E-Paper
AI-driven transmission and storage can unlock green power
This article is authored by Akshay Hiranandani, CEO, Serentica Renewables.
India has proved it can build renewable power fast. Record additions over the past year confirm that scale is no longer the sector’s central challenge. India accomplished its largest power capacity expansion when it added 52.5 GW of new capacity through renewable energy sources between April 2025 and January 2026. The total non-fossil capacity increased to over 270 GW because solar and wind combined to create about 75% of that capacity expansion. This scale now places new demands on how power is managed and delivered.
But scale is no longer the constraint. Reliability is, and increasingly, flexibility is. Flexible power is now the defining requirement for hyperscalers and neoclouds driving the AI revolution. It is a reliable power that can ramp up or down in real time, aligned to grid conditions and compute demand. The transition is about reliable, low-carbon power at scale.
The unpredictable nature of renewable energy sources requires advanced methods to maintain power supply balance according to current demands, which traditional methods and outdated grid systems cannot support. This is precisely where the idea of Grid 2.0 becomes relevant. The next phase of grid development is not just about adding wires. It is about adding cognition. Artificial Intelligence (AI) gives the grid the ability to anticipate rather than react, turning it from a passive network into an adaptive system.
India’s national grid has expanded rapidly in recent years, with transmission lines and transformation capacity growing strongly, reshaping the energy landscape. But planning and operations remain sectoral as generation is separate from transmission, and both are largely separate from digital real-time control. For a rapidly changing energy mix, this approach risks inefficiencies and bottlenecks. Inefficient transmission squanders green energy through curtailment and forces reliance on fossil backup, exactly the opposite of the clean-power promise. Analysts and industry groups now recognise that transmission needs not just expansion, but smarter, data-driven planning and real-time control. Digital intelligence, enabled by AI, allows operators to fine-tune flows, forecast stress points, and reduce congestion that static models often miss.
Rising renewable penetration is increasing grid firming costs globally. In parallel, falling storage costs are making flexibility viable at scale. Batteries, pumped hydro, gas peakers, SMRs, and flexible infrastructure such as onsite BESS, solar-plus-storage, and behind-the-meter systems are now core to enabling AI-scale demand.
The current reliability assessment process depends on storage as its main component. India will experience a substantial increase in its storage-backed renewable capacity, which will reach 25 to 30 gigawatts by 2028. The development shows that people have begun to understand storage functions as a system optimiser instead of its traditional use as a storage buffer. The two storage technologies, batteries and pumped hydroelectric power, have the ability to transform electricity supply throughout multiple hours of operation. The system uses artificial intelligence to optimise charging and discharging processes by analysing weather data along with power needs, market changes, and battery state information. Storage functions as an operational asset within the grid while it works together with the system to maintain balance between power supply and demand.
This enables the shift to round-the-clock (RTC) power, where wind and solar, combined with storage, are engineered into a consistent, dispatchable supply that can replicate and displace baseload generation.
Policy movements in 2025–2026 demonstrate that India’s leadership recognises the need for smarter grids. The extension of transmission charge waivers for storage projects until 2028 is one such step, reducing cost barriers and encouraging storage deployment tied to renewables. But these fiscal incentives must be matched with clear rules for data sharing, grid access, and auction design that give investors’ confidence they can deploy AI-enabled systems at scale. Without coherent regulatory frameworks for digital grid operations, storage, and transmission investments risk being underutilized or slowed by uncertainty.
The phrase AI holds no specific definition when it appears in power circles. In practice, AI for grids means prediction, optimisation, and automation at scales humans cannot manage manually. Predictive analytics can anticipate supply shortfalls hours ahead by integrating meteorological, generation, and load data streams. Optimisation algorithms balance multi-site storage assets, reduce curtailment, and schedule dispatch to maintain system stability. Automation reduces reliance on manual control centre decisions. Together, these capabilities transform the grid into an adaptive system aligned with the principles of Grid 2.0.
For hyperscalers, the implication is direct: AI compute is constrained by power. Access to firm, flexible, and dispatchable renewable power enables better planning of infrastructure, capital, and risk. RTC solutions introduce time-block certainty, aligning compute demand with power availability.
Power is no longer just supplied—it is shaped. Workloads can shift, intensity can be modulated, and non-critical compute can move to periods of high renewable output or lower prices, directly linking power systems with compute infrastructure.
Corporate investors have begun to recognise power-grid intelligence as a vital component of strategic infrastructure. The requirements for substantial funding include $100 billion commitments from major Indian industrial groups to establish AI-enabled renewable energy systems, reflecting the growing integration of power and compute systems.
At Serentica, this approach has been embedded from the outset. As a RTC IPP, the focus has been on solving for continuous industrial demand in sectors such as aluminium and zinc that operate 24/7.
This has translated into execution at scale. A time-block assurance PPA of 60% on a 530 MW load is being delivered through a multi-site wind-solar mix supported by multi-GWh BESS across four locations. In parallel, projects delivering up to 90% assurance are being developed, combining solar and storage to provide up to 14 hours of dispatchable evening power. These are part of a broader 8.5 GWp portfolio delivering firm and flexible renewable power.
India’s power landscape is at a juncture where leadership in the global energy transition depends on solving integration challenges. Clean energy targets and record capacity additions set the direction. Intelligent transmission and storage systems, powered by AI, will determine whether that expansion delivers reliable, affordable power.
The metric is no longer capacity added. It is power delivered—reliably, every hour.
India’s renewable revolution will be judged not by how many megawatts it installs, but by how effectively it delivers power every hour of every day. Grid 2.0 is the architecture that will make that possible, and AI-enabled transmission and storage systems are its essential foundation.
(The views expressed are personal)
This article is authored by Akshay Hiranandani, CEO, Serentica Renewables.
