How to Calculate RECD Particulate Reduction from Test Results

An RECD test report can run to a dozen pages of tables: mode loads, filter masses, dilution ratios, flow rates, corrected concentrations. Somewhere in there is the one number everyone wants, the percentage of particulate matter the device removes. This tutorial shows you how that number is built, so you can check it yourself instead of taking the summary page on trust.

You will need the report’s mode-by-mode results and a calculator or spreadsheet. If you are still at the stage of planning the tests, start with our guide to planning RECD emission tests before and after installation, then come back here when the data arrives.

What You Are Calculating, and the Target

The figure you are after is the particulate matter capture efficiency: the share of particulate mass that the RECD prevents from leaving the exhaust.

For in-use DG sets up to 800 kW, State Board directions call for a CPCB-certified device with a capture efficiency of at least 70%, tested on the ISO 8178 D2 cycle by a CPCB-recognised laboratory (CECB reply before the NGT). Earlier circulars framed the same target as a minimum 70% reduction in particulate matter without adversely affecting other emission parameters (BPCL tender on GeM citing the MPCB circular).

The calculation therefore needs particulate mass. Opacity or optical readings will not do, for reasons we cover in smoke opacity vs particulate mass and can a handheld AQI monitor verify RECD performance.

Gather These Values from the Report

Before doing any maths, pull these values into one place. If any are missing, ask the laboratory or vendor for them.

Value	Symbol	Unit	Where it usually appears
Engine power at each test mode	P	kW	Test conditions or mode summary table
Weighting factor for each mode	W	–	Test cycle description
PM mass rate, baseline (no RECD), each mode	ṁ base	g/h	Particulate results, baseline test
PM mass rate, with RECD, each mode	ṁ RECD	g/h	Particulate results, device test
Weighted specific PM, baseline	PM base	g/kWh	Summary page
Weighted specific PM, with RECD	PM RECD	g/kWh	Summary page
Engine identification and fuel used	–	–	Test set-up section

Some reports give PM per mode directly in g/kWh rather than g/h. If so, multiply each by the mode power to get g/h before weighting. The reason is explained in Mistake 2 below.

Step 1: Check the Two Tests Are Comparable

A reduction is only meaningful if the only real difference between the two tests is the RECD. Confirm:

• Same engine, or an accepted equivalent. State Board circulars describe testing on an engine within ±25% of the rated power or swept volume of the engine the device will be used on (TeamLease RegTech on the MPCB circular).
• Same test cycle and mode loads. For constant-speed gensets, the ISO 8178 D2 cycle.
• Same fuel and lubricating oil. Fuel sulphur alone can shift PM results.
• Same sampling system and filter type.
• Ambient conditions within the method’s correction limits.

Our guides on how to read an RECD test certificate and verifying a CPCB-tested RECD cover the document checks in more detail.

Step 2: Set Up the D2 Weighting

The D2 cycle tests a constant-speed engine at five loads. Each mode is given a weighting factor that reflects how gensets typically operate.

Mode	Load (% of rated power)	Weighting factor W
1	100	0.05
2	75	0.25
3	50	0.30
4	25	0.30
5	10	0.10
Total		1.00

The weights sum to 1.00, and 60% of them sit at 50% and 25% load. A device that performs well only near full load will struggle to reach a good weighted figure.

Step 3: Calculate Weighted Specific PM for Each Test

For each test (baseline and with RECD), multiply each mode’s PM mass rate by its weight, add them up, and divide by the sum of each mode’s power multiplied by its weight:

PM\ (g/kWh) = \frac{\sum_{i=1}^{5} \dot{m}_{i} \times W_i}{\sum_{i=1}^{5} P_i \times W_i}

This is a ratio of sums, not an average of ratios. It gives the grams of particulate per kWh the engine would produce if it spent its time in the proportions the cycle assumes.

Step 4: Calculate the Reduction

Reduction\ (\%) = \left(1 - \frac{PM_{RECD}}{PM_{base}}\right) \times 100

This is the same as (PM base − PM RECD) ÷ PM base × 100, written in a form that is quicker to type into a spreadsheet.

Worked Example 1: Laboratory D2 Test on a 160 kW (200 kVA) Engine

The values below are illustrative, chosen to demonstrate the method. They do not represent any specific product. For converting between kVA and kW, see how to read a generator’s kVA and kW rating.

Mode	P (kW)	W	P × W	Baseline ṁ (g/h)	Baseline ṁ × W	RECD ṁ (g/h)	RECD ṁ × W
1	160	0.05	8.0	30.0	1.50	6.0	0.30
2	120	0.25	30.0	21.0	5.25	5.0	1.25
3	80	0.30	24.0	15.0	4.50	4.2	1.26
4	40	0.30	12.0	9.5	2.85	3.0	0.90
5	16	0.10	1.6	6.0	0.60	2.2	0.22
Sum			75.6		14.70		3.93

Weighted baseline PM: 14.70 ÷ 75.6 = 0.194 g/kWh

Weighted PM with RECD: 3.93 ÷ 75.6 = 0.052 g/kWh

Reduction: (1 − 0.052 ÷ 0.194) × 100 = 73.3%

A neat shortcut: because both tests share the same power and weights, the denominator cancels. The reduction is simply 1 − 3.93 ÷ 14.70, which gives the same 73.3%. That is a handy cross-check on any report.

Step 5: Read the Mode-by-Mode Pattern

The weighted figure is the compliance answer, but the individual modes tell you how the device will behave on your site. Divide each mode’s RECD mass rate by its baseline mass rate:

Mode	Load	Mode reduction
1	100%	80.0%
2	75%	76.2%
3	50%	72.0%
4	25%	68.4%
5	10%	63.3%

Capture falls as load drops. That pattern is common, because exhaust at low load is cooler and carries more unburnt organic material. The device passes on the weighted figure, yet a set that idles at 10 to 25% load most of the time will see real-world capture nearer the lower end.

If that describes your generator, compare it with your actual operating profile using DG load logs to assess RECD suitability, and read choosing an RECD for a generator running at low load.

Worked Example 2: Site Test Upstream and Downstream of an Installed RECD

Sometimes you have a site stack report rather than a laboratory D2 certificate. A testing agency samples before and after the device at a single load and reports concentrations in mg/Nm³ plus gas flow. Here is how to calculate a reduction properly from that kind of data.

Sampling point	PM concentration (mg/Nm³)	Gas flow (Nm³/h)	PM mass rate (g/h)
Upstream of RECD	90	1,100	90 × 1,100 ÷ 1,000 = 99.0
Downstream of RECD	30	1,150	30 × 1,150 ÷ 1,000 = 34.5

Mass-based reduction: (1 − 34.5 ÷ 99.0) × 100 = 65.2%

Concentration-only reduction: (1 − 30 ÷ 90) × 100 = 66.7%

The downstream flow is higher, which can happen when air leaks in at joints or through a sampling port. That extra air dilutes the gas, so a concentration-only comparison flatters the device by 1.5 percentage points here, and by much more if leakage is larger. Always convert to mass rate first. Our post on mg/Nm³ vs g/kWh explains concentration, flow and oxygen correction in detail.

Keep two limits in mind. A single-load site result is not a D2-weighted figure, so it cannot be compared directly with a certificate’s 70% claim. And upstream and downstream samples should ideally be taken at the same time, or at least at the same steady load, because emissions shift minute to minute. Good access to sampling ports makes this easier; see safe access to RECD sampling ports.

A Spreadsheet Layout You Can Reuse

Set up a sheet like this once and reuse it for every report. Enter your data in columns A to E, rows 2 to 6.

Column	Heading	Entry or formula (row 2, copy down to row 6)
A	Mode	1 to 5
B	Power P (kW)	From report
C	Weight W	0.05, 0.25, 0.30, 0.30, 0.10
D	Baseline PM (g/h)	From report
E	RECD PM (g/h)	From report
F	P × W	=B2*C2
G	Baseline × W	=D2*C2
H	RECD × W	=E2*C2
I	Mode reduction	=1-E2/D2 (format as %)

Then add the totals and results in row 7 and below:

Cell	What it shows	Formula
F7	Sum of P × W	=SUM(F2:F6)
G7	Sum of baseline × W	=SUM(G2:G6)
H7	Sum of RECD × W	=SUM(H2:H6)
G9	Weighted baseline PM (g/kWh)	=G7/F7
H9	Weighted RECD PM (g/kWh)	=H7/F7
H10	Weighted reduction	=1-H7/G7 (format as %)
C8	Weight check, must equal 1	=SUM(C2:C6)

If your report gives each mode in g/kWh, add a helper column to convert: g/h = g/kWh × P.

Four Mistakes That Change the Answer

Using Worked Example 1, here is what each common error produces, compared with the correct 73.3%.

Mistake	What it looks like	Result	Why it is wrong
1. Averaging mode reductions	(80.0 + 76.2 + 72.0 + 68.4 + 63.3) ÷ 5	72.0%	Ignores both the weights and the larger mass at higher loads
2. Weighting g/kWh values directly	Σ(g/kWh × W) for each test, then compare	70.5%	Gives low-power modes far too much influence; the cycle formula weights mass and power separately
3. Quoting the best mode	Mode 1 alone	80.0%	Full load carries only 5% of the weight
4. Mixing units between tests	Baseline in mg/Nm³, RECD in g/kWh	Meaningless	Concentration and specific mass cannot be subtracted

Mistake 2 is the most dangerous, because it looks rigorous. In this example it nearly erases the margin above 70%. On another engine it could push a passing device below the line, or a failing one above it.

Interpreting a Borderline Result

Every measurement has uncertainty. Gravimetric PM testing at low emission levels is particularly sensitive, because the filter mass gain can be small. Suppose a report shows a reduction of exactly 71%, and the uncertainty on each weighted PM figure is around 5% for the baseline and 10% for the much lower RECD result. Pushing both in unfavourable directions gives about 66%; pushing both favourably gives about 75%.

That does not mean the result is invalid. Certification is based on the laboratory’s result under the method, not on a buyer’s recalculated range. But it does mean:

• A device certified with a comfortable margin above 70% is a safer choice than one that cleared it narrowly.
• On your own engine, especially an older or lightly loaded one, real performance may land below the certified figure.
• Performance guarantees in your purchase order should be tied to clear test conditions. See RECD performance guarantees and operating conditions and measurable RECD tender specifications.

If the Number Comes In Below 70%

Work through the likely causes in this order before blaming the device:

1. Check the arithmetic using the spreadsheet above.
2. Check comparability of the two tests (Step 1).
3. Check for leakage or bypass, especially on site tests where downstream flow or oxygen is higher than upstream.
4. Check engine condition. High oil consumption or poor combustion can overload any device; see engine oil consumption and RECD suitability.
5. Check device sizing against your engine’s exhaust flow, covered in OEM exhaust flow data for specifying an RECD.
6. Check maintenance status, using our RECD maintenance guide.

Worked Example 3: When the Report Lists g/kWh per Mode

Some reports skip mass rates and list a specific emission for each mode. Using the same engine as Worked Example 1, the baseline might appear like this:

Mode	P (kW)	Baseline PM (g/kWh)	Converted: g/kWh × P = g/h
1	160	0.1875	30.0
2	120	0.1750	21.0
3	80	0.1875	15.0
4	40	0.2375	9.5
5	16	0.3750	6.0

The converted column matches the g/h values from Worked Example 1, so the weighted calculation proceeds exactly as before and gives the same 0.194 g/kWh baseline. Do the same conversion for the RECD test, then apply Steps 3 and 4.

The trap is to skip the conversion and weight the g/kWh column directly. That gives 0.218 g/kWh for the baseline instead of 0.194, because the 10% load mode, with its high g/kWh but tiny power, gets far more influence than it should. That is Mistake 2 in the table above.

Sanity Checks Before You Trust the Summary Page

A two-minute review catches most reporting errors. Before accepting a summary figure, check:

• Weights add up to 1.00. If they do not, the wrong cycle or a typing error is in play.
• Mode powers match the stated load percentages. On a 160 kW engine, mode 3 should be close to 80 kW. Large gaps suggest the engine could not hold load or the wrong rating was used.
• Every RECD mode is lower than its baseline mode. If any mode shows higher PM with the device, ask why. It may point to a measurement issue at that mode.
• The denominator cross-check works. Divide the sum of weighted RECD mass rates by the sum of weighted baseline mass rates. One minus that ratio should equal the reported reduction.
• Units are consistent across both tests. Both in g/h or both in g/kWh before weighting, never one of each.
• Rounding is sensible. A reduction reported to two decimal places from filter masses measured at the limit of a balance’s precision suggests false confidence.

If a report fails any of these checks, it does not necessarily mean the device is poor. It means the report needs clarifying before you rely on it. Our guide on reviewing deviations in an RECD technical offer covers how to raise such points with a vendor.

Turning the Result into a Decision

The number is only useful if it leads to the right action. Here is how to read the outcome in practice.

Weighted reduction	What it suggests	Sensible next step
Comfortably above 70%, with strong low-load modes	Device performs well across the cycle	Confirm the certificate covers your engine size, then proceed
Above 70%, but low-load modes well below it	Passes certification, may underperform on lightly loaded sets	Compare with your load logs; discuss guarantees at your typical load
Close to 70%	Passes, but little margin for engine wear or variation	Prefer devices with more margin, or tighten maintenance and monitoring
Below 70% on a certification test	Does not meet the requirement	Do not rely on it for compliance
Below 70% on a site test of an installed device	Possible leakage, engine issue, sizing or maintenance problem	Work through the troubleshooting list above before concluding

When the decision is a purchase, it usually needs an internal sign-off. Our template for an internal approval note for an RECD purchase shows how to present the test evidence to management.

Keep a Record of Your Calculation

Whoever reviews the RECD next, whether an auditor, an inspector or a new facility manager, should be able to follow your working. Save the spreadsheet alongside the test report, and note:

• the report reference number and laboratory
• the engine and device serial numbers it relates to
• the date you did the calculation and who checked it
• any questions raised with the laboratory or vendor and their answers

This takes ten minutes and saves a great deal of time when documents are requested at short notice, for example during a Pollution Control Board inspection or when renewing a DG set pollution certificate.

Frequently Asked Questions

Is the 70% requirement measured on each mode or on the weighted result?

The capture efficiency is assessed on the ISO 8178 D2 cycle, which produces a weighted result across the five modes. Individual modes can be above or below 70% while the weighted figure passes.

Can I calculate reduction from opacity readings?

No. Opacity is not a mass measurement, and percentage changes in opacity do not equal percentage changes in particulate mass. The requirement is defined on particulate matter.

Why use g/h rather than mg/Nm³ in the weighted formula?

Because the cycle combines emissions across loads in proportion to mass emitted and work done. Concentration ignores how much gas flows at each load, and dilution can change it without any change in actual emissions.

Do I need both baseline and RECD tests on the same engine?

The two tests must be comparable. Certification circulars allow testing on an equivalent engine within a defined range of power or swept volume, but baseline and device results should come from the same test set-up.

What if the report only gives the final efficiency?

Ask for the mode-by-mode data. A reputable laboratory report will include it, and you need it to see low-load performance and to verify the summary figure.

Should the RECD reduce other pollutants too?

The requirement focuses on particulate matter. Circulars have also said the device should achieve this without adversely affecting other emission parameters, so a test report should show that gaseous emissions did not worsen with the device fitted.

The Bottom Line

Calculating RECD particulate reduction is simple once you follow the cycle’s logic: weight the mass rates, weight the power, divide, then compare baseline with device. The traps are in the shortcuts: averaging percentages, weighting g/kWh values directly, or comparing concentrations without flow.

Run the numbers yourself, look at the low-load modes, and insist on mass-based data. If you would like help reviewing a test report or choosing a certified device, explore Aceget’s retrofit emission control devices, see our type approval certificates, or get in touch.