Lithuania Needs 7.13 GW of Flexibility by 2035 – and Power-to-X Is the Piece the TSO might not be Counting On
Litgrid's first flexibility needs assessment under the EU market design reform says Lithuania's flexible capacity must grow from 4.36 GW to 7.13 GW by 2035, with batteries doing most of the lifting and gas plants still holding the long-duration slot. We look at why Power-to-X was scaled down in the 'realistic' case, what the Baltic-wide reserve numbers mean for PtX developers in all three countries, and why the barriers questionnaire closing on 8 September matters.
NEWS
PtXBaltic
9/3/20267 min read
While still challenged by high costs and low availability, green hydrogen and the wider Power-to-X family are increasingly talked about as a flexibility resource for renewable-heavy grids. Lithuania has just produced the first hard numbers that test that claim. In late July, Litgrid published its first national flexibility needs assessment (FNA) under Article 19 of Regulation (EU) 2024/1747, and the headline is simple: the country's need for flexible capacity climbs from 4.36 GW in 2028 to 7.13 GW in 2035, roughly 3 GW of new flexible assets in seven years. The detail matters more for Baltic hydrogen ecosystem stakeholders: the TSO explicitly reduced how much Power-to-X it counts on, and it still hands the long-duration slot to gas.
The numbers: 4.36 GW today, 7.13 GW by 2035, and batteries doing most of the work
Litgrid modelled two scenarios – a National Strategy scenario (NacSc), built on the state's existing planning documents and letters of intent, and a Slower Transformation scenario (LTrSc), calibrated to the 2025 market reality and aligned with ENTSO-E's TYNDP projections. Each ran in a base mode (every tool available) and a "realistic" mode (some held back); the realistic LTrSc run is the decision case. In that run renewables grow from 6.66 GW to 9.47 GW between 2028 and 2035, interconnector capacity available for flexibility sits at only 0.86–1.53 GW, and total flexible capacity rises to 7.13 GW: 1.01 GW of pumped storage (Kruonis), 2.65 GW of batteries, 0.68 GW of Power-to-X, 1.53 GW of cross-border capacity and 0.24 GW of gas. On top of that stack, the model finds a further 0.87–1.04 GW of "additional measures" it can't yet name, with 1.2–2.32 GWh of storage capacity behind it.
What is Litgrid's answer to the unnamed gap? More batteries. The battery (BEKS) requirement grows from 2.23 GW in 2028 to 3.68 GW in 2035, and the TSO's own conclusion is that current developer connection indications (3.12 GW) and letters of intent (4.76 GW) more than cover it. Across the sensitivity runs the 2030 battery need spans 1.11–3.16 GW, and the FNA's first conclusion says why: flexibility needs depend directly on renewable volumes and interconnector access.
Power-to-X was scaled down on purpose, and the reason is availability, not physics
Here's the line that Baltic hydrogen ecosystem stakeholders should read twice. In the notes to the scenario table, Litgrid states that for a realistic assessment of results, "the use of interconnection and P2X technologies for flexibility is reduced." The ACER methodology treats electrolysers as dispatchable assets, defined by power, energy-to-power ratio, availability and efficiency – so the physics isn't in doubt. The TSO simply doesn't believe enough of them will be built and run flexibly by 2030 to lean on. So the "realistic" case carries only 0.68 GW of P2X in 2035, and the difference shows up as the pink "additional measures" bar that batteries are expected to fill.
That is a fair judgement of the market as it stands in 2025, and it is also a challenge. Every megawatt of electrolysis that can demonstrate reliable, contractable flexibility – with real turn-down ranges, ramp rates and availability data rather than nameplate capacity – moves a project from the "not counted" column into the "counted" one. Most announced Baltic electrolyser projects still describe themselves as steady-state hydrogen producers, not grid assets. The FNA is effectively saying: show us the flexibility and we'll plan around it – and with a two-year update cycle, the 2028 edition will reflect whatever the sector actually delivers.
The long-duration slot still belongs to gas – for now
Buried in the conclusions is a small number with a big meaning. Litgrid identifies a long-duration flexibility need – events lasting more than four hours – of 20 to 260 MW, and states this can be covered by the existing flexible gas-fired plants. The storage requirement of 1.2–2.32 GWh against roughly 1 GW of additional power implies a 1.5–2.5 hour profile for the marginal battery, so anything longer than an afternoon is, in the TSO's current thinking, a job for gas turbines.
This is precisely the slot where Power-to-X has a structural argument. An electrolyser paired with hydrogen storage decouples energy from power: the tank sets the duration, not the cell stack. Whether the hydrogen goes to industry, transport or a hydrogen-ready turbine, the service is the same: a large controllable load that sits off for a windless week and runs flat out through a windy weekend. The volumes Litgrid names – 260 MW in 2028, 240 MW in 2035 – are a first-wave industrial electrolyser size, not a gigawatt ambition. I'd flag that long-duration hydrogen flexibility is still unproven economically in the Baltics, and existing gas plants will always beat assets that still need building. But "existing gas can cover it" is a statement about 2028, not about the fleet's remaining life, and the regulation explicitly asks Member States to set indicative objectives for non-fossil flexibility.
Reserves are growing across the whole Baltic LFC block, not just in Lithuania
Since the February 2025 synchronisation with Continental Europe, the three Baltic TSOs run a single load-frequency control (LFC) block and procure reserves together. Litgrid's FNA therefore includes a set of Baltic-wide numbers that matter as much in Riga and Tallinn as they do in Vilnius. Frequency containment reserve (FCR) demand rises from 31 MW to 48 MW across the Baltics between 2028 and 2035, of which Lithuania's share grows from 14 to 25 MW. Manual frequency restoration reserve (mFRR) downward capacity for the block increases from 777 MW to 974 MW, upward from 636 to 744 MW; Lithuania's own mFRR-down need moves from 632 to 773 MW while its upward need stays flat at 633 MW. Automatic FRR is projected flat at 112 MW down and 106 MW up for the block.
Read those numbers from an electrolyser operator's chair. Downward reserves – the ability to absorb energy when the system has too much – are the fastest-growing product, and a flexible electrolyser is fundamentally a downward-flexibility machine: it consumes when told to. Because reserves are procured at block level, a Latvian or Estonian asset can serve a Lithuanian mFRR-down shortfall from across the border. Litgrid also lists two Lithuania-specific products: a fast active-power response service (GAGAP) of up to 154 MW for secure islanded operation, and an LT–PL capacity-enhancement service of up to 146 MW to lift usable LitPol Link capacity to its 500 MW potential. Both pay for controllability rather than energy – the revenue stack that turns a marginal PtX case into a fundable one.
A Baltic first under Article 19, with a two-year update clock
This isn't a one-off study. Article 19 of Regulation (EU) 2024/1747 requires every Member State to assess its flexibility needs at national level, using a common methodology that ACER approved on 25 July 2025 after ENTSO-E and the EU DSO Entity drafted it. Lithuania's report – approved by the national regulator VERT and, according to the presentation, due to ACER by 25 July 2026 – covers 2028–2035 and follows the ACER structure: renewable integration needs, ramping needs, short-term needs, plus DSO and TSO grid needs. Artelys supported the modelling alongside the national resource adequacy assessment and describes Lithuania's needs as "predominantly short-term".
Why does that matter regionally? Because Latvia and Estonia are on the same clock. AST and Elering face the same obligation, reserves are dimensioned at block level, and ACER will aggregate the national reports into the basis for indicative non-fossil flexibility objectives – demand response, storage and Power-to-X. Lithuania's numbers are the first Baltic datapoint. If the Latvian and Estonian assessments also default to "gas covers long duration, P2X reduced for realism", the region will have written hydrogen out of its flexibility planning for the rest of the decade almost by accident.
The barriers questionnaire is open until 8 September, and the market plan follows in November
The final slide of Litgrid's presentation is the one with a deadline. Litgrid and the distribution operator ESO are drafting a flexibility market development plan, with a public consultation promised for November, and ahead of it they are running a confidential questionnaire on the regulatory, technical, financial and other barriers that keep market participants from offering flexibility services. Responses are due by 8 September, the survey takes 15–20 minutes, and "flexibility services" is defined as services bought by the TSO or DSO to manage their networks – the definition that decides which assets get paid.
This is the point where Baltic hydrogen ecosystem stakeholders can actually change a number in the next edition. The reasons electrolysers are "reduced for realism" are exactly the barriers the questionnaire is asking about: prequalification rules written for generators rather than loads, minimum bid sizes, baseline methodologies for demand-side assets, grid-tariff structures that penalise flexible consumption, and the absence of long-term contracts for long-duration services. If PtX developers don't describe those barriers now, the November plan will be designed around the assets that did respond – judging by the connection queue, batteries.
What this signals for the Baltic PtX pipeline
Litgrid's assessment is careful, transparent and reassuring on its own terms: the battery queue covers the gap. But it also shows where hydrogen's value would be if the sector can prove it – in the 20–260 MW long-duration slot currently assigned to gas, in the fast-growing downward reserve products across the Baltic LFC block, and in the 0.87–1.04 GW of "additional measures" the TSO expects to fill with 2-hour batteries. That matters for PtXBaltic because the biennial FNA cycle is now the scoreboard for whether Power-to-X counts as Baltic grid flexibility or stays a footnote. Three practical takeaways: answer the barriers questionnaire before 8 September; make sure any electrolyser project in the pipeline can state its flexibility parameters in the TSO's language (turn-down range, ramp rate, availability, energy-to-power ratio); and watch for the Latvian and Estonian assessments, because the block-level numbers are where the real market will be dimensioned. These are Litgrid's own figures, not independently audited, and they shift with every scenario – a direction of travel, not a procurement target.