Last month, we hosted the Storage & Grid Flexibility Private Breakfast Forum in New York City, convening a curated group of ~50 leaders across utilities, corporates, investors, and startups. In collaboration with ecosystem partners including Urban Future Lab at NYU Tandon School of Engineering, Third Derivative, and Greentown Labs, the discussion focused on how flexible energy systems will shape the next phase of the energy transition.
Across a mix of expert insights, panel perspectives, and small-group discussions, a consistent theme emerged. The challenge is no longer building supply. It is making the system work.
The data supports this shift. By 2030, 104 GW of generation retirements are expected to be replaced by 209 GW of new capacity. At the same time, interconnection queues now exceed 1.8 TW of active generation and storage requests, while transmission congestion costs reached roughly $12 billion in 2024. The system is expanding, but not necessarily becoming easier to operate.
The energy transition is often framed as a generation problem. In practice, it is increasingly a coordination problem.
For the past decade, the mandate was clear. Build renewable capacity at scale. The next phase is different. It is about connecting, coordinating, and operating that capacity in a system where supply is variable and demand is becoming more dynamic.
The system is not short electrons. It is short of flexibility. The ability to shift, store, reroute, or reduce demand in response to real-time conditions.
That flexibility is emerging across five distinct layers of the grid. Each behaves differently. Each attracts capital differently. And increasingly, each scales at a different speed.
The Landscape: Five Ways to Flex a Grid
Understanding this market requires a clear framework. Grid flexibility does not emerge in one form. It exists across five distinct layers:
- Demand-side flexibility: shifting or reducing load through VPPs, demand response, and customer participation
- Storage-enabled flexibility: batteries and other storage technologies that move energy across time
- Grid flexibility coordination: DERMS, orchestration platforms, and marketplaces that coordinate assets
- Network-side flexibility: dynamic line rating, topology optimization, and transmission technologies
- Supply-side flexibility: dispatch optimization and generation-side responsiveness
This taxonomy matters because it reveals a core truth. Flexibility is not a single market. It is a stack. And today, capital is not flowing evenly across it.
Where Capital Went, and What It Reveals
Investment into grid flexibility technologies accelerated rapidly through 2023 and 2024, then pulled back in 2025.
The pattern is visible in the Net Zero Insights dataset. Capital concentrated at later stages during the peak years, followed by a visible slowdown in early and growth stage deployment. This is not a uniform retreat. It is a rebalancing.
The most important signal is how capital was distributed across the flexibility stack.
The data shows a clear concentration:
- Electrochemical storage at $3.2B
- VPP aggregation platforms at $1.8B
- DERMS at roughly $900M
- Network-side technologies significantly smaller
This is not just a picture of where the capital went. It is a picture of what capital understands.
Storage and software are legible to capital. They have defined product boundaries and fit how private markets underwrite scalable assets and platforms. Network-side flexibility does not. Its value is often indirect, realized through avoided infrastructure, reduced congestion, or improved system utilization. At the same time, utility incentives still favor capex over service-based solutions.
The implication is structural. The system is investing heavily in flexibility supply while underinvesting in the mechanisms required to access and coordinate it. That mismatch becomes visible only at deployment.
The Startup Opportunity Gap & The Regulatory Battering Ram
The imbalance across the flexibility stack creates whitespace. The technologies with the highest system leverage are not always the ones attracting the most capital. In many cases, they are the hardest to package, sell, and underwrite.
The ecosystem mirrors the capital flow:
- Storage and demand-side dominate startup formation
- Coordination layers emerging but fragmented
- Network-side innovation remains thin
This is not a signal of low importance. It is a signal of structural difficulty. Network-side solutions often monetize indirectly through avoided grid upgrades, reduced congestion, or faster interconnection timelines.
These benefits are real, but they do not map cleanly to venture-scale business models. At the same time, the primary buyers of these solutions are often incentivized to invest in infrastructure rather than procure services. The result is an opportunity gap that is not just technical, but institutional.
This is where policy must act as the battering ram. Major regulatory shifts, such as FERC Order 2222 in the U.S., which mandates that wholesale markets open up to distributed energy resources (DERs), are designed precisely to force these institutional bottlenecks open. Without policy mandates compelling utilities and grid operators to value and integrate these software and coordination services, the gap will persist.
A glimpse of what happens when that battering ram lands can be seen in NineDot Energy, one of the Urban Future Lab at NYU Tandon’s School of Engineering’s portfolio companies. They are building community-scale battery storage across New York City, with a 60+ project pipeline and hundreds of megawatts in development. Its recent $431 million financing to deploy 28 projects (124 MW / 494 MWh) highlights that these assets are technically viable and financeable at scale.
What makes this possible is not just the technology, but the policy scaffolding around it. Programs like New York’s “Solar for All,” combined with wholesale market access reforms, create revenue visibility from grid services, capacity, and energy arbitrage. In effect, policy converts avoided system costs, reduced congestion, deferred upgrades, and improved resilience.
NineDot sits precisely at the intersection of the “network-side” gap: its batteries are physically distributed, locally sited, and deeply tied to grid constraints, yet monetized through market participation and policy-enabled revenue streams.
Utilities vs. Hyperscalers: Two Theories of Flexibility in the AI Era
Flexibility is not only shaped by technology. It is shaped by who is deploying it. The data reveals two distinct models emerging in parallel, driven by a massive, unprecedented shift in the market.
Driven by the explosion of Artificial Intelligence and widespread electrification, the grid is facing its first major surge in load growth in two decades.
Utilities are focused on demand-side programmes, storage deployment, and operational tools to manage grid constraints. This reflects their role. They are responsible for maintaining system reliability within regulatory and infrastructure constraints.
Hyperscalers (the major tech companies building massive data centers) behave differently:
- deploying demand-side and storage assets behind the meter
- selectively investing in coordination and emerging technologies
They are not optimizing the grid. They are optimizing their own reliability against a strained system.
This leads to a clear divergence. Utilities are operationalizing flexibility within the existing system. Hyperscalers are building parallel flexibility stacks outside it to guarantee uptime for their AI workloads.
The question is whether these paths converge. If behind-the-meter assets remain isolated, they reduce pressure on the grid but fragment system visibility. If they are integrated, they become a highly potent new layer of grid infrastructure.
In New York City, this dynamic is visible, even without hyperscale data centers. Flexibility is instead driven by dense commercial load, constrained substations, and aging infrastructure.Con Edison’s Brooklyn-Queens Demand Management Program\ showed that distributed resources can defer ~$1B+ in traditional upgrades. At the same time, large buildings across the city participate in demand response, effectively acting as dispatchable load. Combined with New York’s Value of Distributed Energy Resources (VDER) framework, which compensates assets based on when and where they reduce congestion, these systems are increasingly functioning as a parallel layer of grid infrastructure.
The result is a distinctly urban model of flexibility: not centralized megaprojects, but highly localized, coordinated networks of distributed assets deployed where constraints are most acute.
Demand Flexibility Scales Faster Because It Avoids the Grid
Not all flexibility technologies face the same friction.
The pattern is clear:
- Demand-side flexibility dominates experimentation and programme deployment
- Storage dominates infrastructure deployment and procurement
The difference is not technological. It is procedural. Demand flexibility can be deployed as a programme. It does not require interconnection, permitting, or large-scale hardware deployment. Storage does.
This creates a speed asymmetry. Demand-side solutions scale through software, incentives, and aggregation. Storage scales through infrastructure, capital, and physical deployment.
But software’s monetization still depends on integration, performance verification, and utility procurement structures. That friction limits how quickly it translates into system value. The implication is that early flexibility markets are shaped by what can be deployed fastest, not necessarily what delivers the most system value.
A System Rebalancing
Across all five layers of flexibility, a consistent pattern emerges.
Capital has moved first toward what is easiest to finance. Batteries, aggregation platforms, and software layers with clear product models.
Deployment is now encountering what is hardest to scale. Interconnection, network capacity, and coordination across distributed assets.
This creates a second phase of the market. The first phase built flexibility assets. The next phase determines whether the system can actually use them. And that shift will define where the next wave of opportunity sits.
Dive Deeper
Get access to the full presentation. Includes additional charts and extended insights on grid flexibility and storage.
