Your Home Battery Is Worth More to a Data Center Than It Is to You. The Builder Who Installed It Didn't Mention That.
On June 24, Sunrun, Tesla, and Renew Home announced a partnership to aggregate more than 16 gigawatts of flexible energy capacity from residential batteries and smart thermostats into a single distributed power plant. Sixteen gigawatts is roughly fourteen nuclear reactors' worth of dispatchable electricity, drawn not from a facility with cooling towers and security fences but from hundreds of thousands of ordinary houses where the homeowner probably thought the battery was for keeping the lights on during storms.
Sunrun's stock jumped 25% in a single trading session, the kind of move that happens when institutional investors suddenly realize an entire asset class has been mispriced. Wall Street understood something that most homebuilders haven't caught up to yet: that battery in your garage is no longer just backup power. It is a revenue-generating grid asset. And the people willing to pay the most for access to it are the companies building AI data centers.
The Grid That Broke First
PJM Interconnection operates the electricity grid across 13 states and Washington, D.C., serving about 70 million people. It also hosts Northern Virginia, the densest concentration of data centers on the planet. What happened next should terrify anyone who pays an electricity bill in the mid-Atlantic and excite anyone who owns a home battery.
PJM's forward capacity price, which is essentially the market's bet on what reliable electricity will cost, went from $28.92 per megawatt-day in the 2024/25 delivery year to $329.17 in 2026/27. That is an eleven-fold increase. Wholesale power costs jumped 76% in the first quarter of 2026 alone, with data center demand identified as the primary driver. AEP, one of the region's largest utilities, has threatened to leave the market entirely. PJM paused applications for new generating sources from 2022 to April 2026 because the queue was so backlogged that processing it had become impossible.
Goldman Sachs projects U.S. data center power demand will reach 41 gigawatts this year and 66 gigawatts in 2027, and by 2028, data centers could account for 12% of total U.S. electricity demand, according to the Department of Energy, a figure that would have seemed absurd five years ago and now looks conservative. Centralized generation cannot keep up. Permitting a new natural gas plant takes three to five years. A nuclear plant takes a decade. A battery in someone's garage goes in before lunch.
The $500 Question Your Builder Should Be Asking
Making a new home "VPP-ready" during construction costs between $500 and $1,000 in marginal materials and labor. That buys a dedicated 240V circuit from the main panel to a battery mounting location, a pre-run conduit path, a mounting surface rated for the weight, and a CT clamp hookup point for grid metering. Panel and Wi-Fi are already there. All that's missing in most new homes is the conduit and circuit connecting them to the spot where a battery will eventually sit.
Retrofitting the same capability into an existing home costs $2,000 to $5,000, depending on panel capacity, conduit routing through finished walls, and whether the existing service needs an upgrade to accommodate bidirectional flow. Most homes built before 2020 have 100-amp or 150-amp panels that need a $2,000 to $4,000 upgrade before a battery can interconnect with the grid, assuming the utility allows interconnection at all, which in some territories requires its own permitting process that takes weeks.
A builder who spends $500 during framing creates the same infrastructure that costs $5,000 to add after the drywall is up, a 4:1 to 10:1 cost penalty for waiting. Almost all of it comes from the difference between running conduit through open stud bays during framing versus fishing wire through finished walls after occupancy, which means cutting into drywall, patching, repainting every surface you touched, and scheduling an electrician for a return trip that costs $150 before he picks up a tool.
What a Home Battery Actually Earns
Base Power, the Austin startup backed by a16z and Thrive Capital, has deployed more than 500 megawatt-hours of residential battery storage in Texas. Its model works like this: you pay $695 to $995 upfront for a 25 or 50 kilowatt-hour battery that Base Power installs, owns, and operates. You commit to buying electricity from Base Power for three years at 8.5 cents per kilowatt-hour plus whatever delivery fees the local utility charges, a rate that undercuts most Texas retail plans and locks in predictability that construction loan officers notice when they calculate your total monthly housing cost. Base Power charges the battery when grid prices are low and discharges it when prices spike, pocketing the arbitrage and passing some savings to you through rates that run 25% below the local utility in Illinois, where Base just launched.
Each battery costs approximately $10,000 to install before a $3,000 federal tax credit under the Inflation Reduction Act, which covers 30% of residential energy storage through 2032. At a net cost of $7,000 after the credit, Base Power estimates a 3.5-year payback from stacked retail margins and wholesale arbitrage, with homeowners saving roughly $300 to $600 per year on electricity depending on usage patterns, the local rate structure, and how aggressively the algorithm trades on their behalf during peak pricing windows. Base Power projects $70 million in revenue for 2026, up from $12 million in 2025.
Buy the battery yourself and the math gets considerably better. In PJM territory, where capacity prices have reached $329.17 per megawatt-day, a home with a 10 kW dispatchable battery system participating in capacity auctions could theoretically earn around $1,200 per year in capacity payments alone, before counting energy arbitrage or demand response incentives. That is on top of whatever the battery saves you during outages, which in a state like Virginia or New Jersey increasingly means avoiding days-long blackouts when the grid buckles under data center load and summer heat simultaneously.
Why Builders Are Missing This
Production homebuilders, the companies building 80% of America's new single-family homes, optimize relentlessly for first cost. Every dollar added to the base price compresses margins and pushes the monthly mortgage payment higher in a rate environment where the average 30-year fixed mortgage hovers around 6.5%, pricing real buyers at the margin out. A $500 pre-wire for a battery that might get installed three years from now does not help sell the house today. Almost nobody does it.
But the counterargument is increasingly quantifiable. In markets where VPP participation generates $500 to $1,200 per year in direct revenue or savings, which now includes most of PJM's 13-state territory and all of ERCOT, a $500 pre-wire has a payback period measured in months once the battery arrives. More importantly, the buyer who moves into a VPP-ready home in Northern Virginia or central Texas in 2027 will face an electricity market where doing nothing means absorbing whatever rate increases the data center boom pushes onto residential ratepayers, which in PJM has already been a 76% wholesale increase in a single quarter, and where participating means their home literally pays them during the hours when electricity is most expensive.
New Jersey is projecting that data centers will consume 10% of the state's electricity within four years. A bipartisan federal bill introduced this month would require large-load customers like data centers to cover the costs of new generation and transmission upgrades, which both Microsoft and Google have publicly supported. But legislation moves slowly, and electricity prices move now. A homeowner with a battery and a VPP contract is hedged. Without one, you're exposed.
The 16 Gigawatt Asterisk
The Sunrun/Tesla/Renew Home partnership deserves scrutiny. Sixteen gigawatts represents the combined installed base of home solar panels, Powerwall batteries, and smart thermostats across three companies' residential customer networks, the theoretical maximum if every device participated at full capacity simultaneously. That will never happen. Not every customer will opt in. Not every battery will be fully charged during peak events. Not every thermostat adjustment translates to meaningful load reduction.
Actually deployable capacity today is far smaller. Sunrun said 300 megawatts is "readily available" in Virginia, with 500 megawatts projected by 2030. That is 1.9% to 3.1% of the headline number. If the partnership enters PJM's reliability program, it could offer 1 gigawatt immediately — 6.25% of the announced total. The gap between theoretical capacity and contracted, dispatchable capacity has historically been wide enough to swallow entire business plans, and it is precisely where VPP announcements traditionally collapse into press releases that never produce a follow-up earnings call.
There is also the degradation question. Every cycle costs you, because lithium-ion batteries lose measurable capacity each time they charge and discharge. A battery dispatched daily for grid arbitrage, charged and discharged through hundreds of cycles per year to exploit wholesale price spreads, will degrade faster than one that sits idle in a garage for months waiting for the next hurricane to justify its existence. Most home battery warranties guarantee 70% capacity after 10 years under normal use. "Normal use" in a VPP-participating home means 250 to 400 cycles per year instead of the 50 to 100 cycles a backup-only battery sees. No major VPP operator has published long-term degradation data from high-cycling residential deployments, which means the 10-year revenue projections everyone quotes are built on assumptions about battery lifespans that haven't been tested at scale.
What This Means If You're Building
If you are a builder working in PJM territory, ERCOT, or any deregulated market where capacity prices are rising, the case for battery-ready infrastructure in new construction is no longer speculative. Five hundred dollars for $500-$1,200 in annual returns once a battery goes in. You will not capture that value in the sale price today, but you will capture it in the listing conversation when the buyer's agent asks about energy costs, and you can point to a pre-wired battery bay instead of a blank wall.
If you are buying a new home, ask your builder two questions. First: is the electrical panel rated for bidirectional interconnection? Second: is there a dedicated circuit and conduit run to a battery mounting location? Yes to both? You have optionality that will cost someone else $5,000 to create after they move in. If the answer is no, you are buying a house that was designed for an electricity market that no longer exists.
Limitations
The $500-$1,000 estimate for VPP-ready pre-wiring in new construction is based on typical residential electrical rough-in costs and does not account for regional labor rate variation, which can range from $55/hour in Texas to $120/hour in the Northeast. PJM capacity payment calculations assume a 10 kW dispatchable system and full participation in capacity auctions, which requires aggregation through a qualified VPP operator and is not available to individual homeowners directly. Base Power's 3.5-year payback estimate comes from Sacra's analysis of the company's financial model, not from independently audited results. Note that the 30% IRA tax credit for residential storage is scheduled through 2032 but subject to legislative change. Battery degradation rates under high-cycling VPP use are estimated from laboratory data and manufacturer specifications, not from field measurements of residential VPP fleets at scale. VPP revenue is inherently volatile and depends on wholesale electricity market conditions, weather patterns, grid topology, and regulatory frameworks that vary by state and utility territory.
Strongest counterargument: The entire VPP revenue thesis depends on electricity markets remaining tight and capacity prices staying elevated. If new generation comes online faster than expected, if data center growth slows, or if grid operators solve their interconnection queue backlogs, wholesale prices could normalize and the revenue case for home batteries collapses back to backup-only value. PJM's 11x capacity price increase could also be partly a market anomaly driven by the queue pause rather than a permanent structural shift. Building homes around today's peak pricing assumptions is the same mistake builders made with solar in 2015 when net metering rates were at their peak and utilities subsequently cut them by 30-60% in multiple states. Real risk: $500 in pre-wiring creates optionality for a market that retreats before the homeowner exercises it.