Data centres reshape South-East Europe’s power economics, from grid planning to bankable offtake

In South-East Europe, the biggest risk for investors may no longer be whether supply arrives on time—it may be whether demand assumptions keep pace with a new class of electricity user. Data centres and other digitally anchored loads are starting to behave like an active pricing force, tightening spreads across hours and altering what “bankable” procurement looks like for renewables and storage.

For years, the region’s investment story has centred on renewable build-outs, cross-border congestion, battery storage and market coupling. Those themes still matter, but they are increasingly incomplete without accounting for the way large digital loads concentrate consumption geographically, require high reliability and often come with commissioning timelines that pull forward grid decisions.

Global growth sets the backdrop; South-East Europe faces a different starting point

The International Energy Agency projects global data-centre electricity demand will more than double to around 945 TWh by 2030. It also expects consumption to rise about 15% per year from 2024 to 2030, with data centres accounting for 10% of EU electricity demand growth to 2030 under current policy settings.

This matters in South-East Europe because many emerging markets remain less saturated than established Western European hubs. Land costs can be lower, power availability is uneven across countries, and strategic geography is drawing cloud, telecom and low-latency applications spanning Europe, the Eastern Mediterranean and the Middle East. CBRE’s 2025 assessment highlights that power availability varies widely and that some projects are being pushed into emerging markets where demand is not yet as saturated. CBRE also links persistent volatility to rising around-the-clock electricity demand from new data centres; BloombergNEF and S&P similarly expect European data-centre demand to double by 2030.

A new regional pattern: linked developments rather than one dominant cluster

The clearest signals appear first in Greece. Data Center Dynamics reports the Greek market is expected to more than double by 2030, supported by submarine cable landings, its position between Europe, Asia and Africa, and a growing list of large-scale digital infrastructure projects. One strategically important step was a joint venture between IPTO and Serverfarm to develop hyperscale data-centre infrastructure in Greece—an arrangement that ties power-grid expertise directly into digital-load development.

The implication for system planning is straightforward: large-load customers are no longer only external consumers of grid capacity; they are becoming part of grid strategy itself.

Romania shows an even sharper move into the large-load category. In December 2025, Data Center Dynamics reported that Accelerated Infrastructure Capital, together with ClusterPower, plans an 800 MW data-centre build-out in southwestern Romania. At high utilisation rates, such a campus implies annual electricity consumption on the order of 5.5–7.0 TWh, depending on load factor and redundancy design—large enough to affect nearby generation economics, transmission reinforcement needs and storage deployment choices.

This shifts Romania beyond a role defined mainly by renewable export potential and balancing dynamics. It becomes a candidate anchor market for very large round-the-clock digital demand.

Serbia is earlier in the curve but still moving in the same direction. Domestic reporting tied to Serbia’s state technology and infrastructure agenda indicates construction of a new data centre in Niš (spelled as “Niš” in source text) is planned to begin in 2026. Serbia has also expanded sovereign compute infrastructure through operation of a second supercomputer. While these steps do not yet amount to a hyperscale commercial hub comparable in scale or positioning to Greece or Romania’s ClusterPower plan, they suggest Serbia’s digital-load story is transitioning from public-sector compute capacity toward broader infrastructure development—important in markets already dealing with connection queues, solar clustering pressures and transmission bottlenecks.

Why this changes price formation—and what it means for renewables capture

The financial significance can be underestimated if analysis focuses only on aggregate demand levels today. Traditional incremental load patterns have been shaped largely by supply-side volatility across South-East Europe—hydrology swings, gas-linked scarcity periods, cross-border constraints and renewable intermittency.

A data-centre campus behaves differently: it typically runs at high load factors, demands very high reliability, tends to grow within concentrated geographic clusters rather than evenly across national systems, and often arrives with strict commissioning schedules that force substation or transmission investment decisions sooner than conventional industrial expansion would.

This alters how prices interact with flexibility resources. South-East Europe’s price system has been shaped by renewable oversupply during some hours alongside gas-driven scarcity at others. Large digital loads add relatively inelastic consumption during off-peak and shoulder periods—raising local price floors without necessarily eliminating midday oversupply entirely. Where data-centres concentrate demand effectively absorbs surplus conditions more readily; evening-to-night demand becomes structurally firmer; imports become more valuable when local flexibility is constrained; batteries gain stronger economic support when peak-price events occur without sufficient hedging via local generation or storage.

The result may not be uniformly higher prices everywhere—but it can produce tighter spreads between near-zero surplus hours and moderate-demand hours while increasing support for peak-price events when loads aren’t fully covered through firming strategies.

This creates a new commercial layer for renewable developers. A solar project sited in a weak-demand zone may still face severe midday capture-price erosion due to oversupply exposure along southern corridors. The same asset located closer to an emerging digital-load cluster could achieve better effective capture simply because there is deeper local absorption capacity for electricity during critical hours.

A financing shift: from merchant assumptions toward credit-anchored contracting

The article points to Romania as where this evolution could become investable at scale first. An 800 MW-class build-out would be big enough to drive dedicated power strategies: nearby renewable developers would compete not only within wholesale markets shaped by HUPX/OPCOM spreads but also as suppliers within structured stacks designed for digital infrastructure requirements.

The range of contracting structures described includes virtual PPAs (and physical sleeved contracts), co-located solar plus BESS configurations, plus longer-term balancing arrangements intended to match continuous load profiles more closely than spot-only approaches can deliver.

A key mechanism here is credit anchoring: instead of relying solely on wholesale-market price assumptions (and moderate leverage) typical of older models where merchant exposure had to be justified upfront, proximity—or direct participation—in supplying digital loads can improve perceived bankability through longer-term contracted cash flows that compress merchant risk for both generation assets and storage portfolios.

The source provides indicative financing ranges tied specifically to renewable-plus-storage projects versus merchant equivalents under congested-node conditions: leverage assumptions closer to the 65–75% range with DSCR around 1.30x–1.40x, compared with leverage nearer 50–60% and DSCR requirements around 1.45x–1.60x.

The grid investment case cuts both ways: CAPEX pressure can unlock wider capacity value

Larger digital loads often force grid reinforcement—but they can also justify it economically through additional value created downstream for surrounding generation types beyond just serving one campus load profile.

The article notes that connection packages (new 110 kV or 400 kV connections), substation reinforcement, reactive power equipment and backup integration can add tens of millions of euros to effective energisation cost for a campus build-out. At the same time those upgrades can unlock broader local capacity for renewables deployment tied into electrification trends elsewhere—including industrial electrification—and help enable storage additions where they were previously constrained by network limitations.

A concrete illustration of scale—and why “premium” firm supply becomes central

A numerical example underscores why these developments matter beyond narrative framing:

• 100 MW
• 90%
• 788 GWh per year

An equivalent calculation yields roughly 788 GWh per year ; scaling up gives about 2.36 TWh per year  for a 300 MW campus at similar utilisation rates; an 800 MW regional cluster runs toward about 6.3 TWh per year  at high utilisation according to the source text.

The argument follows that such volumes are large enough—in principle—to underpin multiple gigawatts of renewable PPAs as well as several hundred megawatts of battery storage alongside major transmission reinforcement work. 

Iberian-style certainty isn’t required—but long-duration contracting is moving closer

The report describes Greece’s attractiveness as tied not only to domestic demand but also international connectivity supporting cloud architecture overlaid onto a power market already influenced by LNG-linked marginal pricing alongside large solar additions. 

This combination keeps Greek prices among the most volatile in the region—supporting battery arbitrage—but growing digital demand strengthens incentives for long-term renewable procurement paired with “24/7 matched energy” strategies aimed at delivering firmed supply rather than accepting delivery uncertainty. 

A specific example cited from Romania involves Orange Romania’s ten-year virtual PPA with Engie Romania covering < 40 GWh per year  of electricity demand. 

Demand shock feeds back into Serbia’s outlook—even before hyperscale clusters arrive at full scale

The source connects this logic directly back into Serbia’s planning horizon through two channels: rising electricity-demand expectations embedded within national energy planning through < 2030  and beyond, and global drivers highlighted by IEA projections including industrial electrification cooling demand plus expansion of data centres & AI.  .