If you are reading this, the decision question is probably behind you. Our BESS decision framework gives C&I buyers a five factor screen for whether battery storage belongs in their plant, and its diesel displacement test is simple: if your DG sets log more than roughly 200 hours a year, storage deserves a serious look. That 200 hour figure is an entry gate, not a conclusion. Passing it tells you the project is worth engineering. It does not tell you how to engineer it.
This playbook is the how. Our storage market outlook asked readers to establish three numbers this quarter: your peak demand pattern, your monthly diesel spend, and the evening and night share of your consumption. Treat those three numbers as your starting inputs. Everything below assumes you have them, or shows you how to get them properly, and then goes several levels deeper: into your DG logbook, your real fuel cost per unit, sizing against an actual load profile, changeover engineering, the retrofit sequence, and the commissioning tests you should refuse to sign off without.
Step 1: Read your own DG logbook before anyone sizes anything
Every diesel genset room in India has a logbook, and most of them are write only. Entries go in at shift change and nobody reads them again. That logbook, plus your fuel purchase records, is the single most valuable engineering document in this entire project, because it is the only honest record of what your plant actually does when the grid fails.
Here is what to extract, and over what period:
- Run hours per month, for at least six months and ideally twelve. One month is not enough. DG usage in most of North India is strongly seasonal: summer grid stress, monsoon trips, winter fog outages and festival season load all behave differently. A logbook read in a mild month will undersize everything.
- Fuel drawn per month, reconciled against purchase invoices, not against memory.
- The load the DG was carrying when it ran. Many logbooks record only start and stop. If yours does, start recording the kW or ampere reading at the top of each running hour from today. Six weeks of that data will transform the quality of the sizing work later.
- Why each run happened. Grid failure, voltage sag, planned maintenance, or a peak clipping habit. Each reason points to a different technical fix.
Now convert fuel into a cost per unit of electricity, because "per litre" hides the truth and "per kWh" reveals it. The conversion runs through specific fuel consumption, which is the litres of diesel a genset burns to generate one kWh:
| Operating condition | Typical specific fuel consumption |
|---|---|
| General real world range | 0.25 to 0.30 litres per kWh |
| Larger modern gensets at roughly 70 to 80 percent load | 0.21 to 0.25 litres per kWh |
| Smaller or lightly loaded gensets | 0.30 to 0.45 litres per kWh |
Indicative field ranges from our project experience. Your genset's actual figure should be established from your own fuel and generation records.
Multiply your specific fuel consumption by your diesel price per litre and you have your fuel cost per unit. Do this with your own numbers, not with anyone's published average, including ours.
Two worked examples show why the procurement route matters so much right now. At a retail pump price near Rs 95 per litre and a healthy 0.25 to 0.26 litres per kWh, fuel alone works out to roughly Rs 24 to 25 per unit. At a formal industrial bulk supply price near Rs 135 per litre, the same genset produces power at roughly Rs 34 to 35 per unit, fuel alone, before a rupee of maintenance, lube oil or depreciation is counted.
Diesel prices as of mid 2026 and subject to change; verify current rates with your supplier before making investment decisions.
That gap between pump and bulk pricing is not a rounding error, and in June 2026 it acquired regulatory teeth: the Ministry of Petroleum and Natural Gas issued a temporary order restricting industrial, commercial and institutional users from buying diesel at retail pumps, directing them to bulk supply channels and capping retail sales at 200 litres per customer per day. The order is temporary and its status should be checked at the time you read this, but the direction is clear. A C&I plant should build its business case on the diesel price it can actually and legally procure at, which for most industrial users today means the bulk rate.
Step 2: Run hours and load factor, where the real money hides
The specific fuel consumption table above contains the most underappreciated fact in DG economics: the same genset produces power at very different costs depending on how hard it is working.
A genset sized generously for the worst case, which is how most were bought, spends most of its life lightly loaded. At 40 percent load, specific fuel consumption climbs toward the top of the range and beyond, and the per unit cost can be half again higher than the healthy load figure. Lightly loaded diesels also wet stack, glaze their cylinder liners and demand more maintenance per running hour. So a plant whose logbook shows long hours at partial load is paying the worst possible price per unit at exactly the moments it thinks backup is cheap.
This matters for the replacement case in a way that a simple annual diesel bill never shows. When a battery takes over the routine, low load, short duration events, it removes the most expensive units the DG produces, not the average ones. The remaining DG hours, if any, are the long, heavily loaded emergency runs where the genset is at its most efficient. In other words, a properly sized solar and BESS retrofit does not just reduce DG hours proportionally. It preferentially deletes the ugliest hours first. Build your avoided cost estimate on that shape, not on a flat average, and the case will be both stronger and more honest. Our CFO avoided cost model shows how these avoided units then flow through a financial model alongside tariff and demand charge effects.
Step 3: Sizing against a real load profile, power versus energy
Here is the question we are asked most often at this stage: "What size battery replaces a 500 kVA DG?" And here is the honest answer: the question cannot be answered as asked, because a genset rating and a battery rating measure different things.
A DG's kVA rating describes the maximum apparent power it can deliver at an instant. A battery system has two ratings that matter: the power its inverters can deliver in kW, and the energy its cells can store in kWh. Replacing a DG means matching the power rating to your real critical load, and choosing the energy rating from how long you need that load carried. There is no reliable fixed ratio between battery kWh and genset kVA, and any vendor quoting one before seeing your load data is guessing.
The correct sizing inputs are:
- The actual critical load in kW, measured, not the DG nameplate. Most 500 kVA sets in our experience carry far less than their rating.
- The required autonomy in hours: how long must the battery carry that load before either the grid returns, solar recovers, or the DG is called.
- The permissible depth of discharge of the chosen battery chemistry.
- Inverter and conversion efficiency.
- An aging reserve, so the system still meets its duty in year eight, not just in month one.
A worked illustration, using round numbers our engineering team would recognise from the field: a plant with a 500 kVA DG whose measured critical load is about 200 kW. Working through usable depth of discharge, conversion efficiency and an aging margin, the preliminary battery sizes come out near:
| Autonomy required | Preliminary battery size |
|---|---|
| 30 minutes | roughly 140 kWh |
| 1 hour | roughly 280 kWh |
| 2 hours | roughly 560 kWh |
| 4 hours | roughly 1.1 MWh |
Preliminary illustration for a 200 kW critical load; actual sizing must be engineered from your measured load profile and chosen battery specification.
Notice what drives the answer: the measured 200 kW, not the 500 kVA on the nameplate. This is why the logbook and load recording work in Step 1 is not bureaucracy. It routinely cuts the required battery investment dramatically compared to sizing from the genset rating, and it is the difference between a system engineered for your plant and one guessed for a brochure. Real deployments follow exactly this logic: the 600 kWh system at an agro processing plant in Mainpuri, Uttar Pradesh described in our storage market outlook was sized from the plant's outage pattern and evening load, not from its genset rating.
One more sizing distinction that changes budgets: covering outages is not the same as covering an evening shift. A battery that rides through grid failures needs energy for the longest credible outage. A battery that also shifts solar energy into evening operation needs energy for the whole evening consumption block, every day, which is a much larger and differently cycled duty. Decide which duty you are buying before anyone quotes you.
Step 4: The changeover, engineered rather than promised
Our decision framework notes the headline difference: a battery inverter picks up load in milliseconds, while a DG needs 30 seconds to 3 minutes to start, stabilise and accept load. On the plant floor, engineering that headline into reality involves choices that deserve scrutiny.
The central question is where the battery inverter sits relative to your changeover switchgear. In a true uninterrupted arrangement, the critical load bus is fed through the battery system continuously or near continuously, so a grid failure is invisible to the load: no contactor race, no restart of drives, no PLC reboot. In a switched arrangement, the battery waits behind a transfer switch much as the DG does, and a break of some milliseconds to a few seconds occurs depending on the switchgear class. Both are legitimate designs. They cost different amounts, and they protect against different things.
Match the design to what an interruption actually costs you. A plant whose losses come from spoiled batches, tripped process lines or crashed controllers needs the uninterrupted class and should say so in the specification. A plant that merely needs lights and essential services back quickly can take the switched class and save money. What you should not accept is an uninterrupted price for a switched design, and the only way to know the difference is to ask, in writing, how many milliseconds of interruption the critical bus will see on grid failure, and to have that number demonstrated at commissioning. We return to that test in Step 7.
Anti islanding protection, DG and inverter synchronisation if both can run together, earthing scheme and protection grading also belong in this conversation. They are standard engineering, but they are the part of the project where an experienced integrator earns their fee, and where a cheap quotation usually reveals itself.
Step 5: Does the DG stay?
In most retrofit projects, we recommend keeping the existing DG as a backup of backup. This surprises buyers who expected "DG replacement" to mean a crane removing the genset, so it deserves a clear explanation.
Solar and battery should eliminate routine and low load DG running, which as Step 2 showed is where the worst economics live. The DG remains valuable for the events a rationally sized battery should not be asked to cover: extended multi day outages, prolonged low solar periods, exceptional load events, and maintenance windows on the battery system itself. Sizing a battery for the worst credible week rather than the worst credible evening multiplies the investment for events that may occur once in several years. The genset you already own covers those events at zero additional capital cost.
We would advise removing the DG entirely only where the site can genuinely tolerate interruption, or where the battery system has been deliberately designed with redundancy to cover the worst credible outage on its own. For everyone else, the practical goal is not "remove the DG". It is reduce DG operation from regular use to rare emergency use: from hundreds of hours a year to a monthly test run and the occasional genuine emergency. The fuel bill collapses, the maintenance contract shrinks, the compliance exposure on emissions and noise falls, and the genset itself lasts longer because it no longer accumulates lightly loaded hours.
Step 6: The retrofit sequence, what happens in what order
A DG displacement retrofit on a running plant is a sequencing exercise. The work falls into phases, and knowing which items sit on the critical path helps you hold a realistic programme:
- Measurement and study. Load logging, logbook analysis, roof and land survey, grid and DISCOM context. Runs while the plant operates normally. This phase feeds everything and rushing it is the classic false economy.
- Design and approvals. System engineering, protection coordination study, and the regulatory workflow with the DISCOM for the solar interconnection where applicable. Approval timelines vary by state and are frequently the longest single item in the programme; start them early and in parallel with procurement.
- Procurement and civil works. Module mounting structures, cable routes, the battery and inverter room or enclosure with its ventilation and fire detection. Almost all of this proceeds without touching the live electrical system.
- Electrical installation. Solar strings, inverters, battery installation, and the new switchgear. The only unavoidable shutdowns sit here, at the tie in points. A competent contractor will consolidate tie ins into one or two planned shutdown windows, agreed with production weeks in advance, often on a weekly off day.
- Testing, commissioning and handover. Covered in the next step, and not to be compressed no matter how close the festival deadline is.
Where do delays actually live? In our experience: DISCOM approvals, battery delivery schedules, and shutdown windows that production keeps postponing. None of these is solved by pressure late in the project. All three are solved by starting them early.
Step 7: Commissioning and acceptance, the tests to insist on
Commissioning is where promises become measurements. Before signing acceptance, a buyer should insist on witnessing, at minimum:
- A real changeover test under load. Open the grid supply deliberately, with the plant running its normal critical load, and measure the interruption seen at the critical bus. The measured figure must meet the number written in the specification. A demonstration on a lightly loaded test circuit does not count.
- A battery duty test. Discharge the battery into the real load for the specified autonomy duration and confirm the delivered energy and the voltage behaviour at the end of discharge. This verifies the sizing arithmetic from Step 3 in metal and chemistry rather than on paper.
- DG interaction tests. If the DG remains, prove the full sequence: battery carries the load, DG receives its start signal at the agreed state of charge, synchronises or transfers correctly, and hands back cleanly when conditions recover.
- Protection and anti islanding tests as per the approved protection study, witnessed and documented.
- Monitoring verification. Every meter and sensor reporting into the monitoring platform should be cross checked against a field measurement. Your future avoided cost reporting is only as good as this instrumentation.
- Documentation handover. As built drawings, protection settings, battery warranty terms with their cycling and temperature conditions, and the operations and maintenance schedule.
Insist that the acceptance criteria and the test procedure are written into the contract at signing, not negotiated at handover. Any capable EPC partner will agree to this readily. Reluctance to commit to witnessed acceptance tests is itself useful information about a vendor.
Where MGetEnergy fits
We have spent 13 plus years building solar and hybrid systems for Indian industry: 45 plus MWp delivered across 400 plus installations in 8 plus states, from our engineering base in Greater Noida serving Delhi NCR and our Mumbai office serving the west. DG displacement retrofits sit squarely inside our hybrid and BESS practice, from the first load logging exercise through design, installation and the witnessed acceptance tests described above.
If your DG logbook is telling you the story this playbook describes, we are happy to read it with you. Bring the three numbers, or let us help you establish them, and we will tell you honestly whether the engineering case is there. Talk to our engineering team.
Frequently asked questions
How long should I log DG and load data before sizing a battery system?
At least six months of run hour and fuel records, and ideally twelve, because DG usage in India is strongly seasonal. If your logbook has no load readings, start recording kW at the top of each running hour immediately; even six weeks of load data dramatically improves sizing quality.
How do I calculate my actual diesel generation cost per unit?
Multiply your genset's specific fuel consumption in litres per kWh by your diesel price per litre. Real world specific fuel consumption typically runs 0.25 to 0.30 litres per kWh, better for large well loaded sets and considerably worse for lightly loaded ones. At mid 2026 bulk diesel prices, fuel alone typically exceeds Rs 30 per unit before maintenance and depreciation.
Why is bulk diesel currently more expensive than pump diesel?
Through 2026, state owned oil companies held retail pump prices down to protect ordinary consumers while bulk industrial prices continued to track the international market, creating an unusual gap of roughly Rs 40 per litre. A June 2026 government order additionally restricted industrial users from buying at retail pumps. The situation is temporary in design and should be verified at the time of reading, but industrial buyers should build business cases on the bulk rate they can actually procure at.
Can a battery really take over from a DG instantly?
Yes, if the system is engineered for it. A battery inverter can pick up load in milliseconds, against 30 seconds to 3 minutes for a DG start. But the achieved changeover time depends on where the inverter sits relative to your switchgear, so specify the maximum interruption in milliseconds you will accept at the critical bus and have it demonstrated under real load at commissioning.
What size battery replaces a 500 kVA diesel genset?
There is no fixed conversion, because kVA measures instantaneous power capability while battery size depends on your measured critical load and required backup duration. As an illustration, a 500 kVA DG carrying a measured 200 kW critical load needs roughly 280 kWh of battery for one hour of autonomy, or roughly 1.1 MWh for four hours, after accounting for depth of discharge, efficiency and aging. Measure first, size second.
Should I remove my diesel generator after installing solar and battery storage?
Usually not. We generally recommend retaining the DG as a rarely used emergency reserve for extended outages, prolonged low solar periods and battery maintenance windows. The goal of a well engineered retrofit is to reduce DG operation from regular use to rare emergency use, which captures nearly all the fuel, maintenance and compliance savings without paying for battery capacity you would use once in several years.
How long does a DG replacement retrofit take on a running plant?
The measurement, design and approval phases dominate the calendar, with DISCOM approval timelines varying significantly by state. Physical installation mostly proceeds without disturbing production, with electrical tie ins consolidated into one or two planned shutdown windows. Start load logging and the approval workflow early; they, along with battery delivery schedules, are where programmes actually slip.
Disclaimer: All figures in this article, including diesel prices, specific fuel consumption ranges and battery sizing illustrations, are indicative, drawn from field experience and public information as of mid 2026, and are subject to change. They do not constitute a quotation or engineering advice for any specific site. System sizing, changeover design and commercial terms depend on your measured load data and site conditions and are confirmed only through a written proposal and separate written agreement.