)
| RECD ID | ΔP baseline (kPa, 75% load) | Latest ΔP (kPa, 75% load) | Hours since cleaning | Last test | PM reduction | Warranty end | Docs present / required |
|---|---|---|---|---|---|---|---|
| ACM-GGN-DG01-RECD | 6.0 | 7.8 | 140 | 2025-04-15 | 76% | 2027-03-17 | 8 / 8 |
| ACM-GGN-DG02-RECD | 6.1 | 9.4 | 290 | 2025-04-15 | 78% | 2027-03-24 | 8 / 8 |
| ACM-NDA-DG01-RECD | 4.5 | 5.4 | 85 | 2025-11-20 | 74% | 2027-09-09 | 7 / 8 |
| ACM-NDA-DG02-RECD | 3.2 | 3.5 | 140 | 2026-02-12 | 81% | 2028-01-21 | 7 / 8 |
ACM’s internal rules, agreed with its supplier, are also illustrative: inspect or clean at 300 hours since cleaning, investigate when ΔP rises 50% or more above baseline at the same load, and retest each RECD every 12 months. Your supplier’s guidance and consent conditions take precedence over figures like these.
The two missing documents are the RECD nameplate photo for NDA-DG01 and the post-swap controller configuration backup for NDA-DG02, whose controller was replaced in June 2026. NDA-DG02’s firmware field still shows the old version, so the controller swap trigger also flags it.
Derived KPIs from the extract
- Coverage by count: 4 ÷ 6 = 66.7%.
- Coverage by capacity: total kVA = 1010 + 1010 + 500 + 380 + 160 + 125 = 3,185. Fitted kVA = 1010 + 1010 + 380 + 160 = 2,560. Coverage = 2,560 ÷ 3,185 = 80.4%.
- Coverage by run hours: total = 410 + 380 + 95 + 620 + 140 + 60 = 1,705 hours. Fitted = 410 + 380 + 620 + 140 = 1,550. Coverage = 1,550 ÷ 1,705 = 90.9%. This is the most meaningful figure, because emissions follow running hours, not nameplate count.
- ΔP rise above baseline: GGN-DG01 7.8 ÷ 6.0 = 1.30 (30%); GGN-DG02 9.4 ÷ 6.1 = 1.54 (54%); NDA-DG01 5.4 ÷ 4.5 = 1.20 (20%); NDA-DG02 3.5 ÷ 3.2 = 1.09 (9%). Sets above the 50% trigger: 1 (GGN-DG02).
- Cleaning interval used: GGN-DG02 is at 290 ÷ 300 = 96.7%. The others are at 46.7% (140 ÷ 300), 28.3% (85 ÷ 300) and 46.7%.
- Test currency: both Gurugram RECDs were last tested on 2025-04-15, fell due on 2026-04-15 and are six months overdue. NDA-DG01 falls due on 2026-11-20, within 60 days. NDA-DG02 is due on 2027-02-12. Current tests: 2 of 4 = 50%.
- Average PM reduction at last test: (76 + 78 + 74 + 81) ÷ 4 = 309 ÷ 4 = 77.25%, with every unit above the 70% CPCB requirement. This is a fleet summary only; each set’s compliance stands on its own result.
- Document completeness: (8 + 8 + 7 + 7) ÷ (4 × 8) = 30 ÷ 32 = 93.8%.
- Warranties expiring within six months (window to 2027-04-16): GGN-DG01 and GGN-DG02, so 2 to review now.
What the example tells ACM
From two short tables, the custodian produces an action list in minutes:
- Clean and inspect GGN-DG02 now: highest ΔP rise and nearly at its cleaning interval.
- Book retests for both Gurugram RECDs, after the GGN-DG02 cleaning and with load matched to the original test.
- Plan NDA-DG01’s retest for November.
- Upload the missing photo and configuration backup at Noida, and correct the firmware field.
- Raise the Gurugram warranty expiries with procurement before March 2027.
- Treat Phase 2 as lower priority than fixing gaps on fitted sets: the two unfitted sets account for only 155 of 1,705 hours (9.1%).
Phase 5: Use the Register
A register earns its upkeep only if people act on it.
Compliance calendar
Generate a rolling 12-month calendar from the date fields: test due dates, consent expiries, sensor calibrations, warranty ends and AMC renewals. A filtered “due in next 60 days” view, sent to site managers on the first working day of each month, is often enough.
For Delhi NCR sites, add a GRAP readiness view showing which sets have RECDs with current test reports and complete documents. The post on GRAP rules for DG sets in Delhi NCR helps decide what that view should contain.
Phasing future retrofits
The register lists unfitted sets with their run hours, duty, location and capacity, which is exactly the data needed to rank the next phase. High run hours, prime duty and stricter regions usually move a set up the list. For mixed fleets it also shows where a common specification is possible; see standardising RECD specifications across multiple factory sites.
Budgeting
Pull three numbers each year: capital (unfitted sets in the next phase, with kVA, for quotations), maintenance (expected cleanings, estimated from run hours and interval) and testing and renewals (retests, consents, warranties, AMCs). When finance asks why the budget changed, the register is the evidence.
KPIs for management
Report a small set quarterly: coverage by run hours, sets above the ΔP trigger, test currency, document completeness and upcoming expiries. Add one data quality KPI, such as the share of rows verified in the last 12 months.
Audit Readiness
An audit by an SPCB officer, corporate auditor or customer’s EHS team tests three things: does the register match reality, can you produce evidence quickly, and can you show what changed and when.
- Match reality. Keep nameplate photos current and re-verify a site’s sets before a known audit.
- Produce evidence quickly. Keep a folder per asset ID with the certificate, consent copy, commissioning report, latest test report, GA drawing and service reports, linked from the register. Aim to produce any document within five minutes.
- Show history. Keep the event log instead of overwriting values, and use version history or the CMMS audit trail. Recording who changed what and when follows the same principles as NIST SP 800-92 on log management, applied to a much simpler record.
Keep a one-page audit pack per site: that site’s rows, its compliance calendar and open actions. Inspectors respond better to an organisation that shows its gaps and the plan to close them than to one claiming to have none.
FAQ
How long does it take to build an RECD asset register?
For 10–20 DG sets across a few sites, a first usable version usually takes two to four weeks, with verification the slow part. Do not wait for 100% completeness; a register that is 80% verified and in use beats a perfect one still being built.
What is the most important field to get right?
The pairing of RECD serial to DG serial, backed by photos, together with the certificate and consent references. After that, the ΔP baseline with its load, because almost every later diagnosis depends on it.
Who should own the register in a multi-site organisation?
Split ownership works best: site teams own their assets’ data, specialists own their field groups, and one central custodian owns the structure and checks. A register with only a central owner goes stale; one with only site owners fragments.
Can the register take data directly from RECD controllers?
If your RECD controller offers a data connection to a BMS or monitoring platform, ΔP, temperatures and run hours can be pulled automatically. Even then, keep the register as the master record for identity, compliance and documents. See integrating RECD monitoring with a building management system.
Building Your Register Around Reliable Equipment
A register is only as useful as the equipment and documents it describes. Aceget supplies CPCB type-approved Retrofit Emission Control Devices for DG sets from 25 kVA to 2500 kVA, with the certification documents your register needs to reference. Review our certifications, or contact our team to discuss RECD planning across a multi-site fleet.