The Energy Transition’s Next Bottleneck Is Intelligence

Battery storage is scaling globally at historic speed. In just a few years, grids around the world have added record volumes of battery capacity, driven by renewable penetration, energy security concerns and surging electricity demand. Global installations reached approximately 270 gigawatts (GW)/630 gigawatt-hour (GWh) last year, achieving 43 percent year-over-year growth, with forecasts projecting exponential expansion to 1,545 GW by 2034.

The industry celebrates the buildout. Investors celebrate the capital flows. Policymakers celebrate the gigawatts. Yet beneath the surge in capacity lies a growing vulnerability: we are building fleets faster than we are building the intelligence required to operate them. 

The next bottleneck in the energy transition will be coordination.

Capacity is scaling. Markets are changing. Complexity is exploding.

When batteries first entered modern power markets, the primary challenge was integration. Electricity systems were built for one-directional flow; large central generators pushing power outward. Storage required rethinking dispatch logic, settlement rules and ancillary services. That challenge was mostly solved relatively quickly.

Today, batteries are indispensable. They sit alongside solar and wind farms, stabilize microgrids, serve industrial facilities and increasingly support data centers whose demand is accelerating with the expansion of A.I. infrastructure, projected to grow 160 percent by 2030. But while deployment has accelerated, operational sophistication has not kept pace. As electricity demand increasingly intersects with the growth of A.I. infrastructure and digital economies, the intelligence layer managing energy assets will become as strategic as the assets themselves. 

Across applications, batteries are often oversized to hedge uncertainty, operated conservatively to manage risk or dispatched based on static forecasts that fail to capture evolving market structure. As a result, revenue opportunities are quietly left on the table as market rules, competition and topology shift.

Installed capacity will continue rising for the foreseeable future. But capacity does not scale in isolation. As batteries scale, they reshape market dynamics themselves. Markets are also evolving for many other reasons, such as fuel retirements, renewable penetration, demand shifts, transmission upgrades and policy changes.

Meanwhile, the diversity of battery applications continues to expand. A battery is no longer a single-use asset. It participates across stacked services, regions and time horizons. Each additional gigawatt of storage adds not only supply but strategic interaction. As deployment deepens and use cases expand, the grid becomes more interdependent and market behavior more sensitive to how storage operates.

What was a winning dispatch strategy last year may underperform this year. This is a time-varying state, not a repeating cycle.

A recent analysis of Australia’s National Energy Market (NEM) illustrates this directly. As grid-scale battery capacity surged—with projects like AGL’s 500 MW Liddell Battery and Neoen’s 560 MW Collie Battery coming online—batteries shifted from passive price-takers to active price-setters, clearing energy prices up between 35 percent and 50 percent of the time during peak intervals. When modelers dropped a hypothetical 500 MW/2,000 MWh battery into the market, using traditional forecasting and dispatch assumptions, projected revenues came in at around AU$400 million over the course of 1.5 years. However, accounting for the battery’s own impact on market-clearing prices, as well as realistic dispatch scenarios in every interval given the market depth, only 25 percent of that figure was realistically capturable, indicating a AU$300 million gap driven entirely by the feedback loop between dispatch strategy and market response. The strategy that would have worked when batteries were minor players in a deep market is now the strategy most likely to result in weak ROIs for the project.

Below are three examples of how complexity and market dynamics continue to evolve. 

Batteries begin to shape the market

In theory, adding batteries to the grid should dampen volatility and enhance stability by absorbing excess supply and discharging during scarcity. In practice, as storage penetration increases, batteries begin to shape the market itself.

In markets such as Australia’s NEM, Texas’s ERCOT and Great Britain’s National Grid, rapid battery deployment has coincided with, or, as some would argue, led to, the compression of certain ancillary service revenues and arbitrage spreads. What was once a deep opportunity pool is becoming increasingly shallow as more assets pursue the same signals.

When too many assets chase the same opportunity, spreads compress and signals shift. Scarcity spikes can trigger synchronized responses, flooding the market and erasing the very premium that triggered the action. At low penetration, storage is a price taker. At scale, storage becomes a price maker.

Traditional forecast-and-position strategies assume the market is exogenous. But as batteries scale, the act of positioning the asset moves the market itself. Strategy is no longer only about predicting price, but it is about anticipating how your own dispatch, combined with others’, reshapes the system.

The complexity compounds at the portfolio level. Dispatch decisions at one asset affect congestion, spreads and revenue opportunities for others in the same fleet. A strategy that maximizes single-asset revenue may cannibalize value across the portfolio.

This is fundamentally a coordination challenge, which calls for advanced intelligence and automation. Hardware does not coordinate itself.

The grid constantly rewrites itself

Structural change adds another layer of complexity. A 500 MW battery connects to the grid. A coal plant retires. A large data center is commissioned. Electric vehicle load surges. A transmission upgrade reshapes congestion patterns. Each of these developments alters not only the magnitude of supply and demand but their profile, geography and timing. Power flows shift. Regional spreads move. Ancillary requirements evolve.

Coordination logic designed for yesterday’s grid becomes inefficient if it cannot evolve alongside these changes. Energy markets are no longer static equilibrium systems, but rather adaptive, time-dependent networks.

Policy and market signals alter operation strategies

Market design decisions also reshape operational strategy. Take Australia’s NEM as an example. For 2025–2026, the market price cap increased to $20,300/MWh, up from $17,500/MWh the previous year. For batteries, even a handful of such price intervals can materially shift annual returns. The increase alters payoff asymmetry, increases tail exposure and reshapes hedging strategies.

Market design changes can also create entirely new operational opportunities. In late 2023, the NEM introduced a 1-second frequency control ancillary services (FCAS) contingency product, expanding the set of ancillary services that fast-responding assets such as batteries can provide. In regions like South Australia, this new product quickly became a material revenue opportunity for batteries capable of participating and co-optimizing across multiple FCAS markets.

Such policy instruments are tied to reliability economics. But they are also signals: the system’s risk profile and operational landscape are evolving, and operational strategies must evolve alongside them.

The missing layer: operational intelligence

The paradox of the current transition is this: we have solved the procurement problem faster than the decision-making problem. 

Global investment in battery energy storage projects exceeded $54 billion in 2024, a 73 percent increase over recent years. Yet investment in operational intelligence—systems that continuously simulate, coordinate and optimize fleets of assets in real time—as well as portfolio coordination and digital testbeds, remains a small fraction of physical capital deployment.

In other capital-intensive industries, such as aviation, finance and logistics, complex distributed systems are managed through layered intelligence architectures. Simulation precedes action. Feedback loops refine strategy. Adaptive systems replace static rules.

Energy ecosystems are becoming equally complex, yet operational practices remain rooted in a simpler era.

What happens when complexity outpaces adoption?

If intelligence and its adoption fail to scale alongside hardware, the consequences will be primarily economic, but potentially systemic. Underperformance reduces investor confidence. Compressed margins slow new deployments. Poorly coordinated fleets may even increase volatility during scarcity events. Deployment goals depend not only on capital and policy, but on confidence that assets will perform as modeled.

The energy transition is entering a new phase. The first chapter was about scale. The next is about precision, intelligence and automation. Batteries are now core infrastructure that require an intelligence layer equal to their physical scale. If we fail to build it, complexity will continue to compound.

Markets rarely forgive unmanaged complexity.