Serbia’s energy system sits at a structural crossroads. It combines large legacy baseload assets, growing renewable penetration, limited flexibility, and a transmission position that increasingly exposes it to regional volatility. Stress-testing the system is therefore not an academic exercise. It is a way to identify where physical limits, financial fragility and institutional liabilities intersect, and how shocks propagate through generation, grids, markets and public balance sheets.
This analysis models Serbia’s energy system under multiple stress scenarios, focusing on technical adequacy, balancing capability, price formation and financial outcomes for utilities, lenders and the state.
Baseline structure: What is being stressed
Serbia operates a power system with installed capacity of roughly 8.3–8.8 GW, of which lignite-based thermal generation represents approximately 4.4 GW, large hydropower around 3.0 GW, gas and oil units roughly 0.5 GW, and wind, solar and biomass together still below 1.0 GW, though growing rapidly. Annual electricity demand fluctuates between 33–36 TWh, with winter peaks increasingly driven by electrified heating and summer peaks emerging due to cooling demand.
System flexibility is structurally limited. Hydropower provides most ramping and reserve capacity, while thermal plants supply inertia and firm capacity but suffer from ageing, maintenance backlog and rising outage rates. Battery storage remains negligible at system level, below 100 MWh installed.
Financially, the system is anchored by the state-owned utility Elektroprivreda Srbije, whose balance sheet carries legacy debt, social tariff obligations and exposure to fuel and carbon costs. Wholesale market prices are increasingly influenced by regional coupling, especially with Hungary, Romania and Bosnia and Herzegovina.
This is the starting point for stress.
Scenario 1: Extreme winter with low hydro and thermal underperformance
The first stress scenario assumes a cold winter similar to 2012 or 2017, with demand reaching 38–40 TWh, combined with below-average hydrology reducing hydro output by 20–25 %, and simultaneous thermal unit underperformance of 10–15 % due to outages.
Technically, this creates a firm capacity deficit of 700–1 200 MW during peak hours. Imports become the primary balancing tool. However, regional systems tend to experience correlated stress in cold spells, limiting available imports. Serbia would face repeated scarcity hours, forcing system operators to activate emergency measures, including demand curtailment for large industrial consumers.
Wholesale prices under this scenario would spike sharply. Average winter baseload prices could rise by €25–40/MWh, while peak prices would see frequent excursions above €300/MWh. For a utility supplying regulated customers, this translates directly into losses. Assuming 5–7 TWh of energy procured at elevated prices and sold at regulated tariffs, EPS could absorb financial losses of €400–600 million over a single winter season.
From a lender perspective, this scenario does not threaten immediate solvency, but it accelerates balance-sheet deterioration and increases reliance on short-term borrowing and state guarantees. Credit metrics worsen, and refinancing risk rises.
Scenario 2: Prolonged drought combined with summer heatwaves
The second scenario models a two-year drought reducing average hydro production by 30 %, combined with recurring summer heatwaves pushing peak demand above 6.5 GW due to air conditioning load.
Technically, the system loses its primary flexibility source. Thermal plants are forced into cycling operation for which many units are poorly designed, increasing forced outage rates and maintenance costs. Gas units, though limited in capacity, become critical marginal producers.
Prices under this scenario rise structurally, not just episodically. Average annual wholesale prices increase by €15–25/MWh, while volatility intensifies. Import dependence grows by 4–6 TWh per year, exposing the system to regional price contagion.
Financially, EPS faces a cumulative impact. Higher fuel costs, lower hydro margins and increased imports erode operating cash flow by €250–350 million per year. Over two years, this equates to €500–700 million in lost economic value, even before accounting for deferred maintenance and accelerated asset degradation.
The state becomes the ultimate backstop. Either tariffs rise sharply, or fiscal support is required. In practice, Serbia’s past behaviour suggests partial socialisation of losses through budget transfers or state-backed borrowing.
Scenario 3: Accelerated coal phase-down without compensating firm capacity
This scenario assumes an accelerated retirement of 1.5–2.0 GW of lignite capacity by 2030, driven by environmental constraints or technical failure, without equivalent firm replacement.
Technically, the system loses inertia and firm capacity. Even with aggressive wind and solar additions reaching 3.0–3.5 GW, adequacy during low-wind winter evenings deteriorates. Loss-of-load expectation metrics rise sharply, breaching accepted reliability standards.
Balancing requirements increase exponentially. Without batteries or pumped storage, Serbia becomes structurally dependent on imports for 10–15 % of annual demand. This dependency is highly seasonal and price-sensitive.
Financially, the system becomes more expensive even if emissions fall. Capital expenditure requirements exceed €6–8 billion over a decade for renewables, grids and flexibility. Yet operating costs remain high due to imports and backup procurement. Wholesale prices trend upward, stabilising only if large-scale storage is deployed.
For lenders, this scenario increases project finance opportunities but worsens sovereign and utility risk. Debt migrates from corporate project SPVs to public balance sheets through grid investments and residual capacity obligations.
Scenario 4: Regional shock and import collapse
In this stress case, Serbia experiences a regional shock: nuclear outages in neighbouring countries, gas supply disruption, or simultaneous weather stress across Central and South-East Europe.
Technically, cross-border imports drop by 50–70 % during critical hours. Serbia’s interconnections become congested but ineffective. Domestic resources must cover almost all demand.
The system responds with emergency measures: industrial load shedding, voltage reduction and reserve activation. The economic cost of unserved energy becomes material. Even 200–300 GWh of curtailed industrial load can translate into €150–250 million in lost GDP.
Financially, this scenario is less about utility losses and more about macroeconomic damage. Energy-intensive sectors such as metals, chemicals and construction face forced shutdowns. Tax revenues fall, while social and political pressure rises.
Scenario 5: High-RES penetration without storage
The final scenario assumes rapid renewable build-out, reaching 45–50 % of annual generation from wind and solar by the early 2030s, but with limited storage deployment.
Technically, the system becomes over-generated during windy or sunny periods and under-generated during calm, dark hours. Curtailment rates rise above 10–15 % of potential RES output. Frequency stability becomes more challenging.
Economically, renewables cannibalise themselves. Average captured prices for wind and solar fall by 20–30 % relative to baseload. Merchant projects struggle. Lenders shorten tenors and demand higher equity buffers unless storage is integrated.
Balancing costs rise for the system operator, while state utilities remain responsible for residual adequacy. The system becomes cleaner but not cheaper or more stable.
Cross-scenario conclusions: Where the system breaks first
Across all scenarios, three structural weaknesses recur. First, flexibility is the binding constraint, not energy volume. Second, financial risk concentrates in the state utility and the public sector, even when private investment grows. Third, imports cannot be relied upon during regional stress, making domestic adequacy indispensable.
From a quantitative standpoint, Serbia’s system requires at least 1.5–2.0 GW of new firm or quasi-firm flexibility by the early 2030s, equivalent to 10–15 GWh of battery storage, pumped storage, or fast gas capacity. Without this, stress scenarios translate directly into fiscal risk.
Financially, the difference between proactive investment and reactive crisis management is measured in billions. Under-investment in flexibility increases expected annual system costs by €300–500 million, while timely deployment could stabilise prices and reduce volatility premiums embedded in wholesale markets.
Final assessment
Serbia’s energy system is resilient enough to withstand isolated shocks, but fragile under compound stress. Climate variability, ageing assets and regional coupling amplify each other. The stress tests show that renewables alone do not resolve this fragility. Flexibility, market design and balance-sheet capacity do.
In the absence of decisive investment in storage, grids and firm capacity, Serbia risks a future where energy shocks are not exceptional events but recurring fiscal and economic disruptions. The technical limits are visible. The financial consequences are quantifiable. What remains is a policy choice between engineering resilience in advance or financing crises after the fact.
Elevated by virtu.energy