How to Track RECD Service Costs per Operating Hour

Ask most facility teams what their emission control device costs to run and you will hear one figure: the annual AMC value. That number is easy to find in the purchase order, but it tells you almost nothing. It ignores the cleaning that was billed separately, the sensor that failed in May, the emission test, the gasket kits, the hours your own electrician spent on site and, eventually, the replacement core. Tracking RECD service cost per operating hour fixes that by putting every rupee spent on the device against the hours the DG set actually ran.

The difference matters most when you run a mix of sets. A standby set in an office tower might run 80 to 150 hours a year. A prime or near-prime set at a factory with weak grid supply might clock well over 1,000. Give both the same AMC and the annual figures look identical, while the real cost of each running hour can differ by a factor of three or four.

This tutorial builds that metric as a monthly spreadsheet: cost buckets, trustworthy run hours, columns and formulas, an illustrative three-site example, and how to read the results.

What “RECD service cost per operating hour” actually measures

The core definition is simple:

Cost per operating hour = total RECD-related cost in the period ÷ DG run hours in the same period

Both halves must cover exactly the same dates, taken from meter readings rather than estimates.

Useful variants and when to use each

The plain per-hour figure is not always the best lens. Four variants are worth having in the same sheet.

VariantFormulaBest used for
Cost per operating hourTotal cost ÷ run hoursTrend for one set
Cost per 100 run hoursCost per hour × 100Management reports
Cost per kWh generatedTotal cost ÷ kWhPrime sets with varying load
Cost per set-yearTotal cost ÷ sets ÷ yearsLow-hour standby sets

Standby sets running under about 200 hours a year will always show a high cost per hour because fixed costs are spread thinly; the worked example shows how large that effect is.

If your team is unsure which of its sets count as standby and which as prime, our explainer on standby vs prime power generators sets out the distinction before you start grouping sites.

Step 1: Decide which costs go into the RECD buckets

The most common mistake is inconsistent scope between sites. Agree the buckets once, write them down, and apply them everywhere.

The bucket table

BucketTypical examplesInclude?
AMC / visitsContract value, per-visit and travel chargesYes
CleaningElement cleaning, ash removal, core transportYes
ConsumablesGaskets, clamps, sealing rings, insulation repairYes
Sensors and partsSensors, wiring, controller parts, where fittedYes
Core replacementFilter element, catalyst module or cartridgeYes, amortised
TestingEmission or opacity tests, lab chargesYes
Internal labourStaff hours for isolation and supervisionYes, at a standard rate
Waste disposalRemoved soot, ash, used gasketsYes
Downtime costHired DG, lost productionOptional, separate column
Extra fuelFuel penalty from high back-pressureOptional, separate column

Downtime and extra fuel are estimates, not invoices, so keep them out of the main total and report them as a second line.

Disposal is easy to forget because it often arrives as a separate small invoice months later. Our sibling guide on handling soot and ash removed during RECD servicing explains what to ask for on that paperwork so the cost can be traced back to the right visit.

Core replacement: capitalise or expense?

A replacement core is the one large, lumpy cost in the life of most RECDs. If you book it as an expense in the month it arrives, that month’s cost per hour will spike and the trend line becomes useless.

The cleaner approach, whatever your accounts team decides about capitalisation for tax and books, is to amortise the core over its expected operating hours in the tracking sheet. If a core costs ₹2,40,000 and the supplier’s guidance, or your own history, suggests it will last 6,000 run hours, each run hour carries ₹40 of core cost. You add that ₹40 per hour every month, whether the core is replaced this year or in five years.

Record where the expected life came from and revisit it yearly. On very low-hour standby sets, cores can age by calendar time as much as by hours, so amortise over years if that gives a higher figure. If you are weighing whether a failing unit needs a new core or a whole new assembly, our guide to RECD filter core replacement vs complete assembly shows how to compare those two costs over several years.

Step 2: Get run hours you can defend

The denominator is where most cost-per-hour sheets quietly go wrong. A cost figure backed by invoices divided by a guessed run-hour figure produces a confident-looking number that is simply wrong.

Sources of run hours, in order of preference

  1. Controller log export, taken on the last day of each month.
  2. Hour meter reading on a fixed date, photographed and filed.
  3. Operator logbook, useful for reconciling but prone to drift.
  4. Fuel back-calculation, only to fill a gap.

Our earlier guide to the RECD shift handover log includes a run-hour field that feeds neatly into this sheet.

Reconciling gaps and oddities

Compare the hour meter difference with the logbook total each month. Larger gaps usually mean unlogged exercise runs, outages spanning shifts, late readings or a failed meter. Write dates unambiguously; the ISO 8601 date format (YYYY-MM-DD) removes the ambiguity entirely.

When the hour meter is replaced

Record three things on the day: the old meter’s final reading, the new meter’s starting reading, and the date. That month’s run hours become (old final − month opening) + (new closing − new start). Add two helper columns for the swap month and use =IF(S2="",C2-B2,(S2-B2)+(C2-T2)) in column D. Without this, the sheet shows a negative month or a sudden thousand-hour jump.

Step 3: Build the sheet

Use one tab per DG set, one row per month, an Assumptions tab for rates, and identical column order at every site.

Assumptions tab

CellItemExample value
B2Replacement core cost (₹)2,40,000
B3Expected core life (run hours)6,000
B4Core amortisation rate (₹ per hour)=B2/B3
B5Internal labour rate (₹ per hour)450
B6Set rating (kVA)500

Monthly tab columns

Row 1 holds headers; rows 2 to 13 hold April to March.

ColumnHeaderEntry or formula (row 2)
AMonth2026-04
BHour meter, openingReading on first day
CHour meter, closingReading on last day
DRun hours=C2-B2
EAMC / visits (₹)From invoices
FCleaning (₹)From invoices
GConsumables (₹)From invoices
HSensors and parts (₹)From invoices
ITesting (₹)From invoices
JInternal labour hoursFrom timesheets
KInternal labour (₹)=J2*Assumptions!$B$5
LWaste disposal (₹)From invoices
MCore amortisation (₹)=D2*Assumptions!$B$4
NTotal RECD cost (₹)=SUM(E2:I2)+SUM(K2:M2)
OCost per run hour (₹)=IFERROR(N2/D2,"no run")
PkWh generatedFrom energy meter
QCost per kWh (₹)=IFERROR(N2/P2,"")
RDowntime estimate (₹)Optional, kept out of N

Column J holds hours, not rupees, so the total in N skips it.

Year-to-date and rolling formulas

Place these below the monthly rows, in row 15 onwards:

  • Annual cost per run hour: =SUM(N2:N13)/SUM(D2:D13)
  • Same, using meter readings directly: =SUM(N2:N13)/(C13-B2)
  • Cost per 100 run hours: =SUM(N2:N13)/SUM(D2:D13)*100
  • Rolling three-month cost per hour (at month 12): =SUM(N11:N13)/SUM(D11:D13)
  • Cost per set-year: =SUM(N2:N13) on each tab, then averaged on the summary tab

Never average column O: a month with 2 run hours and ₹18,000 of cleaning shows ₹9,000 per hour and wrecks the average. Always divide total cost by total hours.

A simple exception flag

Add a column that marks unusual months:

=IF(AND(D2>0,O2>1.25*SUM(N$2:N$13)/SUM(D$2:D$13)),"Review","")

This flags months more than 25% above the annual figure; tighten the threshold once you have history.

The optional fuel penalty column

If back-pressure is measured at each visit, you can estimate extra fuel as run hours × fuel use per hour × penalty percentage × diesel price. Take the penalty percentage from your engine supplier or your own measured data, and keep the column outside the main total. Our explainer on how exhaust back-pressure affects a DG set gives the background.

Step 4: An illustrative three-site worked example

The figures below are illustrative only. They are chosen to show how the arithmetic behaves, not to indicate what any RECD, supplier or service contract will cost. Your own costs depend on the device design, duty, fuel, location and contract terms; our guide to RECD cost in India covers what drives those figures.

The three sites

  • Site A: 500 kVA set at a manufacturing unit, runs during grid outages and peak shaving. 420 run hours in the year.
  • Site B: 500 kVA standby set at an office building. 95 run hours in the year, mostly short weekly exercise runs and brief outages. Site A and Site B use the same RECD model, so the 500 kVA DG set RECD cost structure is directly comparable between them.
  • Site C: 1010 kVA set at a plant with weak grid supply. 1,150 run hours in the year.

Core amortisation assumptions: ₹2,40,000 over 6,000 hours for the 500 kVA units (₹40 per hour) and ₹3,60,000 over 6,000 hours for the 1010 kVA unit (₹60 per hour).

Annual costs by bucket

BucketSite A (₹)Site B (₹)Site C (₹)
AMC / visits48,00048,00072,000
Cleaning36,000 (2 × 18,000)18,000 (1 × 18,000)1,04,000 (4 × 26,000)
Consumables6,5003,00014,000
Sensors and parts14,000021,000
Testing22,00022,00028,000
Internal labour9,0006,00018,000
Waste disposal3,5001,5008,000
Core amortisation16,800 (420 × 40)3,800 (95 × 40)69,000 (1,150 × 60)
Total1,55,8001,02,3003,34,000

Checking the totals:

  • Site A: 48,000 + 36,000 + 6,500 + 14,000 + 22,000 + 9,000 + 3,500 + 16,800 = ₹1,55,800
  • Site B: 48,000 + 18,000 + 3,000 + 0 + 22,000 + 6,000 + 1,500 + 3,800 = ₹1,02,300
  • Site C: 72,000 + 1,04,000 + 14,000 + 21,000 + 28,000 + 18,000 + 8,000 + 69,000 = ₹3,34,000

Cost per operating hour and per 100 hours

MetricSite ASite BSite C
Total cost (₹)1,55,8001,02,3003,34,000
Run hours420951,150
Cost per run hour (₹)1,55,800 ÷ 420 = 370.951,02,300 ÷ 95 = 1,076.843,34,000 ÷ 1,150 = 290.43
Cost per 100 run hours (₹)37,0951,07,68429,043

At first glance, Site B looks like a disaster: nearly three times Site A’s cost per hour on an identical set. Site C, the largest set with the biggest bill, looks the most efficient.

Both conclusions are partly an illusion created by fixed costs.

Splitting fixed and variable costs

Treat the AMC and annual emission test as fixed (calendar-driven) and everything else as variable (driven by run hours and conditions). State your own split in the Assumptions tab.

MetricSite ASite BSite C
Fixed cost (₹)48,000 + 22,000 = 70,00048,000 + 22,000 = 70,00072,000 + 28,000 = 1,00,000
Variable cost (₹)1,55,800 − 70,000 = 85,8001,02,300 − 70,000 = 32,3003,34,000 − 1,00,000 = 2,34,000
Variable cost per run hour (₹)85,800 ÷ 420 = 204.2932,300 ÷ 95 = 340.002,34,000 ÷ 1,150 = 203.48
Fixed cost per set-year (₹)70,00070,0001,00,000
  • Fixed costs explain most of Site B’s headline figure. Its ₹70,000 of fixed cost spread over 95 hours is ₹736.84 per hour on its own (70,000 ÷ 95). Better maintenance cannot remove that; judge standby sets on fixed cost per set-year.
  • Site B’s variable cost per hour is still high. At ₹340.00 against roughly ₹204 at Sites A and C, the set needed a full cleaning after only 95 hours.
  • Sites A and C are almost identical on variable cost per hour, even though Site C is twice the size. That suggests Site C’s device is performing well relative to its rating.

Normalising by size

To compare sets of different ratings, divide variable cost per hour by the set rating in hundreds of kVA:

  • Site A: 204.29 ÷ 5 = ₹40.86 per hour per 100 kVA
  • Site B: 340.00 ÷ 5 = ₹68.00 per hour per 100 kVA
  • Site C: 203.48 ÷ 10.1 = ₹20.15 per hour per 100 kVA

This confirms the reading: Site C is the cheapest to run per unit of capacity, Site A is in the middle, and Site B stands out. Treat these as bands for discussion, not as precise rankings; a few thousand rupees of one-off parts can move them.

Cost per kWh as a cross-check

If the energy meters show 11,400 kWh at Site B (95 h at an average 120 kW), its cost is 1,02,300 ÷ 11,400 = ₹8.97 per kWh. An average 120 kW on a 500 kVA set (about 400 kW at 0.8 power factor) is roughly 30% of rating, a strong clue to its variable-cost problem.

Step 5: Reading the numbers, and what a rising cost per hour tells you

A single year’s figure is a baseline. The value comes from watching the trend and asking why it moved. When the variable RECD service cost per operating hour rises over two or three quarters, the cause usually falls into one of a handful of groups.

Low-load or short-run operation

Engines that run lightly loaded or for short exercise runs produce more soot and run cooler exhaust, so filters load faster and cleaning comes sooner. Site B in the example is a textbook case. Compare cleaning interval in run hours against your load data; our guide to understanding generator load factor and duty cycle explains how to calculate both, and our post on choosing an RECD for a generator running at low load covers the design side.

Rising engine oil consumption

When an engine burns more lubricating oil, ash from oil additives accumulates in filters and cannot be burnt off. DieselNet’s overview of ash in diesel particulate filters describes how this ash builds up over time. In the sheet, it shows as shorter cleaning intervals and an earlier-than-expected core replacement. Cross-check against engine oil top-up records; our post on engine oil consumption and RECD suitability explains what levels should prompt an engine inspection.

If the engine itself needs work, the cost of the repair belongs to the engine budget, not the RECD sheet. Coordinating the two teams so that the RECD is protected during and after the repair is covered in our sibling guide on coordinating DG engine repairs with the RECD service team.

Poor or incomplete cleaning

If cost per hour climbs while cleaning visits get more frequent, check what each visit actually achieved. Service reports that omit back-pressure before and after cleaning, or do not record what was done, make it impossible to tell a thorough clean from a quick wipe. Our sibling post on auditing an RECD service report for missing information gives a field-by-field checklist you can apply before approving each invoice.

Parts becoming scarce

A sudden jump in the sensors and parts column, or long lead times inflating downtime estimates, can be an early warning that a model is reaching end of support. Our sibling guide on planning for parts when an RECD model is discontinued covers what to stock and when to start planning a replacement.

A note on what the figure cannot tell you

Cost per hour is a management metric, not a compliance test. Compliance is established through type approval and testing under the frameworks published by the Central Pollution Control Board, your SPCB consent conditions and, in Delhi NCR, directions from the Commission for Air Quality Management. Never let a cost target push anyone towards bypassing or removing the RECD to keep a set running cheaply; that is a serious compliance risk, and sealing and penalties can follow.

Benchmarking RECD costs across sites without fooling yourself

League tables are useful only when they compare like with like.

Pitfall 1: Mixing standby and prime sets

Standby sets will always lose on cost per hour because of fixed costs. Put them in a separate group and rank them by fixed cost per set-year and variable cost per hour. Rank prime and near-prime sets by cost per hour or per kWh.

Pitfall 2: Comparing different sizes directly

A 1010 kVA device will cost more per hour than a 500 kVA device in absolute terms. Normalise by rating, as in the worked example, and compare within rating bands rather than across the whole fleet. A simple banding that works for most portfolios:

BandRatings
SmallUp to 125 kVA
Medium160 to 500 kVA
Large625 to 1250 kVA
Very largeAbove 1250 kVA

Pitfall 3: Letting one-off events distort a year

A core replacement booked as an expense, a damaged sensor after a flood, or a re-installation after exhaust modification can make one site look terrible for a year. Amortise the core, and tag other one-off costs in a “one-off” column so you can show the figure with and without them.

Pitfall 4: Different contracts and too little data

One site’s AMC may include cleaning while another bills it separately, so totals compare fairly but bucket breakdowns do not. Note the contract type on the summary tab; when renegotiating, our guide on how to negotiate an AMC contract for your generator pairs well with your cost history. Treat any figure under 100 run hours as indicative only.

A monthly dashboard layout for RECD maintenance cost tracking

Managers need one summary page that answers “is anything going wrong?” in under a minute.

KPIs to show for each set

  • Run hours and total RECD cost, month and year to date
  • Rolling twelve-month cost per run hour, total and variable
  • Variable cost per hour per 100 kVA
  • Fixed cost per set-year (standby sets)
  • Run hours since last cleaning and since last core replacement
  • Months flagged “Review” this year
  • Estimated downtime cost (separate line)

Group sets by duty, then rating band, and use one highlight colour for “Review” flags and another for cores past 80% of expected hours.

Review rhythm

FrequencyWhoWhat to do
MonthlySite engineerEnter readings and invoices, explain flagged months
QuarterlyMaintenance headReview trends and cleaning intervals with the provider
AnnuallyFinance and facility headReset assumptions, set budgets per set

The annual review is also the right time to set budgets from the data rather than from last year’s AMC. Budget per set as: fixed cost per set-year, plus expected run hours × variable cost per hour, plus a contingency for one-off events. For the illustrative Site A, if next year’s expected run hours are 450, the budget would be ₹70,000 + (450 × ₹204.29) = ₹70,000 + ₹91,930.50 = ₹1,61,930.50, before contingency.

Do not forget that the sheet only works if the service reports and invoices arrive on time. The broader RECD maintenance guide sets out the routine checks that generate those records in the first place.

FAQ

What is a good RECD service cost per operating hour?

There is no universal benchmark, because device design, set size, duty, fuel quality and contract terms all vary. Build a year of your own data first, then compare sets within the same duty group and rating band. Your own trend is more useful than any published figure.

Should the replacement core be included in RECD running cost?

Yes, but amortise it over expected run hours rather than booking it in the month it is bought. Otherwise one month spikes and the trend becomes unreadable. On very low-hour standby sets, amortise over years if that gives a higher figure.

Why does my standby DG set show such a high cost per run hour?

Because fixed costs such as the AMC and annual testing are spread over very few hours. Split costs into fixed and variable, judge standby sets on fixed cost per set-year, and watch variable cost per hour for genuine problems.

Is cost per kWh better than cost per hour?

For prime-power sets with varying load, cost per kWh is often fairer because it reflects useful output. For standby sets with long periods of light load, cost per hour or per set-year is easier to interpret. Keep both in the sheet.

Should downtime and extra fuel be included in RECD lifecycle cost?

Track them, but in separate columns. They are estimates, not invoices, and mixing them into the main figure weakens its credibility. Report them as a second line alongside the invoice-backed total.

Planning RECD Costs with Aceget

Aceget supplies CPCB type-approved retrofit emission control devices for DG sets from 25 kVA to 2500 kVA. If you are building a cost history across several sites and want help understanding what service intervals and parts to expect for your sets’ duty, our team can talk through it with you. You can review our type approval and certification details, or contact the Aceget team with your sets’ ratings, run hours and current service pattern.

Related Reading



Leave a Reply