Virtual Power Plants in 2025: Markets, Money, and Readiness

The moment VPPs stop being a pilot and start being a product

Virtual power plants are having a very specific kind of breakout year. Not a press-release spike or a lab milestone, but the quieter gear-change when rules, telemetry, and payments line up so that real portfolios can transact at scale. In the United States, three market moves define 2025: Texas enlarged its Aggregated Distributed Energy Resource pilot and published Phase 3 procedures, New York continued enrolling distributed energy resource aggregations into wholesale markets, and PJM firmed a path for full Order 2222 compliance. Together they convert home batteries, smart thermostats, and small commercial loads from promising demos into bankable grid assets. As transmission planning tightens under FERC 1920-A transmission reforms, local flexibility grows more valuable.

In Texas, ERCOT’s official materials document higher pilot caps and updated qualification procedures that matter for revenue today. See the program’s evolving limits and Phase 3 artifacts on ERCOT’s pilot page for the definitive reference to what counts in 2025 and how it is measured and dispatched: ERCOT’s ADER pilot materials. This is not just about technology. It is about market access, dispatch rights, and meter-grade data.

Below we map where virtual power plants can earn now, how to stack revenue streams without breaking rules, what telemetry and operational readiness actually look like, and how to finance portfolios so they scale through 2026 to 2028.

Where to monetize now

Texas: ERCOT

If you want megawatts that move today, start in Texas. ERCOT’s ADER pilot lets aggregators of distribution-connected devices participate directly in the wholesale market through a Qualified Scheduling Entity. The pilot expanded in 2025 with increased participation caps and a Phase 3 governing document that clarifies registration, telemetry validation, and qualification for energy and select ancillary services. The upshot is simple: more room to enroll assets, clearer test procedures, and a smoother path from portfolio to dispatchable resource.

For a developer, that means your fleet can clear energy in the real-time market and qualify for ancillary services that reward speed and controllability.

What actually sells in ERCOT right now:

Energy in the real-time market when scarcity is likely and prices are volatile.
Ancillary services defined in the pilot, subject to caps and qualification tests. These include fast response products where precise control and reliable telemetry are the difference between revenue and penalties.

Operationally, treat Texas like a wholesale market with distribution DNA. You will need nodal awareness, a dispatch interface, and near real time telemetry that satisfies ERCOT’s validation procedures. Think of it as building a miniature control room for thousands of endpoints that acts like a single peaking plant when called.

New York: NYISO

New York’s aggregation program is open and enrolling, which means small resources can join portfolios that bid into wholesale markets. The model supports mixed fleets, so a virtual power plant can include batteries, controllable load, and demand response, provided the aggregation meets NYISO’s registration and coordination requirements with distribution utilities.

The practical upside is access to day-ahead and real-time energy markets, plus a path into ancillary services as technical capabilities are demonstrated and approved.

In New York, price formation is different from Texas. Locational marginal prices and constraints in downstate zones can reward precise timing and multi-hour planning. That shapes the operations playbook: day-ahead scheduling for revenue certainty, real-time dispatch to capture spreads, and careful attention to baselining so performance credit matches actual flexibility delivered.

The near horizon: PJM

PJM’s market is the largest in the country. It also has the most consequential capacity market for virtual power plants over the next few years. The regulator’s latest summary indicates that PJM’s capacity market will accept distributed energy resource aggregations on a timeline that enables participation in the 2028 to 2029 delivery year, while energy and ancillary services follow on a later date. For developers and utilities, that means 2025 is the right moment to pre-build the enrollment, telemetry, and interconnection foundations that capacity obligations will require. For the official, time-stamped details, see the Commission’s overview of Order 2222 implementation milestones: FERC’s Order 2222 explainer and timelines.

The revenue stack that actually clears

Most virtual power plant pitch decks show three buckets: energy, ancillary services, and capacity. That is correct, but the order of operations and the risk profile change by region. Rising data center demand is amplifying peak risk and volatility, as covered in AI demand reshapes grid planning.

Energy: The daily bread. In Texas, scarcity moments drive outsize returns, so fleet operations emphasize fast discharging and recharging windows tied to weather, load forecasts, and constraint patterns. In New York, day-ahead commitments lock in spreads for batteries and predictable demand flexibility from buildings, with real-time dispatch to improve margin or manage risk.
Ancillary services: The grid’s control-room contracts. Regulation and fast reserves pay for speed, telemetry quality, and availability. In practice that means your control stack must deliver accurate setpoint tracking, verifiable state of charge management, and automatic failover when a device drops. Penalties for underperformance can erase energy margins, so a conservative nomination strategy and rigorous device qualification matter more than squeezing every last megawatt out of the fleet.
Capacity: The slow money that enables financing. New York provides a path for aggregations to count toward capacity needs as implementation matures, and PJM’s timeline creates a large window to structure future-year contracts. Capacity requires proof of dependable availability in correlated stress periods. That is where diversified asset classes in a single aggregation shine. Thermostats smooth long events, batteries deliver peak-hour punch, and commercial load control provides the firm floor that credit committees want to see.

A practical sequencing for 2025 portfolios looks like this: sell energy and selected ancillary services now in ERCOT and NYISO, use today’s telemetry and performance data to underwrite capacity-linked contracts, and prepare for PJM capacity participation as the rules lock in. The operations discipline you build for ancillary services will also make your energy revenues more predictable.

Telemetry and operational readiness without the hand-waving

If you cannot measure it, you cannot settle it. Grid operators require telemetry that is frequent, accurate, and aligned with dispatch instructions. Here is what that means in the field, stripped of jargon.

Device-level truth: Each enrolled site needs a verified meter point and secure communications. The data must be traceable from device to aggregation to the market interface. Think of it as a chain of custody for electrons.
Aggregator control plane: You need a control room. Not a fancy wall of screens, but a resilient stack that receives dispatch, calculates device setpoints, and confirms delivery within the operator’s tolerances. That stack should degrade gracefully. If one communication channel fails or a customer overrides a thermostat, the fleet keeps performing.
Baselines with teeth: Whether you are selling load reduction or battery discharge, the performance is measured against a counterfactual. Invest early in baselining methods that survive operator audits and customer behavior changes. Use shadow settlement to reconcile your internal ledger with the market’s numbers every day.
Distribution coordination: In every Order 2222 implementation, the distribution utility has a say. Expect pre-registration coordination, locational screens, and in some cases feeder-level limits. Treat those interactions like you would an interconnection study for a traditional generator, because that is what they are in practice.
Security and privacy: Wholesale telemetry opens a doorway into customer devices. Close it with encryption, device whitelisting, and role-based access. If you plan to finance portfolios with non-recourse debt or insurance, this is not optional. It becomes part of the diligence checklist.

The technical story is simple to summarize. A virtual power plant must act like a small, dependable power station. That means dispatchable, measurable, auditable, and safe to connect to the grid.

Financing playbooks that scale beyond pilots

Pilots are funded by innovation budgets. Real fleets are funded by repeatable contracts. Here are structures that work now, with the why and the how.

Portfolio tolling with a performance floor

What it is: A buyer commits to take a slice of controllable capacity from an aggregation for a term, with a fixed monthly payment and a revenue share above a performance hurdle.
Why it works: The buyer gets a predictable supply of flexibility. The aggregator converts merchant volatility into a financeable annuity with upside.
How to do it: Anchor the deal on hours when the fleet is most dependable. Use historical telemetry to set the floor, and reserve some device headroom to hit the hurdle even on bad days.

Capacity strip paired with energy call options

What it is: In PJM and New York, sell forward capacity rights from the portfolio and buy call options for high-priced hours to hedge underperformance risk.
Why it works: Credit committees understand capacity cash flows. Options cap the tail risk when a heatwave or a storm shifts the load shape.
How to do it: Use weather-normalized fleet performance to size the strip. Calibrate option volumes to worst-case derates from customer overrides and communications outages.

Tariffed on-bill financing for customer devices

What it is: The utility funds batteries or controls at customer sites and recovers costs on the bill tied to the meter, not the person.
Why it works: Customer churn does not break the business case. The utility earns on both system benefits and customer bill savings.
How to do it: Align the tariff with measured peak reduction and participation hours. Keep customer terms simple. Require opt-out rather than opt-in for event participation, with clear comfort bounds.

Transferable tax credits plus flexibility offtake

What it is: Pair the federal investment tax credit for standalone storage with a long-term grid services contract from a utility or load-serving entity.
Why it works: The upfront tax benefit lowers capital cost. The offtake de-risks merchant exposure, which reduces the cost of capital for the remainder.
How to do it: Standardize warranty terms, telemetry obligations, and minimum performance definitions across installer networks. Aggregate at the contract level to avoid bespoke due diligence on every customer.

Insurance for performance and extreme events

What it is: Parametric or indemnity cover that pays out when the fleet misses contracted performance due to defined triggers.
Why it works: Lenders and rated buyers like predictable downside.
How to do it: Use operator-settled data as the reference. Structure exclusions narrowly to match the actual operational risks you cannot control, like a telecom outage affecting a whole neighborhood.

These structures are already being applied to consumer battery programs, commercial load control portfolios, and mixed-fleet aggregations. The differentiator in 2025 is the growing pool of audited telemetry and settlement history that turns a model into a covenant.

Priority markets and sequencing through 2028

2025 to early 2026: Lead with Texas and New York. In Texas, target feeders with frequent scarcity pricing and partner with retailers and cooperatives that already run demand response. In New York, focus on downstate zones with high price volatility and building electrification trends that create flexible load from heat pumps and hot water systems.
Mid 2026 to 2027: Expand New York capacity participation as rules mature. Use two years of operational data to pre-structure multi-year capacity obligations with conservative derates. In parallel, keep enrolling devices in Texas as pilot caps expand or transition toward a permanent market model.
2027 to 2028: Shift resources to PJM capacity opportunities. The scale is unmatched. By then, your fleet control software, telemetry, and customer contracts will be battle tested. That history shortens credit approval and lowers financing spreads.

This sequence maximizes near-term cash while preparing for the largest capacity market opening of the decade. It also diversifies weather risk and customer behavior across regions.

Playbooks for aggregators and utilities

For aggregators:

Treat enrollment like a sales pipeline with engineering checks at each stage. The number one reason portfolios underperform is not device failure. It is messy data and inconsistent customer permissions.
Standardize device cohorts by capability. Make your operations team’s life easy with pre-defined playbooks for batteries, thermostats, and commercial load controllers.
Build a forecasting culture. Daily probabilistic forecasts of participation and state of charge should drive bids. Operator penalties punish overconfidence.

For utilities and load-serving entities:

Create grid services tariffs that pay for verifiable outcomes. Buy peak reduction by hour and location, not vague participation.
Coordinate early with distribution planning. Use locational screens to direct enrollment where it defers upgrades. The avoided cost is your hidden margin.
Offer on-bill or tariffed device financing that pairs customer savings with system benefits. The combination reduces churn and improves program longevity.

For both sides:

Align incentives with real settlement. Revenue shares should reference operator-settled data, not forecasts.
Fund customer education. Households and small businesses need simple rules. Example: the fleet can use up to a set share of their battery during peak alerts, and they can set a minimum backup threshold.
Prepare for audits. Keep a clean chain from device data to market settlement. When a regulator or counterparty asks for proof, you should be able to produce a single report that reconciles everything.

Common pitfalls and how to avoid them

Baselining mistakes: Use control groups or weather-normalized models where allowed. Audit baselines monthly against actual dispatch history.
Telemetry gaps: Build alerts for stale data and automatic device quarantining. Nothing erodes trust like a megawatt that disappears at the wrong moment.
Overpromising capacity: De-rate your fleet for customer overrides, communications failures, and maintenance. Under-commit and over-deliver.
Customer fatigue: Limit the number of events per week and pay more for the ones that matter. Seasonal bonuses keep participation high at the right times.
Fragmented contracts: Consolidate installer agreements and warranty terms. Standard form contracts reduce friction and litigation risk.

What changes between 2025 and 2028

Rules harden: New York finishes implementing its distributed energy resource aggregation features across markets. PJM brings aggregations into capacity first, then energy and ancillary services on its later milestone. The rulebook becomes predictable, which lowers financing spreads.
Telemetry standardization: Operators converge on clearer validation and readiness documents. That reduces the bespoke engineering burden per device class.
Bigger buyers show up: Large retailers and vertically integrated utilities start procuring portfolio flexibility at scale with three to five year terms.
Mixed fleets become the norm: Batteries provide shape. Thermostats and commercial load control provide duration. Together they beat either product alone on both cost and reliability.

The direction is set. Grid operators want dependable flexibility that can be dispatched like a plant and verified like a meter.

A 90-day action plan

Pick one footprint and one product to start. For example, ERCOT real-time energy with ancillary services where your fleet already meets the pilot’s validation steps.
Build the control room checklist: dispatch receipt, device setpoint engine, telemetry quality alerts, and daily shadow settlement.
Close a term sheet with a conservative performance floor. Use your current telemetry record to set the floor so you can beat it.
Stand up locational screens with your utility partners. Put new enrollments where they do the most grid good and earn the most revenue.
Launch customer education. Simple, repeatable rules of engagement reduce churn and improve performance.

The bottom line

Virtual power plants are not waiting for a future grid. They are earning on today’s grid. In 2025, the rules in Texas, New York, and PJM finally line up with what devices can do. The details matter: operator procedures, telemetry, and settlement are the rails your revenues run on. Get those right, and you can finance portfolios on terms that scale. Get them wrong, and a promising fleet becomes an expensive experiment. You do not need a new breakthrough. You need disciplined operations in the markets that have already opened their doors, plus a plan to ride the milestones through 2028. The virtual plant is becoming a real business, one dispatch at a time.