Post: How multi-site businesses can scale energy flexibility 

How multi-site businesses can scale energy flexibility 

Most energy flexibility programmes start the same way: one site, one asset, one revenue stream. A cold store curtails load during a peak event. A manufacturing plant runs its generator during a frequency response call. The economics work, the operations team adjusts, and the energy manager files a good report. The next question is often, can we do this across all our sites? 

The answer is yes but scaling from a single flexible site to a fully integrated Virtual Power Plant (VPP) is not just a matter of replicating what worked at site one across sites two through twenty. It requires a different approach to data, controls, governance, and market participation. Done well, it transforms a cost centre into a coordinated, revenue-generating energy asset. Done poorly, it creates operational risk and complexity without proportionate return. 

This article sets out the practical pathway, what changes at each stage, where the value compounds, and what needs to be in place before scale becomes possible. 

What changes when you move beyond one site 

A single flexible site is relatively straightforward to manage. The asset is known, the baseline is established, the dispatch logic is site-specific, and the operations team understands how events affect the process. Multi-site flexibility introduces a different set of challenges: 

• mixed assets: site A has a 2MW battery. Site B has interruptible industrial load. Site C has an on-site generator and rooftop solar. Each behaves differently, has different response times, and carries different operational risk if dispatched incorrectly 

• market complexity: a single site typically participates in one or two services. A portfolio of assets can participate across multiple markets simultaneously, if the optimisation logic is sophisticated enough to allocate assets to their highest-value use at any given moment 

• inconsistent data: without standardised metering and telemetry, you cannot aggregate assets reliably. Latency differences between sites, varying data formats, and gaps in half-hourly data all undermine the accuracy of a consolidated flexibility position 

• distributed decision-making: at scale, individual site managers may not know when their site has been dispatched or why, which can create friction between central energy teams and site operations 

The stages of scaling 

Stage 1: standardise the foundation 

Before any asset can contribute to a VPP, it needs to be visible, measurable, and controllable from a central platform. This sounds obvious, but in practice, many multi-site industrials have: 

• sub-meters that log data locally but don’t feed into a central system 

• assets that can be controlled manually but not remotely or automatically 

• energy data spread across multiple utility accounts, EMS platforms, and spreadsheets 

The first stage of scaling is data and controls standardisation. This means: 

• metering at asset level, not just site level, so individual flexible assets can be monitored and dispatched independently 

• remote control capability via API or direct integration with the platform, enabling automated dispatch without requiring site intervention 

• a common data format across sites, so half-hourly consumption, generation, and flexibility availability can be aggregated accurately 

Stage 2: establish a baseline  

Once data is standardised, the platform can build an accurate picture of what the portfolio is actually capable of. This is the flexibility position: how much load can be shed, shifted, or increased across the portfolio at any given time, and at what operational cost. Key outputs from this stage include: 

• site-level flexibility profiles: each site’s available flexibility by hour of day, day of week, and season, accounting for production schedules, operational constraints, and asset maintenance windows 

• portfolio aggregation: total flexibility available at any point in time, with enough resolution to respond to short-notice dispatch requests 

• constraint mapping: which sites cannot be dispatched simultaneously, which assets have minimum run times, and which operational processes have hard limits on interruption 

Flexibility profiles shift with production schedules, weather, and market conditions. A platform that updates the flexibility position in real time, rather than relying on pre-agreed static profiles, captures significantly more of the available value. 

Stage 3: multi-market optimisation 

With standardised data and an accurate flexibility position, the portfolio is ready for multi-market participation. This is where the economics of scale become compelling. A single site with 1MW of flexibility might participate in one balancing service and earn a fixed availability payment. A portfolio of ten sites with 10MW of aggregate flexibility can: 

• stack revenue across multiple simultaneous markets  

• shift assets between markets based on price signals 

• participate in wholesale trading, using the portfolio’s generation and storage assets to buy cheap and sell at peak 

• offer firm capacity to the system operator, backed by the diversity and redundancy of multiple sites 

National Grid says flexibility reduces strain on the network and avoids costly upgrades, while NESO’s Demand Flexibility Service delivered more than 3,300MWh across 22 events in winter 2022-23, showing that flexibility already has proven system value at scale.  

The revenue potential at this stage is materially higher than at the single-site level as a result of the optionality that a diverse portfolio creates. An AI-powered platform trading across 70 or more data inputs can optimise that optionality in ways that manual or rule-based systems cannot. 

The revenue for getting it right 

The financial case for scaling from single-site flexibility to a VPP is significant, but it depends on getting the foundations right. 

A single flexible industrial site might generate £50,000-£150,000 per year from a single balancing service, depending on asset size and availability. A well-optimised portfolio of ten sites, participating across multiple markets with AI-driven dispatch, can generate multiples of that figure with revenue compounding as more assets are added and market participation broadens. 

The key variables are: 

• asset diversity: a mix of load, generation, and storage gives the platform more options and reduces the risk of any single asset type being unavailable when a high-value opportunity arises 

• market access: working with a partner with access to balancing, frequency, capacity, and wholesale markets rather than participation a single service captures more of the available value 

• optimisation quality: the difference between a rule-based dispatch system and an AI-powered platform can be substantial, particularly in volatile market conditions where the highest-value opportunities are short-notice and require rapid reallocation of assets 

From cost centre to asset 

By implementing a multi-site VPP, organisations can become active participants in energy markets, generating revenue from assets it already owned, contributing to grid stability, and building the operational infrastructure for long-term energy resilience. The pathway is not without complexity. But for energy and sustainability managers who have already demonstrated the model at site level, the question is no longer whether scaling is possible. It is whether the foundations are in place to do it well. 

GridBeyond’s VPP platform is designed to support owners, operators, and aggregators of distributed energy resources, providing connectivity to energy and service markets, optimisation, and trading support.  

To understand what your portfolio of sites could be worth in energy markets, contact our expert team.