Summary
- VivoPower is studying BESS integration at its 41.5MW Mo i Rana data centre in northern Norway.
- The company is targeting up to about US$4m in incremental annualised EBITDA from additional Nordic reserve-market participation.
- A co-located BESS could add ride-through, power-quality support, and load-step buffering while preserving the site’s full leasable AI compute capacity.
VivoPower is studying the integration of a battery energy storage system at its 41.5MW Mo i Rana data centre in northern Norway, combining AI compute capacity with the economics of Nordic reserve markets.
The company is targeting up to about US$4 million in incremental annualised EBITDA from BESS-enabled participation in additional reserve products, subject to feasibility validation, prequalification, capital availability, market conditions, and board approval.
The study covers potential access to FCR-N, FCR-D, and FFR reserve services, which VivoPower says are not economically accessible to the site’s compute load alone because of endurance, symmetry, and response-speed requirements. The proposed design is outlined in the company’s Mo i Rana BESS update.
Storage enters the compute contract
Mo i Rana sits in Norway’s NO4 bidding zone, where VivoPower says day-ahead power prices averaged about US$0.009/kWh in 2025, below southern Norway and continental European levels. Low-cost power has long been part of the Nordic data centre pitch, but the proposed BESS adds another layer: controllability.
A battery system would not simply sit beside the data centre as a merchant energy asset. VivoPower says it could unlock stacked reserve revenues while extending ride-through and power-quality resilience for AI tenants, preserving the full 41.5MW leasable capacity for compute, and adding operational optionality as the Nordic Balancing Model develops.
The tenant-facing case is grounded in the electrical behaviour of AI workloads. Training and inference clusters can move rapidly between idle and high-load states, creating steep ramps at the point of connection. A battery-backed inverter layer can absorb some of those ramps locally, present a smoother demand profile to the grid, and reduce exposure to short-duration voltage sags, transients, and reconfiguration events.
That changes the role of storage inside the data centre power chain. Traditional resilience has centred on UPS systems, backup generation, redundancy, and switching architecture. A larger co-located BESS can add grid-facing flexibility while also supporting power quality inside the operating envelope of the site. The difficult part is integration: metering, protection coordination, transformer and switchgear capacity, dispatch logic, tenant service levels, and Statnett prequalification all have to work together.
Nordic power advantage gets more technical
Northern Norway’s appeal is not only price. The region offers cooler conditions, industrial land, renewable-heavy power, and a different grid profile from the more congested data centre markets of western Europe. Those advantages do not remove the need for careful electrical design. As data centres become larger and more power-sensitive, connection agreements are increasingly shaped by how controllable and grid-compatible large loads can be.
Reserve-market participation can also complicate operational responsibility. Capacity payments depend on availability, response time, market rules, and prequalification. The battery must be available for grid services without reducing resilience for tenants or interfering with maintenance windows, service credits, or site-level redundancy assumptions. A dispatch strategy that works financially but weakens operational confidence would be a poor trade.
The feasibility study will therefore examine electrical headroom at the existing point of connection, transformer and switchgear capacity, protection coordination, metering and settlement architecture, the prequalification pathway with Statnett, and the interaction between BESS dispatch and tenant service-level agreements. Those are the right questions. The technical interface is where the revenue model either survives contact with operations or starts to erode.
Mo i Rana points towards a wider shift in data centre energy strategy. Large compute facilities are no longer passive loads waiting at the edge of the power system. They are becoming active industrial electrical assets, with batteries, demand response, power-quality controls, and grid-service revenue increasingly tied to the value of powered land.
