Serbia’s quantified exporter green-electricity gap of 0.4–1.4 TWh per year is best treated as a build programme with a proof layer, not as a policy slogan. The number matters because it represents the volume of electricity that CBAM-exposed exporters would need to cover with traceable renewable attributes—in practice, Guarantees of Origin that can be assigned and cancelled against their consumption—beyond what can realistically be allocated to them from the currently auctioned renewable pipeline once you account for competing demand, supplier pooling, and the “residual-mix default” that applies whenever attributes are missing. Serbia’s own residual-mix disclosure for 2024 makes the penalty of “not proving it” explicit: the corrected residual mix is still dominated by brown coal and lignite at 66.60%, with hydropower at 23.81%, natural gas at 5.02%, wind at 0.97%, and solar at 0.36%. That residual baseline is the reason the gap cannot be closed by aspiration alone; it can only be closed by additional projects whose output is contractually and administratively earmarked for exporters, with attributes that survive audit and counterparty scrutiny.
The bridge between “gap in TWh” and “projects needed” is simply annual energy yield per installed megawatt, which depends on capacity factor. To keep the translation conservative and bank-case realistic, the conversion should use net capacity factor ranges that are plausible for Serbia’s operating conditions rather than optimistic developer cases. A conservative onshore wind range of 30–35% implies that 1 MW of wind produces about 2.63–3.07 GWh per year. A conservative utility solar range of 15–18% implies that 1 MW of solar produces about 1.31–1.58 GWh per year. These two yields are the engine of every project-count calculation. Once you accept them, the question becomes mechanical: how many megawatts must be built so that annual renewable MWh equals the missing 400–1,400 GWh?
If Serbia attempted to close the entire gap using wind alone, the math is unforgiving but manageable because wind yields more annual energy per MW than solar. Closing the low end of the gap, 0.4 TWh per year, means delivering 400 GWh annually. Dividing 400 GWh by 2.63–3.07 GWh per MW per year implies an additional dedicated wind capacity requirement of roughly 130–152 MW. Closing the high end of the gap, 1.4 TWh per year, means delivering 1,400 GWh annually. Dividing 1,400 GWh by the same yield range implies roughly 456–532 MW of additional dedicated wind. Those numbers become intuitive when expressed as project units. If a standard Serbian greenfield wind project is approximated as a 150 MW wind park, then the low gap closes with roughly one such park under conservative mid-case output, while the high gap closes with roughly three to four such parks depending on whether you want a buffer against low-wind years, curtailment, and commissioning slippage. The wind-only build envelope therefore maps cleanly to approximately 1–4 wind parks of 150 MW each, explicitly earmarked for exporter PPAs and GO allocation, not left to dissolve into general supply portfolios.
If Serbia attempted to close the gap using solar alone, the gap becomes a larger MW problem because solar’s annual yield per MW is lower. Closing 400 GWh annually at 1.31–1.58 GWh per MW per year implies roughly 253–305 MW of solar. Closing 1,400 GWh annually implies roughly 886–1,069 MW of solar. When expressed as 100 MW project units, that translates to about three 100 MW solar parks at the low end and about nine to eleven 100 MW solar parks at the high end. Solar-only can close the gap, but it does so by multiplying the number of projects, grid connection points, permitting sequences, and performance management variables that must all go right. It also increases the exposure to intraday profile risk and price cannibalisation unless projects are paired with firming or structured through contracts that socialise profile risk.
In practice, Serbia’s exporter-anchored solution is almost certain to be a blended portfolio rather than an ideology of wind-only or solar-only, because exporters are not merely buying annual MWh; they are buying a defensible footprint and a procurement structure that does not collapse under balancing costs, congestion, and counterparty scepticism. The purpose of blending is to reduce dependence on any single resource profile and any single corridor of the transmission system. A conservative planning split that tends to work in coal-heavy systems seeking fast exporter decarbonisation is to target roughly 60% of annual gap energy from wind and 40% from solar. This split is not a claim about what Serbia must do; it is simply a conservative hedge that uses wind’s higher yield for volume and solar’s modularity for diversification and connection optionality.
Under that 60/40 split, closing a 0.4 TWh gap means procuring roughly 240 GWh per year from wind and 160 GWh per year from solar. Translating the wind portion into capacity at 2.63–3.07 GWh per MW per year implies about 78–91 MW of wind. Translating the solar portion at 1.31–1.58 GWh per MW per year implies about 101–122 MW of solar. In project terms, a low-gap blended portfolio is essentially one mid-size wind project combined with one solar park roughly in the 100 MW class, with a buffer that allows for conservative performance and modest curtailment.
At the high end, the same 60/40 split applied to 1.4 TWh implies 840 GWh per year wind and 560 GWh per year solar. Translating those energy targets into capacity yields roughly 274–319 MW wind and 354–427 MW solar. In project language, that becomes approximately two wind parks of 150 MW each combined with approximately four solar parks of 100 MW each, again with the exact count depending on where realised performance lands within the conservative capacity factor ranges and on how much buffer Serbia wants to embed against network constraints and the inevitable project-level delays that appear in any multi-site build programme. The practical takeaway is that eliminating the exporter gap does not require a national-scale multi-gigawatt leap beyond current ambitions; it requires a targeted tranche of additional projects measured in the high hundreds of megawatts whose output is earmarked and proven for exporters.
Turning project counts into an implied CAPEX envelope is where investor logic becomes tangible. Because Serbia’s project costs vary by site, grid works, procurement cycle, and financing structure, the responsible way to quantify CAPEX is to state conservative planning ranges rather than to pretend a single “correct” number exists. For utility-scale solar, a conservative European emerging-market planning range of €0.55–€0.85 million per MW is a defensible envelope for all-in cost, inclusive of typical development and owner costs and standard connection works. For onshore wind, a conservative envelope of €1.10–€1.55 million per MW provides a similarly broad, bank-case range. These ranges are not an assertion about any one Serbian project’s EPC contract; they are planning assumptions meant to translate MW into an order-of-magnitude investment requirement.
Under these planning ranges, a 100 MW solar park implies CAPEX of €55–€85 million. A 150 MW wind park implies CAPEX of €165–€233 million. With these unit economics, the low-gap closure scenarios become immediately readable. A wind-only low-gap solution—one 150 MW wind park—sits at €165–€233 million. A solar-only low-gap solution—three 100 MW solar parks—sits at €165–€255 million. A blended low-gap portfolio around 80–90 MW wind plus 100–120 MW solar implies wind CAPEX of roughly €88–€140 million plus solar CAPEX of roughly €55–€102 million, giving a combined envelope of approximately €143–€242 million before any node-specific reinforcement premium.
At the high end, a wind-only gap closure—three to four 150 MW wind parks—implies 450–600 MW wind and therefore €495–€932 million of CAPEX under the conservative range. A solar-only closure—nine to eleven 100 MW solar parks—implies 900–1,100 MW solar and therefore €495–€935 million. A blended closure around 275–320 MW wind plus 350–430 MW solar implies wind CAPEX of roughly €303–€496 million and solar CAPEX of roughly €193–€366 million, resulting in a combined envelope of approximately €496–€862 million before reinforcement premiums. The magnitude is therefore consistent across approaches at the high end, but the system and execution risks differ: wind concentrates risk in fewer, larger nodes and corridors; solar spreads risk across more sites and connections but increases the number of moving parts and tends to require more deliberate handling of profile economics.
That raises the crucial question of connection priorities, because Serbia’s exporter gap is not only an energy quantity issue; it is a deliverability issue. Exporter-anchored renewables must connect where the grid can absorb them without making their delivered MWh unreliable through curtailment and congestion, and where the system can move incremental output toward the industrial basins that drive CBAM exposure. In Serbia’s case, the dominant exporter-driven demand centre sits in the Belgrade–Danube basin because that is where the largest single CBAM-exposed industrial load sits, and it is also the basin where supplier portfolios and industrial supply contracts are most likely to be concentrated. This is why transmission reinforcement programmes matter for exporter economics: they determine whether the marginal renewable MWh that an exporter has contracted will be deliverable in the volumes assumed in the PPA and will therefore generate the GOs needed to displace residual-mix electricity in the exporter’s footprint.
The most rational connection priority is therefore to anchor additional wind capacity in corridors whose evacuation paths are designed to serve the Belgrade and Srem load basins and to stabilise transfer from wind-rich zones into those demand centres. Serbia’s BeoGrid 2025 reinforcement programme has been described publicly as a major transmission upgrade enabling higher renewable integration and new high-voltage connections in that broader corridor. In practical exporter-PPA terms, such reinforcement acts like an industrial competitiveness asset: it reduces the probability that contracted renewable MWh become curtailed MWh, and it reduces the balancing and congestion premiums that suppliers and offtakers bake into PPA pricing when the grid is fragile.
The second connection priority is to deploy solar in a way that diversifies nodal exposure rather than stacking every exporter solution on a single wind corridor. Because solar is modular, it can often be placed where connection capacity exists or where incremental upgrades are cheaper than new long-distance evacuation lines. In a portfolio context, solar also helps exporters cover more of their annual consumption with attributes without being hostage to one resource’s interannual variability. When a steelworks, cement plant, or chemical complex is trying to reach a high share of GO-backed electricity, the difference between a single-asset PPA and a diversified portfolio is not abstract; it is the difference between consistently cancelling enough GOs each year to displace the residual mix and falling short in a low-wind year when you most need the proof layer.
The third priority is not physical at all; it is institutional. Serbia can build every megawatt implied by the gap and still fail to close it if the attribute allocation layer is not designed for exporter use. The same EMS residual-mix report that quantifies the coal-heavy default also quantifies the GO market’s scale, showing that 2024 cancellations were 2,447,795 MWh. If CBAM-exposed exporters need an incremental 0.4–1.4 TWh of GO-backed electricity beyond what is realistically allocable from existing and auctioned supply, that incremental demand will reshape the GO market. The only way it closes the exporter gap is if the additional projects’ GOs are contractually assigned to exporters and cancelled against their consumption, rather than being absorbed into generic supplier claims. That is why the phrase “earmarked for exporter PPAs/GOs” is not a rhetorical flourish; it is the key mechanism that converts renewable generation into defensible exporter footprints.
Once you frame the problem this way, Serbia’s path to closing the exporter gap looks like an infrastructure tranche with a clear commercial logic. At the low end, Serbia can eliminate the deficit with something as simple as one 150 MW wind park or three 100 MW solar parks, or a blended package around one mid-size wind project plus one ~100 MW solar park, provided the attributes are ring-fenced. At the high end, Serbia can eliminate the deficit with approximately three to four 150 MW wind parks or nine to eleven 100 MW solar parks, or a blended portfolio around two 150 MW wind parks plus four 100 MW solar parks. The implied investment envelope across these cases sits roughly between €0.15 billion and €0.9 billion under conservative CAPEX assumptions, with the key execution risk being grid deliverability and the key commercial risk being whether GO allocation is truly exporter-specific rather than diluted into a pooled market.
The strategic payoff for Serbia is that closing this gap does not require solving every process-emissions challenge in steel and cement overnight. It requires a targeted procurement and build programme that removes the electricity component from the “coal-linked by default” category and replaces it with a measurable, auditable, contractable attribute stream. In a residual mix where 66.60% remains lignite and brown coal, that shift is one of the fastest ways Serbia can protect the competitiveness of CBAM-exposed exports while deeper industrial decarbonisation capex is planned and financed.
Elevated by cbam.engineer