Continuous optimization: the key to real VPP value

The VPP value proposition is stronger than ever. For US utilities especially, virtual power plants (VPPs) are becoming an increasingly compelling answer to a growing challenge: how to maintain reliable, affordable power as demand rises and supply and demand patterns become more difficult to manage. The ability to coordinate flexible consumer devices unlocks huge potential for balancing the grid, slashing associated costs while increasing reliability in the process.

Today, the way many VPPs are managed puts a ceiling on their value. Devices are still often dispatched during predefined, inflexible events, which keeps utilities from making full use of their capabilities – and reaping the associated rewards. The real solution for maximizing VPP value for grids, and end-customers, is ‘continuous optimization’.

Continuous optimization is more efficient than event-based demand flexibility

Demand response has historically operated according to a ‘static’, event-based model, where customers are incentivized to use less electricity over defined periods (often with contractual limits on how frequently events can take place). Outside those windows, a huge portion of VPP potential goes untapped.

Additionally, during an event, amending dispatch to reflect live grid conditions is difficult, making these programs a fairly blunt grid balancing tool. Growing demand volatility makes this approach even less effective.Modern consumer energy assets are unlocking a different set of possibilities. EVs, chargers, batteries and other flexible devices can adjust their charging and discharging continuously and automatically, allowing a VPP to operate precisely and efficiently across a greater proportion of the day. This affords utilities a more consistent source of flexibility, suited to increasingly complex demand patterns.The value within reach is substantial. Research from Pacific Northwest National Laboratory found that coordinating devices continuously reduces peaks by 9–15% compared to static events. For a region comparable in size to Texas, the approach could deliver annual net savings of $3.3–5.0 billion. (DSO+T Study, Vol. 1, Reeve et al., PNNL-32170-1, Jan 2022, page i.)

Delivering value by catering to customers’ needs

For customers, continuous optimization offers a way to benefit from flexibility without repeatedly adjusting their behavior around events. Critically, utilities can manage devices automatically while still respecting customer preferences. An EV driver, for example, can specify when charging should take place and how much charge they need by a particular time, while retaining the option to override the automated schedule. Respecting these boundaries keeps customers engaged, maintaining the trust a successful automated VPP program needs.

Octopus Energy’s Intelligent Octopus Go demonstrates the appeal of this approach. Designed for EV drivers, the next-generation UK tariff (rate), powered by Kraken, has attracted 500,000 customers, making it the country’s most popular EV tariff. Drivers enter their preferences and connect their cars overnight; their utility then takes care of managing and optimizing charging. The financial benefits are stark. In its AI in Charge study, the UK’s Centre for Net Zero found that managed charging reduced peak grid demand by 42% and annual customer electricity bills by around £343 ($450). Running successful next-gen VPP programs

The capabilities involved in running an effective and scalable VPP (customer operations, billing, grid planning and operation, for example) must be brought together, supported by shared systems and data.

Leading utilities are addressing this by giving cross-functional teams responsibility for rolling out and leveraging VPPs, with remit, authority and visibility extended across traditional departmental siloes. These teams benefit from being able to define desired outcomes – deferring a substation upgrade, for example – and how success will be defined and measured. This also provides a basis for deciding which specialists from other teams should contribute, how to encourage customer engagement, and whether separate programs should be consolidated under a single umbrella.

Connecting the data, devices and systems

Optimizing thousands of devices in real time depends on reliable, high-quality data and software capable of handling the relevant streams. The best VPPs therefore need an underlying platform that can handle and leverage this data. Intelligent Octopus Go, for instance, uses Kraken to factor in grid imbalance, wholesale energy costs and customer schedules into decisions that shift peak demand on the grid.

Cross-functional teams also need a common tech foundation. VPP teams need visibility over grids, devices, optimizations and customer engagement in order to make decisions across these traditional siloes. Disparate legacy systems, often cobbled together to talk to each other in different languages make this incredibly difficult.

Fortunately the platforms needed to get the most from VPPs are here. Kraken’s unified data model dismantles siloes and connects devices with broader utility systems, giving teams visibility and room to act effectively across departments, simplifying rollout. This, in turn, makes other parts of developing and running a VPP easier. Rate creation and customer engagement, once challenging and time-consuming tasks, can be streamlined through the same platform.

Unlocking benefits by matching continuous dispatch to local grid conditions

Static, event-based dispatch generally responds to a single common price or dispatch signal over a wide geographical area. Next gen systems, like Kraken, that enable continuous VPP optimization benefit from being able to simultaneously handle multi-channel inputs, such as locational factors, and so enable utilities to maximize benefits from more fine-grained locational pricing (empowering VPPs to capitalize on fine-tuned locational optimization, more of the time).

Consider a standard, static off-peak charging offer available to EV drivers across the entirety of a network that covers a large geographic area. During a demand peak, a lower price signal should encourage customers to charge within the same window. However, while across the system overall demand may be low, if one particular neighborhood has an especially high concentration of EVs, all those cars charging together could continue to strain local infrastructure, increasing congestion and the cost of delivering electricity.

In this case, if dispatch was continuous, and location-specific, utilities could gain massive benefits from continuously reducing charging in constrained areas while allowing it to continue where energy is abundant. This distinction matters financially. According to PG&E’s Electrification Impacts Study, through 2040, orchestration at the local, secondary-network level reduced projected distribution costs by approximately $1.8 billion. Alternatively, the savings from system-wide orchestration alone were just $150 million – and that approach also caused new local overloads. In other words, accounting for local conditions unlocked roughly twelve times the VPP value. (PG&E, Electrification Impacts Study Part 2, CPUC R.21-06-017, 2026.)

Turning greater capability into greater value

As utilities face spikier supply and demand patterns, restricting VPPs to static events, rolled out through siloed programs, increasingly limits what they can achieve. Connected devices already have the capacity to respond with the continuity, precision and efficiency these conditions require.

Fully capturing this potential requires a new operating model, and new tech to power it. Good data, cross-functional teams, and a unified platform like Kraken give utilities the foundations to make their devices work more effectively – delivering far greater benefits for grids, utilities, and their customers.