India’s previous hydel push fell to political rent-extraction and speculation. The current one faces competition from emerging storage technologies.
The first instalment of this series ended with a question. By 2030, India wants to get 500 GW of energy from wind and solar. At that point, India will need as much as 85 GW of storage to balance the grid. Accordingly, the BJP-headed NDA government is making a huge push on not just batteries and green hydrogen, but also hydel.
It wants to hike India’s installed hydel capacity to 70,000 MW by 2030—a 50% jump from the current 45,700 MW. It’s also readying a large push on pumped storage. As energy minister RK Singh told the Economic Times, the NDA has identified 63 projects adding up to a generating capacity of 96,000 MW.
The catch: India is pushing hydel storage at a time when rival storage technologies like batteries and electrolysers are evolving rapidly. Not only are battery prices falling steeply across the world, the cost of green hydrogen is expected to plummet as well.
This raises the obvious question. These hydel projects are expected to go live by 2030. Their viability, however, hinges on an external variable—the price of battery and electrolyser storage by 2030.
Essentially, how big a role will hydel play in decarbonising India?
How the prices compare
Early in 2020, SECI floated an interesting tender. It wanted 1,200 MW of renewable power (solar or wind) with storage for six hours. Hyderabad-based Greenko won 900 MW with a peak power tariff rate of ₹6.12/kWh, and the rest went to Renew Power, which bid ₹6.85/kWh. Interestingly, while Renew used lithium-ion batteries for storage, Greenko bet on pumped storage. Can pumped storage retain this price advantage?
Hydel is a mature technology with little scope for dramatic cost-reduction. If anything, cement is getting costlier. So is resettlement. With banks, given the energy transition, wary of financing projects that will take 10 years to come up, financing costs will be high as well.
In contrast, prices of batteries and electrolysers are expected to fall. “As manufacturing scale builds up, the density of these technologies improves, and new methods to reduce rare earth dependence come up, prices will fall,” said Rahul Walawalkar, the president of India Energy Storage Alliance.
Take lithium-ion batteries, which account for 90% of the global grid battery storage market today. In April 2020, four energy researchers—from Prayas Energy Group, Maharashtra Electricity Regulatory Authority and the Lawrence Berkeley National Laboratory—published a report titled “Estimating the Cost of Grid-Scale Lithium-Ion Battery Storage in India”. Studying American battery storage projects to gauge how prices are moving, they pegged the cost of a Battery Energy Storage (BESS) at ₹1.44/kWh in 2020, ₹1.0/kWh in 2025, and ₹0.83/kWh in 2030.
What about electrolysers? Reliance, as we know, has promised to produce green hydrogen at the cost of $1/kilo. With one kilo of hydrogen—which accounts for 80% of the cost of power generation—producing 33.6 units of electricity, that works out to, at the current exchange rate, a little below ₹3 a unit. More conservative estimates peg the price of green hydrogen at $2 by 2030. Which works out to about ₹5.50/unit.
And then, there is pumped storage. A feasibility report for the 1,200 MW Kurukutti Pumped Storage Project in Andhra Pradesh, which is coming up in Andhra’s Vizianagaram district, seeks to meet the peaking demand for the south grid in general and Andhra Pradesh in particular. Its levelised tariff—tariff averaged over the life of the project—has “been found to be ₹7.85/kWh considering cost of pumping @ ₹3.00/kWh,” says the report, prepared by Tata Consulting Engineers in August 2021.
There is no reason why that tariff should not be the industry norm. At 1,200 MW, Kurukutti’s capital costs stand at ₹4,766 crore. The 1,200 MW pumped storage component of Greenko’s Pinnavaram project costs a similar ₹5,468 crore. Or take Greenko’s 2,520 MW Shahpur Pumped Storage project, which costs Rs11,579 crore.
Taken together, they suggest a range of ₹3.9 crore-₹4.59 crore per MW.
The cost of a Run of The River (RoR)—or a multipurpose—project can be even higher, said Jayant Kawale, the former head of Jindal Steel and Power’s hydel business. “The cost of power from these projects might be as high as ₹12/13. Even the best of projects will not be less than eight rupees a unit.”
This, however, doesn’t mean there is no room for hydel.
Looking beyond price
Mention low solar tariffs to hydel executives and they bristle. Part of the reason solar is cheaper than hydel is the subsidies it enjoys. “Too many ancillary costs of solar are not being factored in—like the O&M [Operation and Maintenance] costs of power plants kept idling that DISCOMs have to pay,” said Probodh Mallick, a senior executive at Marubeni.
According to them, the overweening focus on low tariffs is narrow-minded and hurts India. In the past, for instance, the focus on tariffs has resulted in India spurning offers by Kazakhstan, China and Russia to set up large solar plants through the MoU route. “Why is everything being decided only through competitive bidding? Why not through the MoU route?” a former bureaucrat at Solar Energy Corporation of India had told CarbonCopy last year. “What is happening as a result is that we are ending up with low quality infrastructure.”
A similar critique applies to storage as well. Price cannot be its sole yardstick. We need other parameters as well.
At this time, amongst all these technologies, hydel can best mirror fluctuations in national power demand. We saw this in April 2020 when India switched off its lights at 9 pm (for 9 minutes) after Modi said this would show solidarity against Covid. That night, power demand dropped by as much as 32 GW in a few minutes before amping up again.
It was an unprecedented drop. Even smaller mismatches between power supply and demand can result in power frequency spiking or dropping, putting users’ (and the country’s) electrical infrastructure at risk. India rode that evening out by damping thermal power production while increasing the share of hydel power by 8:30 PM, using it to mirror power demand over the next hour, and then ramping thermal back up again.
Can that change by 2030?
The rise of dispersed storage
In energy circles, opinion is split on that question. One school of thought says BESS and its ilk can take care of relatively smaller loads—but that heavier balancing will need to be done by hydel.
“Nowhere in the world do we have battery installations at a grid-level,” said R Srikanth, the dean of the school of natural sciences and engineering at Bangalore-based National Institute of Advanced Studies. Referring to the Central Electricity Authority’s projection that BESS storage in India will rise to 27 GW by 2030, he said its projections hinge on the assumption that battery prices will fall by two-thirds by 2030.
Others disagree. According to Walawalkar, storage will evolve granularly. “About 30-40% of India’s power demand is commercial load,” he said. “This segment of the market keeps backups. Even today, RE and storage is cheaper than diesel generators. The latter costs around ₹10-12 per unit.”
He cited mobile phone towers. Reliance, he said, chose lithium-ion batteries over diesel gensets. Data centres are doing the same. “It is a trap to think the market will take off at $50—not at $100,” he told CarbonCopy. “Instead, we should see this as a question of the customer segments that can switch at the higher price – and keep driving price down through the scale they generate,” he said. “At each $5 gap, there is a 10x market available. At every 10x rise in scale, there will be a 10-15% cost reduction as well.”
What facilitates such reduction, he said, is the sector’s level of technological maturity. “Unlike solar, which rode on large government tenders, storage has reached a technological point where it will move anyway,” he said.
Walawalkar also alluded to thermal power plants. “Instead of building for peak demand and subsequently running them at lower plant load factors, generators could set up smaller capacities with storage, and use that to smooth out their production variations.”
Listening to him, it seems that India might see the rise of more dispersed storage solutions. This mirrors trends in generation.
This raises a new question. If BESS/electrolysers evolve into grid-level storage—or if the market finds dispersed solutions for grid stabilisation viable—what does the future hold for hydel?
The answer depends on the size of their reservoir.
Looking at scale
Most storage technologies discharge their stored energy quickly. A battery with 1 MW capacity and 4 MWH of usable energy will have a storage duration of 4 hours. RoR projects too come with small reservoirs. So do pumped storage projects. Built with two reservoirs, these generate power by letting water flow to the lower one—and then use cheap off-peak power to pump water up to the higher one.
All of these are better at tackling short-lived variations. The only long-duration storage solution we have is multi-purpose hydel projects. Given their large reservoirs, they can respond to seasonal variations in power supply and demand mismatches—apart from serving other ends like irrigation. The rest will have to compete.
What will such competition look like? One part of the answer again comes back to price. RoR projects come with high initial costs, but almost nothing thereafter. Pumped storage projects, given their need to push water back into the upper reservoir, come with higher operating costs. The batteries comprising BESS will need to be periodically replaced. “Will battery storage of a very large capacity have similar life and running costs as a large pumped hydro?” asked Santosh Khatelsal, the founder of Bangalore-based Enerparc Energy. “Perhaps not. There will be a size/capacity till which this may be true but not above this size.”
At the same time, as Walawalkar said, India is moving towards a future where both generation and storage will see decentralisation. In generation, rooftop solar was a step in that direction. Micro-grids, of the kind Reliance is said to be planning, are another.
In storage, too, as generators and users create their private arrangements, the nature of supply-demand mismatches will change.
There will still be room for large, centralised generation and storage. If there is a sudden surge of demand in a grid, said Khatelsal, batteries cannot amp up their output. India will need stand-by generation as well. Needing to ramp up quickly, this might take the form of hydel or gas.
And yet, all hydel projects don’t have the same capacity to step up when needed. Just as solar’s output drops during the monsoon, hydel has its ebbs as well. “You will not get your full storage of 10-15 hours all days in a year,” said Walawalkar. “Maybe you will get 10 hours worth of storage for 100 days, 6 hours’ worth of storage for 200 days, and 2 hours worth of storage for 300 days or maybe even the full 365 days,” he added to illustrate the problem.
And so, the question is about the frequency of use. Will hydel be more competitive than BESS et al for meeting smaller variations in demand? If not, how frequently will it be called into service?
NHPC, Neepco and Greenko did not respond to questions sent by CarbonCopy asking for their perspective on this matter.
The end game
A clutch of factors complicate this picture further. At a time when India’s new energy architecture is taking shape, long-term lending can be riskier than usual. “With solar and storage, I will get delivery after two years,” said Kawale. “The value of this predictability is very high.” Hydel, on the other hand, will struggle to raise funds.
To make matters worse, the Indian government has rosy projections for renewable energy expansion. India’s energy targets are increasingly top-down, not bottom-up. The country will need 80 GW of storage and grid stabilisation if it adds 343 GW of renewable capacity between now and 2030. This may not happen. Not only has the country missed previous targets—shortly after coming to power, the current administration said India will get 175 GW from renewables (excluding large hydro) by 2022 target; what it managed was 100 GW.
“This is not just bad planning,” said a former energy researcher at Delhi’s Centre for Policy Research. “Most of the current crop of bureaucrats cut their teeth during the nineties and 2000s when high growth rates could be taken for granted. And so, there is a certain techno-optimism that these projections will take care of themselves.” This is what happens, he said, when the state works in mission mode.
Put it all together and a hazy picture emerges. On one hand, unlike hydel, batteries and electrolysers are yet unproven technologies at the level of a grid. Despite being more expensive, hydel comes with fewer import dependencies too.
At the same time, there is a danger that the current state of affairs, where low solar tariffs have elbowed out hydel power, might repeat itself in storage.
And so, the government has created hydel purchase obligations. DISCOMs will have to buy power produced by all the dams commissioned after March 8, 2019, even if it is costlier. But DISCOMs are cash-strapped. Compelled to buy costly power, they will delay payments; or their economics will weaken further; or, the volume of electricity they buy will fall.
They will also pass these costs onto their customers, which might create a situation similar to that of rooftop solar. “These price distortions are borne mostly by the industrial users,” said Srikanth, the professor at Bangalore-based National Institute of Advanced Studies. “And so, whenever they get a chance, they desert DISCOMs.”
If hydel pushes up the price of power from DISCOMs, will users opt for more decentralised solutions? In that case, given the caveats about the government’s targets, how big a market will RoR and pumped storage projects be left with?
Given these risks, the private sector is not stepping into hydel. The sector is back with the public sector. On their shoulders lies the gamble.
(This is the second part in a three-part series. Read the first part here. Next: The costs of an abortive hydel boom)