For a century, balancing the grid meant building bigger power stations. A different model has now reached scale: thousands of small, distributed resources, coordinated to act as one.
A virtual power plant (VPP) aggregates thousands of distributed energy resources — home and grid batteries, EV chargers, smart thermostats, flexible industrial loads — and coordinates them through software to behave as a single dispatchable power plant. In 2025, falling battery costs, mass EV adoption and market-rule reform pushed VPPs from pilots into mainstream grid infrastructure.
For most of the last century, keeping the electricity grid balanced meant one thing: when demand rose, you fired up another power station. Supply chased demand, and the bigger and more centralised the generation, the better.
That logic is now being inverted. The fastest-growing source of grid flexibility isn’t a power station at all. It’s a coordinated swarm of small things — batteries in garages, chargers on driveways, water heaters, air conditioners and industrial processes — orchestrated by software into something that behaves like a single, dispatchable plant. This is the virtual power plant, and in 2025 it stopped being a pilot-project curiosity and became mainstream grid infrastructure.
A virtual power plant is not a building. It’s an aggregation. It takes thousands — eventually millions — of distributed energy resources (DERs) and coordinates them so that, from the grid operator’s point of view, they look and behave like one controllable resource.
The constituent parts are already all around us: home and commercial battery systems, EV chargers, smart thermostats and HVAC, electric water heaters, and flexible commercial and industrial loads. Individually, none of these matters to grid stability. A single home battery discharging is noise. But aggregate fifty thousand of them and coordinate the timing, and you have hundreds of megawatts of flexible capacity that can be dispatched in seconds — faster than almost any conventional plant.
The power station of the future may not be a place at all. It may be a network — and the most valuable asset is no longer the generation, it’s the coordination.
Three trends converged.
The hardware reached scale. Years of falling battery costs and rapid EV adoption meant the underlying distributed assets were finally numerous enough to matter. There were simply enough batteries and chargers in enough places to aggregate into something grid-significant.
The grid needed them. As coal and gas retire and variable renewables grow, system operators face a widening need for fast, flexible capacity to manage volatility. VPPs provide exactly that — and they can be built far faster than new peaker plants or transmission.
The rules caught up. Market reforms in several regions opened wholesale and ancillary-service markets to aggregated DERs, letting VPP operators actually get paid for the flexibility they provide. Without that, the technology is a science project; with it, it’s a business.
The rise of the VPP reframes the grid from a one-way system — large plants pushing power out to passive consumers — into a two-way, coordinated network where the edge participates actively. That has consequences:
It’s tempting to see a VPP as simply a large pile of batteries. It isn’t. The difficulty — and the value — is in the orchestration. A VPP operator has to forecast both grid conditions and the behaviour of its own fleet, decide which assets to call on and when, respect the physical limits and the owners’ preferences of every device, and do all of this continuously, in real time, while meeting market commitments.
This is a genuinely hard control and software problem, sitting exactly at the intersection of energy markets, distributed systems and real-time optimisation. Getting it wrong means unhappy asset owners and missed market obligations. Getting it right means a power plant that costs nothing to build and grows every time someone plugs in a new battery.
The interesting work in VPPs isn’t the hardware — it’s the coordination layer: forecasting, optimisation, device integration and market participation, all running reliably in real time. That’s precisely the kind of full-stack energy engineering we do, spanning control logic, data systems and market strategy. If you’re building or scaling a VPP, let’s talk.
The value is in the coordination layer — forecasting, optimisation, device integration and market participation, running reliably in real time. That's the full-stack engineering we do.