The state of alert was declared on 31.07.2026 by decision No 15/2026 of the National Committee for Emergency Situations, published in Monitorul Oficial I No 637 of the same day, and extended on 28.08.2026 by decision No 21/2026, Monitorul Oficial I No 722. Under it, ANRE and Transelectrica were instructed to fast-track the authorisation of completed solar, wind and storage projects still waiting on operating permits. More than 400 MW cleared within days, about one third of it storage.
None of it was built during the crisis. It was already standing, commissioned and connected, held behind paperwork. The drought did not add generation to the Romanian system, it removed an administrative queue. Cernavoda Unit 1 has been in safe shutdown since 27.07.2026 and Unit 2 is running on volatile parameters, so the queue is still the only lever that moves at speed.
The precedent is now spreading. On 26.08.2026 lawmakers overrode the president's veto to simplify environmental permitting for strategic hydropower. On 09.09.2026 the two promised support programmes landed with a number attached: EUR 650 million for public institutions to install solar and batteries for self-consumption. Two days earlier the government set out a target of 10 GW of new wind and solar by 2030 with storage as the stated focus.
Deregulation under pressure is easy to start and hard to aim. The alert has now run past the month it was declared for, and on 10.09.2026 Romania was still importing around 472 MW with OPCOM at EUR 202.46/MWh. The open question for the rest of 2026 is not whether the shortcut outlives the emergency. It is whether anything else in the system can be made to move at the same speed once the emergency ends.
| Mkt | Project | MW | MWh | Status | Note |
|---|---|---|---|---|---|
| BG | Total operational BESS | 3 432 / 4 823 | n/a | Disputed | ESO official reporting at 20.05.2026 gives 3 432 MW. Capital counted 4 823 MW in June 2026 project by project. |
| BG | Lovech, Balkan Industrial Zone | 124.1 | 496.4 | Operational | Ministry of Energy calls it the largest in the EU. Co-located with 106.2 MWp PV on an 86.2 MW connection. |
| BG | IPS battery factory, Sofia | n/a | 3 000/yr | In production | Expanding to 5 GWh/yr. Domestic cell and pack supply is now a regional factor. |
| RO | Emergency fast-track approvals | 400+ | n/a | Authorised | Completed projects awaiting operating permits, cleared within days under the energy alert. Around one third storage. |
| RO | Enery Ogrezeni, Giurgiu | 543 | 1 000+ | Under construction | Paired with 761 MWp PV. Targeted for completion by end 2027. Subsidised BESS component. |
| RO | Enery standalone BESS | 400 | 400 | In development | 80 containers in 20 groups, one converter each. Grid connection via 400 kV underground cable. |
| RO | Nova Power and Gas, phase 1 | n/a | 200.6 of 802.4 | Setup authorisation | Part of a scheme including a PV park and a 220/20 kV substation. Phase 1 clearing final approval. |
| RO | Premier Energy, Iasi | 200 | 400 | Under construction | Q4 2026 to Q1 2027 COD. EUR 100 million financing. Commissioning timing now carries system value. |
| RS | 7 standalone, EMS contracts | 724 | n/a | Grid contracts | No construction start confirmed. Separately, Nocaj 1 (90 MW PV) is being paired with a 36 MWh system supplied by Pomega. |
| MK | 2026 annual plan, standalone | 637.5 | 1 495 | Cleared in plan | 19 standalone facilities cleared in the 2026 construction plan, plus 77 integrated systems at 3 466 MWh. Grant call open. |
| GR | 3 tender programmes | 900 | n/a | 0 MW connected | Awarded across 2023, 2024 and 2025 rounds. None operational. |
| GR | Merchant standalone programme | 4 700 | n/a | Open | Utility-scale standalone projects with priority connection, merchant basis, no subsidy. Where developer attention has moved. |
| XK | NECP storage target | 170 | n/a | Target only | Target set for 2031. No operational storage. Wind fleet reached four plants with Zatriq on 25.08.2026. |
| ME | No storage pipeline | n/a | n/a | None | First solar premium auction annulled on procedural and grid eligibility grounds. Combined solar and wind round pending. |
The state of alert was extended, not lifted. Declared on 31.07.2026 by CNSU decision No 15/2026, Monitorul Oficial I No 637, and extended on 28.08.2026 by decision No 21/2026, Monitorul Oficial I No 722. The extending act records Cernavoda Unit 1 in safe shutdown since 27.07.2026 and Unit 2 operating on volatile parameters. Under the alert, ANRE and Transelectrica fast-track authorisation of completed projects waiting on operating permits, hydropower is prioritised during the evening peak and all available thermal capacity runs at full output. On 09.09.2026 the two promised support programmes were launched, worth EUR 650 million combined, for public institutions installing solar and batteries for self-consumption. On 08.09.2026 the government set out a target of 10 GW of new wind and solar by 2030 with storage as the stated focus. Acting PM Bolojan argued that batteries should become mandatory for future prosumers. The grid capacity auction on 30.10.2026 remains on schedule.
Danube recovering, and the first Slovenian storage scheme lands. Hungary restarted a turbine at Paks on rising water levels and emergency works, reported 20.08.2026. Slovenia is launching its first state aid scheme for standalone BESS, announced 03.08.2026 and not yet published as a numbered act, making it the third market in the region to fund storage directly. The Energy Community Secretariat assessed the first NECP cycle on 17.08.2026 and put the transition funding requirement at EUR 151 billion.
The RESTORE commissioning deadline of 31.07.2026 has passed, with no third round announced. Round one ran as procedure BG-RRP-4.034 under investment C4.I8 of the national recovery plan and approved BGN 1 149 013 428, about EUR 587 million, for 82 projects totalling 9.71 GWh. Round two ran as BG-RRP-4.036 under C13.I2, closing for applications on 19.11.2025, and signed off BGN 228.9 million for 31 projects exceeding 4 GWh. Aurora Energy Research estimates that of roughly 5 GW of supported capacity in the pipeline, only about 1.8 GW will be operational in 2026, reaching 3 GW by 2030, and points to a shortage of contractors.
A 30% cut in end-user electricity prices announced on 08.09.2026, and a EUR 4.8 billion Social Climate Plan approved on 27.08.2026. Both are aimed at the consumer. The three-year price plan and the Social Climate Plan together do nothing for the producer side: curtailment holds at 3.3 to 3.7 TWh for 2026, around 12% of renewable output, CfD parks lost roughly half their revenue year on year in April, and 724 MW was still added in May. None of the 900 MW awarded across three storage tenders is connected.
The Federation moved closer to issuing guarantees of origin, so far as an administrative step rather than a published act. Without them, corporate PPAs and green tariff products in the market have nothing to certify against. A small regulatory step with disproportionate commercial effect. Sarajevo Energy and Climate Week runs 21 to 25 September.
The Bistrica pumped storage construction tender is set for 2027, minister Dedovic Handanovic said on 09.09.2026. Engineering design and land expropriation are under way, with a consultant sought on 03.09.2026 to complete the tender documentation. Serbia is answering its flexibility problem with pumped storage on a decade horizon while 724 MW of contracted battery capacity has still not started construction.
A draft law introduces investment screening in strategic sectors, energy included, reported 14.08.2026. Still a draft, so no promulgation number yet. A first formal screening mechanism. Relevant to any foreign-backed generation or storage entry into the market.
A note on our own numbers first. The snapshot above carries twenty-four capacity figures. I sourced all twenty-four for this issue rather than carrying them forward, and six of them turned out to be wrong by more than a factor of two. North Macedonia was carried at 0.30 GW of solar and is nearer 1.2 GW, which makes it the largest solar base in the Western Balkans rather than one of the smallest. Kosovo was carried at 0.10 GW against 21 MW at IRENA, and its hydro was three times too high. Montenegro's solar was three times too high, while Bosnia's wind was less than half the real figure. Bosnia's solar I could not resolve at all: IRENA says 212 MW at end 2024, Bankwatch says 608 MW for the same date, and the regulator counts 305.59 MW in larger plants. It is printed above as disputed because that is what it is.
Every figure now shows what it was checked against. That is the change I would ask other people in this market to make, because the errors above were not exotic. They came from carrying numbers forward without asking where they came from, which is the same habit that produces a 1 391 MW disagreement about how much storage Bulgaria has.
Two markets printed the same inversion this month. In Bulgaria the reported average peak price for August came in below the baseload average, EUR 111.15 against EUR 143.15/MWh. In Serbia on 10.09 the off-peak average, EUR 188.2, sat above the peak average of EUR 160.5, with a maximum hour at EUR 350. When the expensive hours stop being the peak hours, a solar park is no longer selling into the shape it was modelled against. That is cannibalisation showing up in the settlement data rather than in a forecast, and it is the strongest argument I have seen this year for pairing an existing park with storage rather than building another one.
The number that matters this month is not the Danube level. It is 1 391 MW. That is the gap between the 3 432 MW of battery storage ESO reported for 20.05.2026 and the 4 823 MW Capital counted in June from its own project-by-project list. Both numbers are defensible. They measure different things, one an operator register and the other a journalist's tally of what has been energised. I do not think either is wrong. I think the market has grown faster than the reporting around it, and an asset base nobody can count precisely is an asset base nobody can dispatch precisely.
The RESTORE deadline passed on 31.07.2026 and the honest answer is that we do not yet know what was commissioned. Round one carried EUR 587 million across 82 projects and 9.71 GWh. Round two added BGN 228.9 million across 31 projects and more than 4 GWh. Aurora Energy Research puts only about 1.8 GW of a roughly 5 GW supported pipeline in operation during 2026. If that estimate holds, the binding constraint in Bulgaria was never the subsidy and never the equipment. It was the number of teams able to build.
Romania proved the same point from the other direction. Over 400 MW moved from paperwork to authorised in days, not because anything was built but because somebody decided to clear the queue. In our own permitting and owner's engineer work the pattern is the same across all eight markets we cover: the technical file is rarely the delay. The delay is the administrative throughput sitting on top of it, and it is the one input no investor can price. Five weeks later Romania put EUR 650 million behind solar and batteries at public institutions. That money will meet the same queue the fast-track suspended, and there is no emergency running to suspend it a second time.
The smaller markets made the same point with their own arithmetic this month. Kosovo energised 70 MW of wind on a EUR 112 million package assembled from four institutions and an export credit insurer, in a market with no storage at all. Republika Srpska tendered 60 MW and asked the bidder to bring the financing with it. Montenegro annulled its first solar auction because the bids did not survive procedural and grid eligibility checks. Greece, with 11.5 GW of PV installed, still has zero of 900 MW of awarded storage connected. None of those four outcomes was decided by technology or resource. All four were decided by whether the institutional side of the project was ready.
What I would do with this going into 2027. Treat contractor availability, not module or cell pricing, as the scarce input in any BESS schedule signed from here. Assume commissioning slippage in merchant models rather than subsidy risk. And where the option exists, hybridising an existing amortised PV park with storage or wind beats greenfield on the only metric that is now genuinely tight, which is time to energisation. The grid connection is already there. That is worth more this year than it was last year.