October 16, 2026
Posted by: editor
Category: RECD & Emission Control Devices

“Our RECD reduces particulate by 74%” is a useful statement for an engineer. It means much less to a board, a sustainability committee, a lender or a resident welfare association. They tend to ask a simpler question: how much pollution has it actually stopped?

Answering that means turning a percentage into a quantity, kilograms of particulate matter (PM) that would have left the chimney but didn’t. The arithmetic is simple. Getting a figure you can defend takes more care, because the result depends on how much the generator runs, at what load, and how dirty its exhaust was to begin with.

This guide explains how to estimate kilograms of PM avoided after an RECD retrofit, using three methods, worked examples for a single DG set and a multi-set campus, a sensitivity check, and guidance on what you can and can’t claim.

The Core Formula

Every method comes down to the same idea:

PM avoided = PM the engine would have emitted without the RECD − PM it actually emits with the RECD

Since the RECD removes a fraction of the engine’s particulate, this simplifies to:

PM avoided (g) = Baseline PM emission rate × Reduction fraction × Amount of operation

The three inputs are:

Baseline PM emission rate: how much particulate the engine emits without the RECD, per unit of operation (g/kWh, g/h or g/litre of fuel)
Reduction fraction: the share of that particulate the RECD removes, as a decimal (for example 0.74)
Amount of operation: kWh generated, hours run, or litres of fuel burned over the period

The three methods below use different units for the first and third inputs. If done carefully, they should give similar answers.

Getting the Inputs Right

Input 1: The baseline emission rate

This is the input that matters most and is most often guessed. In order of preference:

Measured on your engine. A baseline from your own acceptance or before-and-after test, at the load your set normally runs. This is the best source. See planning an RECD emissions test before and after installation.
Engine type-approval or OEM data. The emission certificate for your engine model gives a weighted g/kWh value under test conditions. It is useful, but a new engine in a test cell is often cleaner than a worn engine on site.
Published emission factors. Generic factors for diesel engines, such as those in the US EPA’s AP-42 compilation of emission factors, are a last resort. They are averages across many engines and may not match Indian DG sets or their condition.

Whichever source you use, state it. And make sure the baseline reflects the load your set actually runs at, because specific emissions in g/kWh change with load. See why generator load must be matched in RECD comparison tests.

Input 2: The reduction fraction

Again, a measured value on your installation is best:

Site acceptance test result at the typical operating load
Type-approval result for the RECD model. Use this with care. It is a weighted multi-load figure on a test engine, and your set may run mostly at light load, where many devices perform below their weighted average
The regulatory minimum (70%) as a conservative floor, if nothing better is available

If your acceptance test showed different reductions at different loads, use the one that matches how the set actually operates, or weight them by time spent at each load. See how to calculate RECD particulate reduction.

Input 3: Amount of operation

Choose the record you trust most:

Hour meter readings, which almost every DG has. Read and record them at fixed intervals. See how to check a used generator’s engine hours for reading tips.
kWh from the DG controller or an energy meter, which is ideal for the g/kWh method
Fuel purchase or consumption records, which are often the most complete for sites that track diesel spend

Also note the average load during running hours, from your DG load logs. Without it, hours alone can’t be turned into kWh.

Method 1: Energy-Based (g/kWh × kWh)

Best when: you have a baseline in g/kWh and kWh records.

PM avoided (g) = Baseline (g/kWh) × Reduction fraction × Energy generated (kWh)

Worked example: a single 250 kVA set

The figures are illustrative.

DG set: 250 kVA, about 200 kW electrical rating
Annual running hours (hour meter): 600 h
Average load while running: 90 kW (45% of rating)
Energy generated: 600 × 90 = 54,000 kWh/year
Baseline PM at about 45% load, from the site test: 0.30 g/kWh
Measured reduction at that load: 74%

PM avoided = 0.30 × 0.74 × 54,000 = 11,988 g ≈ 12.0 kg/year

Method 2: Mass-Rate Based (g/h × hours)

Best when: your test report gives emissions as a mass rate (g/h) at the typical load, and you have hour meter readings.

PM avoided (g) = (Baseline g/h − Post-RECD g/h) × Running hours

Same example by Method 2

From the site test at 90 kW:

Baseline mass rate: 0.30 g/kWh × 90 kW = 27.0 g/h
Post-RECD mass rate: 27.0 × (1 − 0.74) = 7.0 g/h
Avoided per hour: 27.0 − 7.0 = 20.0 g/h

PM avoided = 20.0 × 600 = 12,000 g ≈ 12.0 kg/year

The same result as Method 1, as it should be, since the inputs come from the same test.

Method 3: Fuel-Based (g/litre × litres)

Best when: fuel records are your most reliable data, and kWh or hours are patchy.

You need the baseline PM per litre of fuel burned. You can derive it from a g/kWh baseline and the engine’s fuel efficiency at the typical load:

Baseline (g/litre) = Baseline (g/kWh) × kWh produced per litre

Same example by Method 3

At 90 kW the engine burns about 24 litres/hour, which is about 3.75 kWh per litre (check against your own fuel and load records; see diesel consumption in generators)
Annual fuel: 600 h × 24 L/h = 14,400 litres
Baseline PM per litre: 0.30 × 3.75 = 1.125 g/litre

PM avoided = 1.125 × 0.74 × 14,400 = 11,988 g ≈ 12.0 kg/year

All three methods agree because they share the same underlying data. In practice, small differences between methods are a useful cross-check. A large disagreement usually means one of the records (hours, load, kWh or fuel) is wrong.

Scaling Up: A Multi-Set Campus

Large sites, such as hospitals, universities, malls and industrial estates, often run many sets of different sizes and duty cycles. Calculate each set or group of identical sets separately, then add them up. Never apply one average across the whole fleet.

The figures are illustrative.

Set group	Rating (kW)	Qty	Hours/yr each	Avg load (kW)	kWh/yr (group)	Baseline g/kWh	Reduction	PM avoided (kg/yr)
500 kVA	400	2	300	200	120,000	0.20	78%	18.7
160 kVA	128	3	400	51.2	61,440	0.35	72%	15.5
62.5 kVA	50	3	250	17.5	13,125	0.45	70%	4.1
Total		8			194,565			38.3

Two things stand out:

The smaller, older sets have higher baselines per kWh. They run at lighter loads and are often older engines. Their contribution per kWh avoided is larger than their size suggests.
The large sets dominate the total because they generate most of the energy, even though they run fewer hours.

Sector-specific guides such as RECDs for university campuses, RECDs for hospitals and RECDs for housing societies discuss how duty cycles vary by site type.

How Sensitive Is the Estimate?

Every input carries uncertainty. A simple way to see which inputs matter most is to vary each one between a reasonable low and high value while keeping the others at their central value. Using the 250 kVA example:

Input varied	Low value	High value	PM avoided (low)	PM avoided (high)
Running hours	450 h	750 h	9.0 kg	15.0 kg
Average load	75 kW	105 kW	10.0 kg	14.0 kg
Baseline PM	0.22 g/kWh	0.38 g/kWh	8.8 kg	15.2 kg
Reduction	70%	78%	11.3 kg	12.6 kg
Central estimate			12.0 kg

The reduction percentage, which gets most of the attention, has the smallest effect on the total. Running hours and the baseline emission rate matter far more. If you want a better estimate, improve those records first: read the hour meter regularly, and get a measured baseline at your real operating load.

If every input were at its low value at once, the estimate would be about 5.2 kg. If every input were at its high value, it would be about 23.3 kg. Those extremes are unlikely to happen together, but they show why a single figure without a range can mislead.

Special Cases

Estimating a full year from part-year data

Reporting deadlines often arrive before the year ends. If you have hour meter readings for part of the year, scale them up using last year’s seasonal pattern, not a simple pro-rata. DG use is rarely even across the year: it often rises in summer with grid outages, and in Delhi NCR it can fall sharply in winter when GRAP restrictions limit DG operation.

For example, with illustrative figures: last year, the set ran 210 hours from April to July out of 620 hours in the full year, so those months accounted for about 33.9% of annual running. This year, April to July shows 180 hours.

Projected annual hours = 180 ÷ 0.339 ≈ 531 hours

A simple pro-rata (180 × 12 ÷ 4 = 540 hours) gives a similar figure here, but in years with an unusual summer the two can differ widely. State clearly that the figure is a projection, and update it when the full-year data is available.

Sets with no test data at all

For a set that was retrofitted without a site test, you can still produce an indicative figure, as long as it is clearly labelled.

For example, with illustrative figures: a 125 kVA set (about 100 kW) runs 350 hours a year at an average of 40 kW, generating 14,000 kWh. Using the engine’s certification value of 0.20 g/kWh as the baseline and the regulatory minimum of 70% as the reduction:

PM avoided ≈ 0.20 × 0.70 × 14,000 = 1,960 g ≈ 2.0 kg/year (indicative)

Two cautions apply. A worn engine on site is often dirtier than its certification value, so this approach tends to understate the baseline. And the device’s real reduction at 40% load is unknown. Treat the figure as a conservative placeholder until a site test is done.

Reporting the Figure Honestly

A defensible PM-avoided statement includes:

The figure as a range or with a central estimate and range, for example “about 12 kg per year (likely range 9–15 kg)”
The period it covers, such as a financial year
The method used (energy, mass-rate or fuel based)
The source of each input, such as “baseline from site acceptance test, 12 March 2026, NABL-accredited lab”
The assumptions, such as average load or kWh per litre
Exclusions, such as sets not fitted with RECDs

A short statement for an annual report might read:

“RECDs fitted to eight diesel generator sets prevented an estimated 38 kg of particulate matter from being emitted in FY 2026–27 (likely range 30–48 kg). Estimates are based on site acceptance test results, hour meter and energy records, and are calculated per set.”

What Not to Claim

Some claims sound impressive but can’t be supported by a PM-avoided calculation:

Health outcomes such as lives saved or illnesses avoided. These need specialist health-impact modelling, local exposure data and population information. A kilogram figure on its own can’t support them.
Carbon dioxide or climate benefits. An RECD reduces particulate, CO and HC. It doesn’t reduce CO₂, and the extra back-pressure can increase fuel use slightly. Keep PM and carbon reporting separate.
City-wide AQI improvement. A single site’s contribution is real but small relative to all sources of urban pollution. See causes of air pollution in Indian cities.
Type-approval percentages applied to every operating condition. As explained above, light-load sets may perform below the weighted type-approval figure.
Double counting. If a set runs on gas or in dual-fuel mode for part of the year, only count the diesel hours. See dual-fuel generators explained.

Putting the Number in Context

A dozen kilograms a year can sound small next to tonnes of CO₂. But particulate from diesel exhaust is mostly fine, combustion-generated particles, emitted close to the ground in dense urban areas, often near people. The weight is low but the exposure relevance is high. That is why diesel particulate is regulated so closely. See PM2.5 vs PM10 and how air pollution affects children’s health.

Fair context comparisons are:

Against the site’s own baseline: “Our DG sets now emit roughly a quarter of the particulate they did before retrofit.”
Over the device’s life: 12 kg per year over 8 remaining years of engine life is about 96 kg.
Across a portfolio: totals across all sites of an organisation.

Avoid comparisons with vehicles or other sources unless you have a sound, sourced basis.

Keeping the Estimate Current

The estimate is only as good as its inputs, and they change:

Running hours change with grid reliability, seasons and GRAP restrictions in Delhi NCR
Engine condition changes the baseline over time. See how fuel and engine condition can confound RECD performance tests
Device performance can change. Periodic checks show whether the reduction fraction still holds. See the RECD maintenance guide

Recalculate annually, using the year’s actual hours or kWh, and refresh the baseline and reduction inputs whenever a new test is done.

A Simple Calculation Template

Field	Value	Source
DG set ID / rating (kW)		Nameplate
Reporting period
Running hours in period		Hour meter readings (start/end)
Average load while running (kW)		Load logs / controller
Energy generated (kWh)		Controller / energy meter, or hours × average load
Baseline PM (g/kWh) at typical load		Test report reference
Reduction fraction at typical load		Test report reference
PM avoided (kg)	= baseline × reduction × kWh ÷ 1,000
Cross-check method and result		Fuel or mass-rate method
Likely range (low–high)		Sensitivity on hours and baseline

Frequently Asked Questions

Can I calculate PM avoided without any test data?

You can make a rough estimate using engine certification data or published emission factors and the 70% minimum reduction, but label it clearly as indicative. A measured baseline and reduction on your own set make the figure far more defensible.

Why does my estimate seem so small?

DG sets used for standby often run only a few hundred hours a year, and modern engines emit fractions of a gram per kWh. Small mass doesn’t mean small impact, because diesel particulate is fine and emitted near people. Report the figure honestly and give context instead of inflating it.

Which method is most accurate?

The one that uses your most reliable records. Energy-based is usually best if you have kWh data and a g/kWh baseline. Fuel-based is useful when fuel records are complete. Running two methods and comparing them is a good check.

Should I include the RECD’s effect on CO and HC too?

You can, using the same approach with baseline CO and HC emission rates and their measured reductions. Report each pollutant separately, and don’t add them together with particulate.

What about periods when the RECD was bypassed or out of service?

Exclude them. If the device was removed for cleaning, bypassed during a fault, or the set ran before the RECD was installed, those running hours didn’t benefit from the device. Subtract them from the operating total, using maintenance logs and hour meter readings taken at the start and end of each outage. Including them overstates the benefit and is easy for an auditor to spot.

How should I report PM avoided over several years?

Calculate each year separately with that year’s actual running hours, and refresh the baseline and reduction inputs whenever a new test is done. Then report the cumulative total alongside the annual figures. Don’t multiply one year’s result by the number of years, because running hours, engine condition and device performance all change over time.

Can I use this figure for CSR or ESG reports?

Yes, if you document the method, sources and assumptions, and present a range. Avoid claims beyond what the calculation supports, such as health outcomes or carbon savings.

Turn Your RECD’s Performance Into a Number You Can Defend

A PM-avoided figure built on measured data and honest assumptions is a valuable addition to compliance and sustainability reporting. Aceget supplies CPCB type-approved Retrofit Emission Control Devices for DG sets from 25 kVA to 2500 kVA. We help clients gather the test data and operating records needed for a reliable estimate. Read about DG set emission regulations and the CPCB framework, or contact our team with details of your DG fleet.

Diesel Generator Emissions Emission Calculation ESG reporting PM avoided RECD sustainability reporting

Estimating Kilograms of PM Avoided After an RECD Retrofit