Europe now has market-based price benchmarks for battery storage. Türkiye does not. Turkish battery files written on European templates are running a risk their credit committees have not been shown.
If you are financing a Turkish battery project in 2026, the practical move is to record in the financing memo where your central forecast sits against the Pexapark Germany benchmark, and to run a scenario set rather than a point estimate on Turkish ancillary revenue. Both are free to do and neither is currently standard. A file that shows only a central case is now distinguishable from one that shows a central case plus a market reference, and the credit committee should be the one deciding whether that matters.
The reason it matters here is scale and timing. Türkiye has 568 active pre-licensed storage-integrated wind and solar projects totalling 28,592 MW, against 370.60 MW commissioned across eight plants, on EPDK data compiled at the end of July 2026. Almost all of that pipeline will be financed during the same window in which the European market has acquired its first observable price anchors for battery storage — and Türkiye has not.
Note on dates. This page was first published as a LinkedIn brief on 30 April 2026 and was re-checked against published data on 7 August 2026. The commissioned-capacity and pipeline figures have moved materially since first publication and have been restated; the conflict the original brief carried between two pipeline numbers is resolved below. Each figure carries its own date.
On 27 April 2026 Pexapark published The BESS Valuation Trap, arguing that the forecast-led valuation approach that struggled with capture-rate downside in solar and wind is now being applied to battery storage, with more complexity and a comparable risk profile.
Two market events made the question commercially live rather than theoretical.
On 16 April 2026 the European Commission approved Germany's €3.8 billion state-aid scheme for industrial electricity-price relief under the Clean Industrial Deal State Aid Framework. The scheme runs retroactively from 1 January 2026 to the end of 2028, with a €50/MWh (5 ct/kWh) floor, relief of up to 50% of the wholesale price on a maximum of 50% of a company's annual consumption, and a requirement that at least half the aid be reinvested in decarbonisation measures within 48 months. On the same date the German economics ministry said relief could extend to sectors outside the standard list, including metalworking, mineral raw materials and food processing. Applications to BAFA are submitted retroactively, with the first 2026 applications from early 2027.
In parallel, Pexapark launched market-based BESS pricing benchmarks for Germany and Spain, built on observable market activity — market consensus, polling, bid/ask and transaction evidence — drawing on 100+ counterparties and 500+ monthly data points, refreshed monthly. That is a product description as at April 2026 rather than an independently verified figure.
The consequence is narrow and specific: a forecast-derived asset valuation can now be checked against observed market pricing, not only against another forecast.
The weak version of this argument is that solar got mispriced so batteries will too. The stronger version is different and more useful.
A solar PPA is dominated by one revenue stream: a long-dated price for delivered energy, modelled against expected capture rates and merchant tails. There is essentially one forecast under the file.
A battery earns across day-ahead arbitrage, intraday, balancing energy, ancillary services, capacity mechanisms, tolling and floor structures, and optimiser routing across all of them. Diversification can in principle reduce dependence on any single forecast. That is the case for batteries.
The difficulty sits one layer down. Each revenue stream carries its own assumption set, the streams interact, and the optimiser routing assumes liquidity and price-formation behaviour that may not hold — particularly where ancillary product design is still being adjusted, balancing-market depth is shallow, or capacity-mechanism rules are politically live. The risk has not disappeared. It has moved from a single forecast to a stack of assumptions about market structure, which cannot be hedged the way a price can.
| Revenue stream | What it assumes | Audit difficulty |
|---|---|---|
| Day-ahead arbitrage | A persistent intra-day price spread of the modelled magnitude as renewable penetration rises | Medium |
| Intraday market | Intraday liquidity and bid-ask depth comparable to day-ahead — holds in Germany, varies by market | Hard |
| Balancing energy (aFRR / mFRR) | Stable activation rules and TSO procurement design, both subject to regulatory change | Hard |
| Ancillary services (FCR, voltage, black-start) | Product design, procurement volume and price formation that are still maturing in many markets | Very hard |
| Capacity mechanism | Political durability of the mechanism across the asset life | Very hard |
| Tolling / floor structures | Counterparty creditworthiness and contract enforceability | Cleanest |
| Optimiser routing | That every assumption above holds simultaneously, and that other batteries do not crowd the same trades | Very hard |
Audit difficulty rises as the assumption set leaves the contract layer and enters market-design behaviour. The optimiser routing layer compounds every assumption above it. This table is Kickoff's framing of the revenue stack, not a sourced classification.
This is the part of the original brief that has changed most, and where a six-month-old figure would have been misleading.
EPDK data compiled at the end of July 2026 shows 5,968 pre-licence applications for storage-integrated generation, totalling 260,354 MW. TEİAŞ allocated roughly 33,300 MW of capacity against those applications. After cancellations of projects that missed their obligations, 568 active projects remain, totalling 28,592 MW — 240 wind projects at 16,804 MW and 328 solar projects at 11,787 MW.
Further down the funnel: 357 projects with a combined 16,110 MW have received environmental impact decisions, 62.8% of the active pre-licensed set. Since generation licensing began in March 2025, 82 projects totalling 3,538.24 MW have received generation licences. Eight plants are in operation, with 370.60 MW of installed capacity and 405.61 MWh of storage.
The original brief carried two numbers that appeared to conflict — 33 GW from Ember and 25.6 GW from Energy-Storage.News. They are not contradictory. They are readings at different dates on different bases, and here is the sequence:
So the pipeline has not grown from 25.6 GW to 33 GW and stopped. It was allocated at around 33 GW and has since been trimmed by cancellation to 28.6 GW, while the funnel below it has started to convert.
| Stage | Value | As at |
|---|---|---|
| Pre-licence applications | 5,968 applications, 260,354 MW | 31 Jul 2026 (EPDK) |
| Capacity allocated by TEİAŞ | ~33,300 MW | 31 Jul 2026 (EPDK) |
| Active pre-licensed projects | 568 projects, 28,592 MW | 31 Jul 2026 (EPDK) |
| — of which wind | 240 projects, 16,804 MW | 31 Jul 2026 (EPDK) |
| — of which solar | 328 projects, 11,787 MW | 31 Jul 2026 (EPDK) |
| Environmental impact decisions granted | 357 projects, 16,110 MW (62.8% of active) | 31 Jul 2026 (EPDK) |
| Generation licences granted since Mar 2025 | 82 projects, 3,538.24 MW | 31 Jul 2026 (EPDK) |
| Commissioned and operating | 8 plants, 370.60 MW / 405.61 MWh | 31 Jul 2026 (EPDK) |
| Commissioned, previous published figure | 208 MW | Feb 2026 (EPDK, via Ember) |
| National storage target | 7.5 GW at 2-hour duration | 2035 (National Energy Plan) |
| Average duration, licensed and pre-licensed | 37 GWh across the set, about 1.1 hours | Ember, April 2026 |
EPİAŞ runs day-ahead and intraday markets and a balancing market exists. The issue is relative depth and liquidity against Germany and Spain — the markets European forecasting templates are calibrated on. A revenue-stack model trained on German intraday liquidity may overstate the realisable Turkish intraday spread.
Ember records the licensed and pre-licensed storage set at 37 GWh, an average duration of about 1.1 hours, because Turkish criteria for storage-integrated projects have pushed investors toward one-hour systems. The national 2035 target is 7.5 GW at two-hour duration. Global installations in 2025 averaged 2.5 hours. A one-hour asset earns a different revenue stack than a two-hour asset, and templates that do not separate the two will misprice both.
Lira hedging adds a layer of uncertainty over multi-year revenue stacks. On the capital side, Türkiye applies a 30% import duty on LFP cells, modules and systems — announced in late 2023 and applied from early 2024, on Energy-Storage.News reporting. That is directly relevant to files whose CAPEX assumptions were built on European cell-cost trajectories.
CBAM-driven offtake demand from Turkish industrials is real and largely absent from underwriting templates. Turkish exporters under CBAM pressure have a structural reason to contract for renewable-plus-storage offtake on hourly-matched terms. European market models do not yet price that demand. It runs in the project's favour, and leaving it out is a different kind of error from leaving out a risk.
Our working thesis is that a material share of the Turkish pipeline is being underwritten through European IPP templates that import forecast logic into a market without the same day-ahead, intraday and balancing-market depth as Germany or Spain. We have not pressure-tested this against named lender underwriting notes. We flag it as a thesis for Turkish IPPs and lenders to test against their own files, and we would rather say that than present it as established.
None of these are Kickoff services. They are what we would expect anyone financing Turkish battery storage in 2026 to be running, given what changed in the European market.
370.60 MW of installed capacity across eight commissioned plants, with 405.61 MWh of storage, on EPDK data compiled 31 July 2026. That is up from 208 MW as at February 2026. Against it sit 568 active pre-licensed projects totalling 28,592 MW and a national target of 7.5 GW at two-hour duration by 2035.
Neither, as at today. Those are readings from different dates. 25.6 GW was the pre-licensed colocated total reported by Energy-Storage.News around the introduction of the 30% LFP import duty. 33 GW is the capacity allocated to storage-integrated wind and solar since 2022, per Ember's Türkiye Electricity Review 2026. The current figure is 28,592 MW across 568 active projects, after cancellation of projects that missed their pre-licence obligations, on EPDK data compiled 31 July 2026.
A solar PPA is dominated by one revenue stream and therefore one forecast. A battery earns across day-ahead arbitrage, intraday, balancing, ancillary services, capacity mechanisms, tolling and optimiser routing. Each layer carries its own assumptions, the layers interact, and the optimiser assumes liquidity and price formation that may not hold. The risk moves from a single forecast to a stack of assumptions about market structure, which cannot be hedged the way a price can.
A state-aid scheme approved by the European Commission on 16 April 2026 under the Clean Industrial Deal State Aid Framework, running retroactively from 1 January 2026 to the end of 2028. It provides relief of up to 50% of the wholesale price on a maximum of 50% of a company's annual consumption, with a €50/MWh floor and a requirement to reinvest at least half the aid in decarbonisation within 48 months. It matters to storage because it changes the industrial demand side of the German power market that European BESS revenue models are calibrated against.
Not as at 7 August 2026. Pexapark publishes market-based BESS pricing benchmarks for Germany and Spain. There is no Türkiye equivalent. That is why the practical move is to reconcile a Turkish file against the German benchmark and record the gap, rather than to wait for a domestic reference that does not yet exist.
If a Turkish battery file is going to a credit committee this year, the cheapest thing you can add to it is the gap between your central case and the nearest observable market price.
Talk about the work >