A battery project and a solar project of the same size do not get the same loan. Developers who have financed solar in India are often surprised by how much harder the storage conversation is — not because lenders dislike storage, but because they have far less history to price it with.
This article explains what the lender is actually doing, in plain language, and what you can do about it.
Why storage borrows at a premium
A lender funding a solar plant can model two decades of output with real confidence. Irradiation data goes back a long way, panel degradation is well characterised, and the failure modes are understood.
A battery is a younger asset class. Analysis of India’s storage build-out has pointed to storage projects carrying a higher cost of capital than other renewable ventures — with commissioning delays and contract complexity together adding up to around 400 basis points, or four percentage points, to the cost of capital. That is not a rounding error. On a fifteen-year loan it is the difference between a project that clears its return threshold and one that does not.
Three things drive most of that premium:
- Thin performance history. There is limited long-run field data on how a given cell behaves at Indian temperatures over ten years of daily cycling.
- Replacement cost uncertainty. Most projects will need augmentation — topping up capacity as cells fade. Nobody knows what cells will cost when that day arrives.
- Revenue uncertainty, on merchant projects especially, where earnings depend on market spreads rather than a signed contract.
Contracted versus merchant: two different loans
The single biggest fork in the road is whether you have a firm offtaker.
Read that chart as an equity story, not a debt story. If a merchant project raises half its cost as debt where a contracted one raises nearly two-thirds, the sponsor has to write a much bigger cheque for the same megawatt-hours — and equity is the expensive money.
What DSCR actually means
DSCR stands for debt service coverage ratio, and it is simpler than it sounds. Take the cash the project generates in a year, and divide it by the loan repayment due that year. A DSCR of 1.20x means the project earns twenty percent more than it owes.
Lenders do not use DSCR to grade you. They use it to size the loan. The bank picks the minimum coverage it will accept, looks at your projected cash flows, and lends whatever amount keeps you above that line in the worst modelled year. A higher required DSCR mechanically means a smaller loan.
You will also see P50 and P90 attached to these numbers. P50 is the central case — the outcome you would beat half the time. P90 is the conservative case you would beat nine years in ten. Testing coverage against P90 is a much stricter test than testing against P50, which is why the contracted structure above carries both a lower ratio and the tougher probability case: the underlying revenue is simply more predictable.
Where the revenue comes from matters too. Lenders generally model capacity payments, ancillary services and energy arbitrage as three separate streams with three different confidence levels rather than one blended number — the logic we set out in our guide to revenue stacking.
What is improving on the policy side
Two developments in the Union Budget for 2026 are worth knowing about:
- Basic customs duty was exempted on capital goods used in manufacturing battery energy storage systems, which pulls down the equipment cost base that everything else is calculated from.
- An Infrastructure Risk Guarantee Fund (IRGF) was proposed, offering partial credit guarantees to lenders. The intent is straightforward — if a guarantee absorbs part of the downside, lenders can extend longer tenors and lower pricing to projects they would otherwise treat cautiously.
The market these measures are aimed at is no longer small. As of March 2026, operational storage capacity above 1 MWh stood at roughly 798 MWh, with about 26,729 MWh under construction and a further 61,013 MWh at the tendering stage. Financing that pipeline is the binding constraint, not finding projects.
The documents that actually move a lender
Most of what improves your terms is evidence you should be assembling anyway:
| What the lender is unsure about | What settles it |
|---|---|
| Will the battery still perform in year ten? | Cell test data and a clear degradation guarantee from a supplier with a balance sheet behind it |
| What will augmentation cost? | A costed augmentation plan in the model, not a footnote |
| Will revenue actually arrive? | Signed offtake, or a defensible market study if merchant |
| What happens if there is a fire? | A placed insurance programme — see insuring a storage project |
| Can the sponsor build it on time? | Track record, and a contractor who has commissioned storage before |
Commissioning delay is worth singling out. It appears in the cost-of-capital gap for a reason: interest accrues during construction whether or not the asset is earning, so a project that slips two quarters has already spent part of its return.
What this means for you
- If you are a developer: get the offtake conversation and the lender conversation running in parallel, not in sequence. The structure you sign determines the loan you can raise, and discovering that after you have committed to a merchant model is expensive.
- If you are an IPP weighing merchant exposure: model the equity cheque, not just the internal rate of return. Merchant upside is real, but it is funded with the costliest capital in the structure.
- If you are a C&I buyer: most of this applies in miniature. Your bank will still ask who stands behind the warranty and what happens at end of life, even on a single ADESS 1000 unit behind your meter.
- If you are early and just sizing the opportunity: start from the cash flows before you start from the loan. Our savings calculator gets you a first-pass number, and our team can pressure-test a specific site with you — get in touch.
Financing terms, guarantee schemes and duty exemptions change by notification and by lender credit policy, and the ranges above are market observations rather than offers. Treat this as an August 2026 snapshot and confirm current terms with your own lenders and advisers.