Summary
- DDC1 will be an 11.5MW multi-tenant facility at the Port of Esbjerg, with commissioning targeted for October 2027.
- Plexar's energy management system is intended to adjust electricity demand around available grid capacity, with storage supporting stable operation.
- The design also incorporates closed-loop cooling and planned heat recovery through local utility DIN Forsyning.
Copenhagen Infrastructure Partners, Thylander, and PensionDanmark have started construction of an 11.5MW data centre in Esbjerg designed to operate more flexibly alongside Denmark’s electricity system.
Dansk Data Center 1, or DDC1, is being built at the Port of Esbjerg and is scheduled for commissioning in October 2027. PensionDanmark, Thylander Group, and CIP own the project, with the latter investing through its CI Microgrid Electrification Fund. Thylander will take responsibility for operations, maintenance, and asset management once the facility enters service.
The site will connect to local distribution network operator N1 with 11.5MW of capacity. Rather than treating that connection as a fixed block of demand, the project will use an energy management system from Plexar to adjust electricity consumption in response to available grid capacity, while energy storage is intended to help maintain stable operation.
Flexibility reaches the facility design
Large electricity users are increasingly being asked to fit into a power system where grid capacity varies by location and time. Data centres present a difficult version of that problem: the IT load needs a dependable supply, tenants buy availability, and cooling, networking, and electrical systems cannot simply be switched off when the wider network becomes constrained.
DDC1’s design therefore places the practical limits of demand flexibility at the centre of the engineering challenge. Moving consumption around the day may be possible for some supporting systems, batteries can temporarily reduce grid demand, and certain computing workloads can be scheduled differently, but a multi-tenant colocation facility cannot assume that customer load is freely interruptible.
The storage system will be central to how far that flexibility can extend. Neither its capacity nor its technology has been disclosed, leaving open whether it will mainly provide short-duration power support, reduce grid peaks, participate in electricity markets, or combine several roles. The usable energy capacity, discharge duration, control strategy, and interaction with the site’s UPS architecture will determine how much load can genuinely be separated from instantaneous grid demand.
Power management also has to sit behind the facility’s resilience requirements. A grid response that creates additional operational risk would undermine the service being sold to customers, so automated controls must work within constraints imposed by redundancy, battery state of charge, cooling conditions, and tenant agreements.
The same systems approach runs through the thermal design. DDC1 will use closed-loop cooling in which water is recirculated rather than routinely discharged, while surplus heat is intended to feed the district heating network operated by DIN Forsyning.
Heat recovery still needs a customer for every megawatt
Having a named utility counterpart gives the heat reuse plan more substance than a general commitment to make waste heat available. Useful recovery nevertheless depends on several pieces of infrastructure outside the data hall: heat exchangers, pumps, potentially heat pumps, distribution pipework, suitable temperature levels, and enough local heat demand to absorb the output.
The relationship between the data centre and the heat network will also change as the facility fills. A lightly loaded building produces less recoverable heat than a mature site, while district heating demand varies seasonally. The commercial and technical model has to accommodate both curves rather than assuming that supply and demand will always coincide.
Cooling water performance will need similar scrutiny. A closed circuit can keep routine water consumption low, but total resource use depends on the final heat rejection system, ambient temperatures, maintenance requirements, and the density of the computing equipment installed. Water recirculation inside the plant does not by itself establish the site’s eventual WUE.
Esbjerg provides an unusually relevant setting for the project. The port is already an established energy and industrial centre, giving DDC1 access to infrastructure and organisations accustomed to operating physical energy assets rather than developing a data centre in isolation from the systems around it.
CIP brings the same background from the investment side. Its wider portfolio spans renewable generation, storage, transmission, distribution, and other energy infrastructure, so DDC1 extends an existing energy investment model into digital capacity rather than approaching power as a procurement exercise after land and buildings have been secured.
At 11.5MW, DDC1 is modest beside the multi-hundred-megawatt campuses now attached to hyperscale and AI development plans, but that scale could make the interaction between tenants, storage, grid controls, cooling, and heat recovery easier to test without depending on a single enormous customer.
By the time commissioning begins in 2027, the useful measures will be operational rather than promotional: how much load the energy management system can move, how the batteries are used, how much heat DIN Forsyning can take, and how much water the cooling plant consumes as utilisation rises. Those figures will show whether energy integration survives contact with a live colocation workload.
