Summary
- LG Energy Solution is showcasing AI data centre power products at ees Europe 2026 in Munich.
- The portfolio spans grid-connected BESS, UPS rack systems, and server-level battery backup units.
- The company is leaning on European LFP production in Poland and Battery Passport compliance as data centre power systems face closer scrutiny.
LG Energy Solution is targeting Europe’s AI data centre power market with systems that span grid-connected battery storage, UPS racks, and server-level backup battery units.
The company is using ees Europe 2026 in Munich to present a data centre power portfolio under the theme “Powering the Future of AI”. Its line-up includes JF2S DC LINK 5.0 for grid-connected storage, the JP6 UPS rack system for momentary power interruptions, and a next-generation 2170 battery backup unit for high-performance server environments.
The product range places batteries across several layers of the power chain. Grid-scale storage, UPS ride-through, and server-level backup are being presented as connected parts of the same AI infrastructure problem: denser compute requires more stable, more flexible, and more closely managed power.
The battery layer is widening
Data centre backup power has long centred on UPS systems for short ride-through and diesel generators for longer standby operation. AI infrastructure is complicating that model because higher rack densities, larger campuses, and more volatile power requirements place new stress on electrical design.
A grid-connected BESS can support demand management, power quality, and potentially flexibility strategies where large loads need to demonstrate more responsive behaviour. UPS batteries protect critical loads through short disturbances. Server-level backup units move protection closer to the workload, which may become more relevant as high-performance systems concentrate power draw inside dense racks.
LG Energy Solution is also putting local production into the European story. The JF2S DC LINK 5.0 uses LFP cells produced at the company’s Wrocław plant in Poland, and the company is emphasising Battery Passport readiness under EU battery rules. Those two points speak to the supply-chain and compliance pressures now sitting alongside performance and cost.
LFP chemistry has become more prominent in stationary storage because of its safety profile, cycle life, and cost characteristics. In data centres, chemistry choice is only the beginning. Operators still need evidence on fire strategy, spacing, cooling, monitoring, maintenance, replacement cycles, fault isolation, grid-code interaction, and insurance acceptance.
AI changes power behaviour
Dense AI systems can concentrate demand in ways that make ordinary data centre assumptions less comfortable. GPU clusters require stable electrical delivery, and power transients can become more challenging as workload scheduling, accelerator utilisation, and cooling systems interact.
That does not mean every AI data centre requires the same battery architecture. Some sites will prioritise grid-scale storage to manage connection constraints or peak demand. Others will focus on lithium-ion UPS upgrades, higher short-duration resilience, or more granular protection close to the rack. Colocation providers and hyperscalers will also differ in how much control they have over customer equipment and workload behaviour.
Battery suppliers are responding by selling power infrastructure as a layered system rather than a single cabinet or container. That is commercially attractive because AI data centres are large, power-hungry, and under pressure to demonstrate both resilience and lower emissions. It also creates a risk of oversimplification: products that perform well in isolation still have to integrate with switchgear, protection schemes, building controls, fire systems, and operating procedures.
Battery Passport readiness adds another dimension. As batteries become a more significant part of critical infrastructure, operators will need clearer data on materials, manufacturing, carbon footprint, performance, safety, and end-of-life handling. Compliance evidence may become a procurement filter rather than an ESG appendix.
Power constraints create a supplier race
European data centre growth is being shaped by electricity constraints as much as by customer demand. Grid queues, transformer lead times, substation capacity, planning conditions, and local political scrutiny are forcing developers to show how new capacity will be powered and how sites will behave during system stress.
Battery systems can support parts of that answer. They can provide ride-through, reduce generator starts for short events, smooth load behaviour, support grid services, and help manage peak demand. They cannot replace firm grid capacity for a major AI campus, and they do not remove the need for resilient electrical design or backup generation.
The best use cases will be precise. A BESS may help a constrained campus manage peaks or participate in flexibility markets. A UPS upgrade may reduce footprint and improve monitoring. Server-level backup may support equipment-level resilience where rack power density is rising. The weakest use cases will attach storage to a project without changing the underlying power risk.
LG Energy Solution’s push into AI data centre power lands in a market looking for credible alternatives to conventional backup and distribution models. The company’s challenge is not only to supply batteries, but to prove integration, safety, serviceability, compliance, and delivery at the pace expected by cloud and AI infrastructure customers.
The Munich showcase points to a broader shift in data centre power design. Batteries are moving from supporting equipment to strategic infrastructure. AI load will not be solved by one product layer, but the power stack is now extending from the grid connection to the rack itself.
