BESS Contract Numbers May Hide Long-Term Delivery Gaps

India is building battery storage at scale. But if project sizing does not fully account for how batteries age, some of these assets could quietly under-deliver in the very years the grid may need them most. India’s storage story is, for now, mostly a story of volume. The CEA’s National Electricity Plan projects India’s energy-storage requirement to increase from 82.37 GWh in 2026-27 to 411.4 GWh by 2031-32, representing an approximately fivefold increase. Battery energy storage systems (BESS) are expected to form the bulk of this growth. Tenders are being structured. Developers are moving quickly to lock in capacity. Most of this activity gets described in one simple phrase: X MW/Y MWh. It may not, however, tell us much about what the grid will actually receive, year after year.

A battery system does not send out its full nameplate DC capacity to the grid. What actually reaches the interconnection point is AC energy, after a few unavoidable losses. There is the usable depth of-discharge window, kept narrower than full range to protect battery health. There are conversion losses through the inverter, transformer, and cabling. Some industry white papers on LFP sizing suggest this gap can be significant – a system rated at 100 MWh DC might deliver only around 85 90 MWh AC at the measurement point, even on day one, before any ageing has taken place.

Day one is, in some ways, the easier part. Lithium iron phosphate (LFP) cells, now common in utility scale BESS in India and globally, tend to lose capacity over time. This happens through a few well studied pathways – growth of a protective layer on the anode, gradual loss of active electrode material, and, under poor charging conditions, lithium plating. Heat tends to speed all of this up. So does holding a battery at a high state of charge for long periods and cycling it deeply and often. This may matter more in India than in many other markets. Ambient temperatures here often run higher than the mild conditions many international degradation curves are based on. Storage assets are also often expected to sit near full for long stretches, simply to stay available. Calendar ageing, and not just cycle wear, may deserve a closer look at the design stage.

It may also be worth noting that ageing does not only reduce how much energy a battery can hold. It can reduce how much power it can deliver as well. As internal resistance rises with age, voltage drop under load tends to increase. This can push the system toward earlier cut-offs and thermal derating at a given discharge rate. In other words, a battery could show adequate remaining energy on paper and still struggle to offer its contracted MW output for the full contracted duration. For storage meant to support firm capacity, ancillary reserves, or peak-shaving over 15-20 years, this could be a meaningful factor in whether a project earns its expected revenue in year 12, not just in year 1.

A simple example may help illustrate this. Suppose a procurer asks for 100 MWh of energy (50 MW for 2 hours) at the grid interconnection point, every year, for 15 years. There are broadly three ways a developer could approach sizing for this:

Table: Indicative dispatchable AC capacity, by year (MWh)

Year	No oversizing	Upfront oversizing	Oversizing + augmentation
0 (COD)	100	135	110
2	95	128	104
4	91	123	110 (augmentation #1)
7	87	117	105
10	83	112	110 (augmentation #2)
15	77	104	102

Note: Augmentation events at Year 4 and Year 10 briefly lift delivered energy above the 100 MWh requirement, before the next degradation cycle brings it closer to the requirement again by Year 9 and Year 15.

Indicative numbers, based on a broad market assessment. The actual profile would depend on the specific OEM’s DC block design.

The no-oversizing column shows roughly what could happen when a project is sized only to pass commissioning tests: AC output may gradually fall to around 77 MWh by year 15, a fairly wide gap from the original 100 MWh commitment. Meeting the full 15-year requirement instead would mean either installing around 135 MWh upfront, to absorb all future degradation in one step, or starting smaller and adding capacity in stages, roughly twice over the asset’s life. The second path may mean lower upfront capital, and it could leave room to adopt better battery technology later, once it becomes commercially available. It does, however, bring its own complexity – multiple battery blocks of different ages operating together, some difference in internal resistance and ageing behaviour, and a need for more careful controls, protection design, and warranty planning. Space, conduit, and switchgear capacity for future additions would also need to be reserved at commissioning.

There is a useful real-world reference point for staged capacity addition. The Hornsdale Power Reserve in South Australia, one of the first large grid-scale lithium-ion batteries, started out in 2017 at 100 MW/129 MWh. In 2020, its owner added a further 50 MW/64.5 MWh, taking the system to 150 MW/193.5 MWh, without needing to rebuild the plant from scratch. The expansion was driven mainly by a wish to add new grid services, rather than by degradation alone. Even so, it shows that staged capacity addition at an operating BESS site is a workable, proven approach, provided the site and controls are planned for it from day one. That is a reassuring data point for developers weighing whether an augmentation strategy is practical.

Neither approach, oversizing or augmentation, appears clearly better in every case. The point for Indian developers, lenders, and EPC contractors may be this: it helps to make the sizing choice in an informed manner, with the help of degradation curve, efficiency assumptions. Two systems both described as “50 MW/100 MWh” could end up sized quite differently, depending on where that capacity is meant to be verified. In most Indian tenders, the measurement point sits at the point of interconnection, which would normally include downstream transformer and cabling losses that a test at the inverter output alone would miss.

This suggests that technical due diligence need to go hand in hand with procurement processes. As India gradually moves toward a market that will increasingly reward storage through ancillary services, capacity contracts, and time-of-day arbitrage, the value of a BESS project may depend less on how many MWh are commissioned, and more on whether it keeps delivering its committed output, at the committed power level, across the life of the contract. It may help for a lender or offtaker reviewing a project to ask a few simple questions –

whose degradation curve is being used, and is it specific to the chosen cell supplier and to Indian conditions;
is the usable depth-of-discharge window stated;
are efficiency curves provided across different load levels, etc.

None of this is an argument against India’s storage ambitions. Our grid needs the capacity and needs it fairly quickly. It may simply be an argument for treating Battery Energy Storage System sizing with something like the same care the solar industry eventually applied to irradiance and panel degradation modelling. A storage asset that shrinks to one-fourth of its contracted value by year 15 may not serve anyone particularly well – not the developer facing availability penalties, not the discom counting on it for peak support, and not the grid operator relying on it for reserves. A little more sizing discipline today could go a long way toward making today’s storage targets into tomorrow’s dependable grid assets.

Volume gives India a battery fleet. Thoughtful, degradation-aware sizing may be what decides whether that fleet is still doing its job well into year 15.