A solar plant has one job and one revenue line: it generates, and somebody pays for the generation. A battery is different. It can sell energy, it can sell the promise of energy, it can sell fast response to the grid operator, and it can quietly save its owner money without selling anything at all.
Getting more than one of those to happen on the same asset is what the industry calls revenue stacking — or value stacking. It is not a clever optimisation; for merchant storage in India it is closer to a necessity, because no single market is yet deep enough to pay for a battery on its own.
The layers
Energy arbitrage
The oldest idea in storage: charge when power is cheap, discharge when it is expensive. In India this has been the dominant merchant stream by a wide margin — around 80 to 85 per cent of total earnings in the observed merchant mix, with ancillary services making up the rest.
It works because Indian prices swing hard within a single day. In April 2026, surplus daytime solar pushed real-time market prices close to zero while evening day-ahead prices approached the ₹10/kWh ceiling. That gap is the product. We cover the mechanics in energy arbitrage on the power exchange.
Ancillary services
Ancillary services are what the grid operator buys to keep frequency stable — the ability to inject or absorb power within seconds. Batteries are unusually good at this because they respond faster than any thermal plant.
CERC’s Ancillary Services Regulations formally allow batteries to provide fast-response secondary reserve and tertiary reserve, with procurement, deployment and compensation mechanisms spelled out. Our article on ancillary services and frequency regulation goes through how the reserves differ.
The layer is smaller than arbitrage, but it has a useful property: you are often paid to stand ready rather than to deliver, so the cycles consumed can be modest relative to the revenue.
Capacity and tolling payments
Here somebody pays a fixed fee for the right to use your battery, and takes the market risk themselves. It is the least exciting layer and the one that makes projects financeable — which is why so many Indian tenders are structured this way. See tolling agreements and capacity contracts for how these are written.
Behind-the-meter savings
For a factory, hospital or commercial building, the stack is much simpler and often more dependable: avoid the peak demand charges on your bill, and shift consumption around your time-of-day tariff. No exchange registration, no scheduling, no market risk. The money shows up as a smaller bill.
Why you cannot just add the layers up
This is the part that trips up business cases.
A battery holds a fixed amount of energy and has a finite number of cycles it can do before its warranty is affected. Every megawatt-hour you discharge into the evening arbitrage peak is a megawatt-hour you did not have available for a reserve call. Capacity you commit to an ancillary contract is capacity you cannot bid into the day-ahead market.
Stacking is therefore a scheduling problem, not an addition problem. The realistic question is not “what could each stream pay?” but “given twenty-four hours, one to two cycles, and a cycle budget I must not breach, what is the best sequence?” A model that sums the theoretical maximum of each stream will overstate revenue substantially.
The cycle constraint is a real commercial limit, not a technical footnote — it is written into supplier degradation guarantees, and exceeding it can void the cover you are relying on.
Is the Indian stack deepening?
The direction of travel is favourable. Day-ahead market revenues available to storage have grown several-fold in recent years, and average one-hour price spreads widened to about ₹7.42/kWh in 2025 from ₹6.95/kWh in 2024 — modest in percentage terms, but spreads are the entire business model for an arbitrage asset. Analysis of a system installed in 2024 has suggested returns in the region of 17 per cent from arbitrage alone.
Demand is coming too. The Central Electricity Authority’s generation adequacy planning has put required battery storage at roughly 147 GWh by 2031-32 and 321 GWh by 2035-36. And India’s Electricity (Amendment) Rules, 2025 explicitly permit consumers to participate in these markets with their own storage assets — which over time widens the stack from a grid-scale story into a consumer one.
What this means for you
- If you are a developer: build the revenue model stream by stream with a shared cycle budget across all of them, and show the lender that budget. A stack that respects the cycle constraint is far more credible than one that maximises each line independently — and credibility is what determines how much debt you raise.
- If you are an IPP: treat contracted capacity as the floor and merchant streams as the upside, rather than betting the project on spreads holding.
- If you are a C&I buyer: ignore most of the above. Your stack is demand charges plus tariff shifting, and it is the most reliable stack in this article. A peak shaving system is sized against your own load curve, not against a market forecast.
- If you want a number for your own site: the savings calculator will give you a first pass on the behind-the-meter layers, and our team can model a specific load profile with you — get in touch.
Market rules, ancillary service products and exchange price behaviour change by regulation and notification, and the shares quoted here are observations of past market outcomes rather than forecasts. Treat this as an August 2026 snapshot and verify current CERC regulations and market design before committing to a revenue model.