Grid Memory, Virtual Power Plants, Vehicle-to-Grid Economy

'Households are leading the energy transition, turning their homes into mini power plants.'. David Parker, Chair, Australia's Clean Energy Regulator

❓ What You'll Learn

• Why did a single Google-Xcel-Form Energy deal become the largest battery project ever announced?
• How does a retired coal plant end up holding 92 Tesla Megapacks?
• Why will storage production outgrow EV battery production 4-to-1 in 2026?
• What turns a parked car into a paid grid asset?
• Where does the real margin sit once battery cells become a commodity?
• Which state wrote the rulebook that's now pulling V2G pilots toward it?
• Why do utilities suddenly care more about permitting speed than battery supply?

💎 Why It Matters

Solar and wind are now the cheapest power on Earth.

They're also the least reliable, minute to minute.

Storage is the layer that turns cheap into dependable.

🔍 Problem

Solar peaks at noon.

Demand peaks in the evening.

Wind disappears for days at a time.

Grids without enough storage burn gas to cover the gap.

💡 Solution

Stack three kinds of storage, each built for a different timescale.

Lithium-ion for hours. Long-duration chemistries for days. Millions of small distributed batteries, coordinated by software, for everything in between.

🏁 Players

Grid-Scale Storage

• Tesla • Megapack backlog worth $29 billion in tracked projects, a new $200 million Houston plant targeting 50 GWh a year, and 92 Megapacks already trucked into a retired Alabama coal plant for its first utility-scale system.
• Fluence • The AES-Siemens integrator. It's Mosaic software, not its hardware, that is the actual product, orchestrating over 20 GWh of deployed storage.
• CATL & BYD • CATL is piloting sodium-ion with China's State Grid. BYD controls its own supply chain from the lithium mine to the power-conversion system.

Long-Duration Chemistries

• Form Energy • Its iron-air batteries just landed a 300 MW / 30 GWh deal with Google and Xcel Energy, built to discharge for days, not hours, to keep a Minnesota data center running.
• Sodium-Ion & Flow • Lithium-free, cost-first chemistries entering pilot-to-commercial scale through 2026-2028.

Orchestration & V2G

• Tesla Autobidder • Batteries as active traders in real-time power markets, not passive backup.
• Nuvve & The Mobility House • The software layer letting EV fleets discharge back into the grid.
• Sunrun × BGE • Piloting vehicle-to-home with Ford F-150 Lightnings across dozens of Maryland households.

🔮 Predictions by Aybars D.

• Storage becomes a bigger business than the batteries in your car. LG Energy Solution expects stationary storage production to outgrow EV battery production 4-to-1 in 2026. The grid, not the road, is now the bigger buyer.
• Long-duration chemistries stop being a lab story by the early 2030s. Form Energy didn't sign a pilot. It signed a utility contract for the largest battery project ever announced.
• Vehicle-to-grid capacity closes in on stationary storage by 2035-2040. Maryland wrote the first comprehensive V2G interconnection rules in June 2025. Every EV added after that is a grid asset nobody had to build separately.

☁️ Opportunities

• Build orchestration software, not battery cells • Lithium-ion pack prices are already headed under $100/kWh. Cells commoditize. Dispatch, bidding, and market-access software don't.
• Build for fleets before individuals • School buses, delivery vans, and municipal vehicles are easier to coordinate into V2G than millions of scattered private cars.
• Build interconnection intelligence • In markets like ERCOT, a battery that can't connect can't earn; permitting speed is now the binding constraint, not battery supply.
• Build the trust layer for V2G • Battery-health data, transparent degradation warranties, and simple compensation models are what turn a skeptical driver into a participant.

🏔️ Risks

• Hardware Commoditization • Lithium-ion prices have fallen 97% since 1991. Great for adoption, brutal for anyone competing on cell margin alone.
• Grid Bottlenecks • Interconnection queues, not manufacturing capacity, are the real constraint in the fastest-growing U.S. markets.
• Chemistry Lock-In • Betting early on one chemistry risks being undercut by whichever one wins the next cost curve.
• Consumer Trust • V2G doesn't scale if drivers believe it wrecks their battery; warranties have to answer that before adoption does.

🔑 Key Lessons

• Storage isn't backup power anymore. It's market access, the difference between selling electricity when the sun allows it and selling it when the grid actually needs it.
• The car is becoming a power plant. Most vehicles sit idle over 90% of the time. V2G is what ride-sharing did to idle cars, applied to idle batteries.
• Software is where the margin moves once hardware is cheap. Fluence and Tesla Autobidder are the tell; the money is in the layer that decides when a battery charges, not the battery itself.

🔥 Hot Takes

• The next energy unicorn may not own batteries • It may coordinate everyone else's.
• The future grid will become a protocol • Millions of assets will respond to price, weather, demand, and grid stress in real time.
• By 2040, 'the grid' won't be a place • It'll be a protocol, millions of batteries, EVs, and buildings responding to price and weather in real time.
• The real AI-energy story right now isn't nuclear • It's storage sits next to the data center, running years before any new reactor comes online.
• The real AI-energy story is not only nuclear • It is storage. Data centres need reliable power before new reactors arrive.
• Battery companies without software will become suppliers • The margin moves to dispatch, aggregation, trading, financing, and control.
• The parked car becomes a power plant • Mobility is the first use case. Grid flexibility is the second.

😠 Haters

'Batteries only last four hours.'

• The bet isn't on nuclear's past. It's on a 20-year contracted demand nobody had before. AI gave nuclear something it never had: a buyer willing to sign the check today for power delivered a decade from now.

'Vehicle-to-grid will destroy car batteries.'

• That risk is real and worth watching closely, but it's also why so many of these deals restart idle plants (Three Mile Island) or fund new baseload capacity, rather than simply outbidding households for existing supply.

'Storage is just backup power.'

• True today, mostly false by 2030. NuScale already holds NRC certification, and Rolls-Royce SMR's Czech approvals are targeted for the same year construction is meant to start elsewhere. The gap between announcement and operation is closing fast.

'Lithium-ion already won.'

• Lithium-ion only won the first storage layer. Hours, days, seasons, congestion, peak demand, and resilience are all different markets, and each one needs a different tool.

'Utilities move too slowly.'

• Utilities move slowly until reliability becomes political. Blackouts, AI data-center demand, and grid congestion are now turning storage from an innovation budget item into core infrastructure.

🔗 Links

• IEA Energy Storage Outlook • Baseline research on why storage must scale rapidly to support renewable growth.
• Utility Dive on Long-Duration Storage • Coverage of grid-scale long-duration storage projects and utility deployment.
• Mordor Intelligence Grid-Scale Battery Market Report • Market sizing and competitive dynamics for utility-scale storage.
• Nuvve Vehicle-to-Grid Outlook • A grounded view of where V2G stands between pilot stage and commercial scale.
• Virtual Power Plant Primers • Technical explainers on how distributed batteries, EVs, and smart buildings become one dispatchable resource.

See you in the next one. To be continued..

Written by Aybars Dorman | July 1, 2026