Serbia’s 2026 electricity balance points to a modest annual net-export position, but the headline conceals a continued reliance on sizeable gross imports to manage seasonal shortages, plant outages, hydrological variability and hourly mismatches between production and consumption.
The official balance anticipates electricity imports of approximately 6,931GWh in 2026, around 14% below the 2025 level, alongside exports of about 7,410GWh, an increase of roughly 2%. On that basis, Serbia would finish the the year with net exports of only 479GWh.
That net figure is small relative to the volume traded in both directions. Gross imports are equivalent to approximately 23% of projected final electricity consumption of 29,972GWh, while combined imports and exports approach 14.3TWh. Serbia should therefore not be described simply as either an electricity importer or exporter. It is an increasingly active transit, trading and balancing market whose commercial position changes by hour, season and generation availability.
The distinction is financially important. Serbia may export surplus electricity during periods of high lignite, hydro, wind or solar generation and then import at substantially higher prices during cold evenings, drought periods or unplanned thermal-plant outages. A positive annual physical balance does not automatically produce a positive trading margin.
EPS remains at the centre of this exposure. The utility’s domestic portfolio is dominated by lignite-fired thermal generation and hydropower, with a growing but still smaller contribution from wind and solar. Imports become necessary when coal quality, mining operations, thermal-unit availability or hydrological conditions reduce production below domestic demand.
The planned reduction from an estimated 8.1TWh of imports in 2025 to 6.9TWh in 2026 assumes improved domestic availability and a more stable generation fleet. The margin for underperformance is limited. A single large thermal unit unavailable for an additional month can create several hundred gigawatt-hours of replacement demand, depending on its capacity and expected utilisation.
At an illustrative wholesale replacement price of €80/MWh, the planned 6,931GWh of gross imports would have a market value of around €555m. At €100/MWh, the value rises to approximately €693m; at €130/MWh, it reaches about €901m. These are indicative gross values rather than EPS’s actual procurement cost, because imports occur across different hours, products, contracts and counterparties.
The timing of imports can be more important than annual volume. Electricity purchased during low-price solar hours may cost considerably less than energy imported during winter evening peaks. Serbia’s generation structure makes it particularly exposed to the latter. Lignite plants provide baseload but have limited operational flexibility, while hydropower is valuable for peak coverage but depends on reservoir levels and inflows.
Wind can reduce imports during winter and overnight periods, when solar contributes nothing. Its higher capacity factor and seasonal profile give it a different system value from solar. Solar can reduce daytime imports during spring and summer, but concentrated expansion can also create midday surpluses followed by steep evening requirements. Treating both technologies as interchangeable additions to annual generation would therefore overstate their contribution to import reduction.
Serbia’s planned 1GW solar programme with battery storage can change the import profile, but its value will depend on dispatch strategy. Storage can shift part of the midday surplus into the evening, provide balancing services and reduce purchases during more expensive hours. It cannot eliminate multi-day renewable deficits or replace firm thermal and hydro capacity on its own.
The same principle applies to the wider renewable pipeline. Serbia has reached approximately 1,232MW of installed wind and solar capacity, compared with just over 400MW several years ago, and is targeting around 3.5GW by 2030. Additional capacity should reduce annual fossil generation and some import requirements, but the effect on gross imports will depend on grid availability, curtailment, storage and the production profile of each project.
Imports also reflect Serbia’s strong cross-border position. The country is interconnected with Hungary, Romania, Bulgaria, North Macedonia, Montenegro, Bosnia and Herzegovina, Croatia and Albania, giving traders access to several markets with different generation structures. Hungary brings exposure to Central European pricing and the HUPXreference, Romania and Bulgaria provide nuclear, hydro and renewable interaction, while Montenegro and Bosnia and Herzegovina add hydro-driven volatility.
These connections allow Serbia to import when neighbouring prices are lower, even when domestic production could technically cover demand. Some imports are therefore commercial optimisation rather than evidence of physical insecurity. Electricity can also enter Serbia and subsequently leave through another border, reinforcing its role as a regional transit zone.
Congestion determines whether that opportunity is available. Cross-border transmission capacity is finite and allocated through auctions and market procedures. During periods of regional scarcity, import capacity becomes more valuable at the same time that wholesale prices rise. The cost of securing the border and the price of energy must both be included in the delivered import cost.
The gross import requirement also has implications for Serbia’s current account and EPS’s liquidity. Electricity purchases require collateral, credit lines and working capital, particularly when transacted on exchanges or under short-term bilateral agreements. A sudden price increase can raise cash requirements before the additional cost is recovered through regulated or commercial electricity sales.
An increase of €20/MWh across the planned import volume would represent approximately €139m of additional gross procurement expenditure. The actual impact would depend on hedging, bilateral contracts, simultaneous export revenue and the proportion of imports procured at spot prices. Nevertheless, the sensitivity shows why relatively small changes in regional power prices can materially affect EPS’s cash flow.
Industrial buyers are not insulated from this exposure. Large Serbian consumers purchase through suppliers whose pricing reflects domestic production costs, SEEPEX prices, cross-border alternatives, balancing costs and risk premiums. When Serbia needs imports during expensive regional hours, the cost is ultimately transmitted through supplier contracts and future offers.
This makes load management increasingly valuable. Industrial consumers capable of moving production away from peak hours can reduce their exposure to the most expensive import periods. Behind-the-meter solar can lower daytime grid purchases, while battery storage can shave peaks and improve scheduling. For continuous-process industries, long-term PPAs with wind or hybrid wind-solar portfolios may provide a better match than solar-only procurement.
CBAM adds a separate commercial consideration for electricity exports to the European Union. From 2026, the definitive EU carbon-border regime changes the economics and documentation requirements for Serbian electricity sold into the EU. The effect is not simply a border charge applied uniformly to every megawatt-hour; it depends on the applicable emissions methodology, importer obligations and the ability to substantiate the characteristics of the supplied electricity.
This may alter trading incentives. Exporting lower-carbon electricity while importing higher-priced or more carbon-intensive power during other hours could preserve an annual net-export position but weaken the economic and carbon quality of Serbia’s domestic balance. Hourly production, import and export data will therefore become more important than annual totals.
The main risk to the 2026 import plan is domestic generation underperformance. Poor lignite availability, delayed overhauls, thermal-unit failures or weak hydrology could push imports above 6,931GWh. A simultaneous regional scarcity event would magnify the financial effect because replacement volumes would be purchased when prices and cross-border capacity values are elevated.
A favourable hydrological year, stronger thermal availability and faster renewable commissioning could produce the opposite outcome. Imports might fall below plan while exports rise, improving EPS’s cash generation. Yet even under that scenario, gross imports would remain commercially rational during lower-priced hours.
Serbia’s expected 479GWh net-export balance is therefore too narrow to serve as the main measure of security or profitability. The more revealing indicators are the 6.9TWh gross import requirement, the hourly spread between import and export prices, the availability of EPS’s generation fleet, hydro reservoir conditions, cross-border capacity costs and the share of purchases covered before the day-ahead market.
Electricity imports are no longer an exceptional emergency measure for Serbia. They form part of the normal operating model of a regional power system that is becoming more interconnected, more renewable and more price-sensitive. The financial objective is not necessarily to eliminate imports, but to prevent involuntary purchases during the region’s most expensive hours while preserving the flexibility to import when neighbouring electricity is economically attractive.
