Germany N-12.1 (2023)

Germany — N-12.1 Offshore Wind Concession Auction (2023)

1. At a glance

Sources (3 primary): bnetza-dynamic-bidding-announcement, bnetza-2023-results-press-release, totalenergies-press-release. Plus BSH Site Development Plan 2023 (URL-referenced; parser could not convert the document).

2. What makes N-12.1 structurally different from AR6 AND Thor

This is the point of running a third pilot — Germany differs from both prior cases on axes that neither UK nor DK stressed.

The most striking schema insight: the "auction model" label has to carry more than a single value per round — Germany's 2023 round is simultaneously a "sealed bid" auction (step 1) and a "dynamic ascending" auction (step 2), and the schema needs to model both stages as first-class within a single auction record.

3. Market context — the 2022 WindSeeG amendment and the 30 GW target

Germany's offshore wind regime has been through two major reforms under the same overarching act:

• 2017 WindSeeG (original): introduced the auction-based regime for offshore wind, replacing the earlier feed-in tariff scheme. Sites identified by BSH via the FEP; BNetzA runs auctions. Produced Germany's first zero-subsidy offshore wind awards (0 €/MWh), widely celebrated at the time as evidence that offshore wind could clear the market without public support.
• 2022 WindSeeG amendment (in force mid-2022, EC state aid approved December 2022): doubled Germany's offshore wind target to 30 GW by 2030, with further targets of 40 GW by 2035 and 70 GW by 2045. Introduced the two-track model: centrally pre-investigated sites (BSH does the site work, bidders compete purely on concession fees) and non-centrally pre-investigated sites (bidders conduct their own investigations post-award). Created the legal basis for the dynamic bidding procedure.

The 2023 round was the first major deployment of the new framework. 4 sites were offered simultaneously — three 2 GW North Sea sites (N-11.1, N-12.1, N-12.2) about 120 km northwest of Helgoland, and one 1 GW Baltic site (O-2.2) about 25 km from Rügen — totalling 7 GW. All four were non-centrally pre-investigated, so the winners would inherit responsibility for site investigation work.

At the same time as Germany was running this auction, the rest of Europe was learning hard lessons about offshore wind cost inflation. The UK's AR5 had just failed (September 2023, two months after the German award) with zero offshore wind bids. Denmark's Thor had been won at 0.01 øre/kWh two years earlier (December 2021) at the peak of the 2021 energy crisis. The German result — 4 sites cleared at €12.6 bn combined concession fees — was therefore a significant market signal that went the opposite way: far from needing subsidy, offshore wind developers were willing to pay the state billions for the right to build. This framing, led publicly by BNetzA President Klaus Müller, became the argument for the German model over the CfD-based approaches used in the UK and Denmark.

With retrospective hindsight from the 2024 and 2025 rounds, the 2023 result looks more nuanced. The 2024 central round secured only 2.5 GW at ~€3 bn — reflecting both smaller volumes and softer bidder interest as the oil-major bidding war of 2023 cooled. RWE famously withdrew from the 2024 round citing insufficient returns. The 2023 "€12.6 bn bonanza" may turn out to have been a one-off spike driven by BP and TotalEnergies' strategic push to enter European offshore wind at scale, rather than a durable willingness-to-pay level. This is a live story as of April 2026.

4. Timeline

Full structured chronology: ../source-docs/de_n12_1/timeline.yaml. Key dates for N-12.1 specifically:

Elapsed time summary:

• Site development plan → sealed bid deadline: ~4.5 months
• Sealed bid deadline → dynamic bidding start: ~2 weeks
• Dynamic bidding start → results published: ~4 weeks
• Total from FEP publication to winner: ~6 months
• Total from winner announcement to planned commissioning: ~7 years

Very different pace from both UK AR6 (~10 months from Core Parameters to results) and Thor (~3.5 years from Energy Agreement to winner).

5. The auction mechanism — two stages, different rules

5.1 Stage 1 — Sealed bid on price per kWh

Bidders submit a single sealed bid stating the price per kWh (in € cents/kWh) they would accept as subsidy for the electricity generated. The bid is formally a "subsidy request" — lower is better, zero cents is the legal floor. This stage is analogous to the first step of Thor (single sealed bid-price) except Germany uses € cents/kWh rather than DKK øre/kWh, and the German mechanic evaluates whether the bidder needs support at all rather than setting a strike price for a CfD.

If a single bidder submits the lowest bid, they win the site directly at that price and receive a CfD-style subsidy at the bid price.

If multiple bidders submit the same lowest price — which in the 2023 round was zero cents/kWh across all four sites — the sealed bid stage cannot discriminate between them. The Offshore Wind Energy Act then requires the auction to proceed to a dynamic bidding procedure, which is a different mechanism on a different competitive dimension.

In the 2023 round, every site went to dynamic bidding: 8 zero-cent bids for each of the three North Sea sites and 9 for O-2.2. Zero companies asked for subsidy across the whole 7 GW — which was the market signal Klaus Müller publicly celebrated as "good news for Germany's energy transition".

5.2 Stage 2 — Dynamic ascending auction on concession fee per MW

Once triggered, the dynamic bidding procedure changes both the competitive dimension and the auction format:

• Bid dimension changes from €/kWh (subsidy) to €/MW (concession fee) — bidders now compete on willingness to PAY for the site rather than willingness to ACCEPT subsidy.
• Format changes from sealed one-shot to online dynamic ascending rounds — bidders can see the current highest bid and raise it in successive rounds. No bid caps.
• Procedure runs online via BNetzA auction software, with multiple rounds per day. BNetzA does not publish bidder identities during the auction; bidders see only the current highest standing bid per site.
• Parallel auctions — all 4 sites in the 2023 round ran their dynamic bidding procedures in parallel, so bidders had to allocate their budgets across multiple simultaneous auctions. A bidder could participate in as many sites as they had pre-qualified for.
• Auction terminates per site when no bidder raises the bid in a round and only one bidder remains active at the highest level. That bidder wins the site at their last bid.
• No explicit maximum number of rounds — the auction continues until it resolves. In 2023 the rounds-per-site were: N-11.1 = 64, N-12.1 = 65, N-12.2 = 55, O-2.2 = 72.
• No lottery fallback — unlike Thor, the dynamic bidding procedure always produces a single winner because the ascending price eventually discriminates between any set of bidders.

5.3 Why this design

The BNetzA rationale for the two-stage design (explicit in the 15 June 2023 announcement):

Several bids with a value of zero cents per kilowatt hour were submitted for each of the sites by bidders deciding not to request funding. The Offshore Wind Energy Act provides for a dynamic bidding procedure in the event that this happens. The award is then given to the bidder willing to pay the highest amount for a site.

In other words: the sealed bid stage tests whether the market needs subsidy. If it doesn't, the auction automatically converts to a value-extraction mechanism where the state captures the economic rents that would otherwise accrue to developers.

Contrast with Thor: Denmark's auction also produced a cluster of bids at the legal floor (0.01 øre/kWh, all 1,000 MW), but Danish law resolved the tie via physical lottery rather than dynamic bidding. Germany extracts ~€12.6 bn of additional value via its dynamic bidding tiebreak; Denmark leaves that value on the table (or captures it elsewhere via the concessionaire-built grid).

Contrast with UK AR6: the UK CfD auction never has a "zero subsidy" case because the ASP ceiling and pot budget constraints always produce competitive clearing prices above zero. The UK cannot — and has not yet needed to — switch its auction to a value-extraction mode.

6. N-12.1 result and winner

6.1 The winner

North Sea OFW N12-1 GmbH & Co. KG — a special-purpose vehicle owned by TotalEnergies (via its TotalEnergies Renewables subsidiary). The SPV name "N12-1" maps directly to the site designation from the FEP.

TotalEnergies also won O-2.2 (Baltic) via a separate SPV (Baltic Sea OFW O2-2 GmbH & Co. KG) in the same round. TotalEnergies' combined commitment across both sites: 3 GW of capacity and €5.82 bn of concession fees.

6.2 The bid

€1.875 million per MW × 2,000 MW = €3.75 billion — TotalEnergies' binding final bid in the dynamic auction for N-12.1 after 65 rounds. This was the second-highest €/MW figure in the round (O-2.2 cleared at €2.07m/MW). N-11.1 cleared at €1.83m/MW and N-12.2 at €1.56m/MW, suggesting the North Sea sites were priced at a slight discount to the Baltic on a per-MW basis despite similar project economics — possibly reflecting the larger 2 GW vs 1 GW site size and therefore a lower marginal €/MW willingness-to-pay for the incremental capacity.

6.3 Payment mechanics

Per the BNetzA proceeds allocation:

• 3.125% Meeresnaturschutzkomponente = €117.19m for N-12.1 — paid to the Bundeshaushalt (federal budget) within one year (so by ~12 July 2024) per §23(1a) no. 2 (NOT 5% as my earlier draft said — the 5% figure was the §23(1) pre-amendment rule; the 2023 round used §23(1a) which applies a different split)
• 1% Fischereikomponente = €37.50m for N-12.1 — paid to the Bundeshaushalt within one year per §23(1a) no. 3 (NOT 5% as my earlier draft said)
• 5.875% Transformationskomponente = €220.31m for N-12.1 — paid to the Bundeshaushalt within one year per §23(1a) no. 4 (this recipient was entirely missing from my earlier 5/5/90 framing)
• 90% electricity cost reduction = €3.375 bn for N-12.1 — paid to TenneT (North Sea TSO) in equal annual instalments over 20 years starting when the wind farm becomes operational (~2030)
• TenneT applies these payments against grid connection charges that would otherwise be recovered from all German electricity consumers. Net effect: German grid charges are ~€169m/year lower (per N-12.1 alone) for 20 years starting 2030 because of the TotalEnergies bid.

The mechanism is unusual: TotalEnergies is effectively pre-funding a portion of offshore grid infrastructure cost recovery, even though the grid is built and owned by TenneT. The value of the site development right flows to consumers via reduced grid charges rather than to the state general budget.

6.4 Subsequent revenue for TotalEnergies

N-12.1 will legally receive a statutory Marktprämie entitlement per WindSeeG §24(1) no. 2 + EEG §19 with anzulegender Wert fixed at 0 ct/kWh per §21(7) — an instrument whose value parameter is zero. Economically this functions as pure merchant, but saying "no revenue mechanism whatsoever" overstates the legal position. TotalEnergies' economic revenue model for the project is pure merchant:

1. Spot market sales — selling generation on the Nord Pool DE/AT/LU zone (or its successor after the Germany/Austria split)
2. Power Purchase Agreements — the TotalEnergies press release explicitly states the intent to enter PPAs "with end-buyers, enabling them to reduce their carbon footprint"
3. Merchant curtailment risk — Germany has no negative-price protection for offshore wind under this framework; TotalEnergies bears the full risk of curtailment and zero/negative hours

This is the opposite end of the spectrum from UK AR6 (where CfD guarantees a 15-year floor) and Thor (where the CfD caps downside in both directions). TotalEnergies bought the right to a merchant position on a 2 GW offshore wind farm for €3.75 bn — a pure call option on German wholesale electricity prices over 25 years.

7. Post-award — what has happened

As of April 2026, the N-12.1 project is in the site investigation and permitting phase. Key milestones:

• ~July 2024 — initial €582m payment (combined N-12.1 + O-2.2) made by TotalEnergies to the federal budget.
• 2024–2026 — site investigation work, environmental studies, wind resource measurements, geophysical surveys. These would have been pre-completed by BSH in the centrally-pre-investigated track; on N-12.1 (non-central) the work is TotalEnergies' responsibility.
• 2026 (expected) — environmental permits submission.
• 2027 (target) — Final Investment Decision.
• ~2028–2029 — construction phase begins. Grid connection by TenneT coordinated alongside.
• 2030 (target) — commissioning, revenue begins, 20-year tail payments to TenneT begin.

Nothing has been cancelled or renegotiated as of this writeup. N-12.1 is the single largest German offshore wind concession bid by absolute euro value in 2023 and a flagship TotalEnergies project in Europe. Unlike Hornsea 4 in the UK (which Ørsted cancelled 9 months after AR6 award), there has been no walk-away pressure — partly because merchant revenue is not tied to a pre-set strike price that can become uneconomic, and partly because TotalEnergies is an oil major with a substantially higher hurdle for writing off €3.75 bn of capital commitment than a pure-play developer like Ørsted.

9. Schema implications — delta from AR6 and Thor

N-12.1 adds substantial new schema surface beyond what AR6 (§20) and Thor (§10) already identified. Grouped by category:

9.1 Revenue mechanism — a bigger enum than just "CfD yes/no"

New field on auctions:

revenue_mechanism: enum {
  cfd_two_way,                    # UK AR6 style
  cfd_two_way_with_caps,          # DK Thor style
  cfd_one_way,                    # older UK, early German rounds
  feed_in_tariff,                 # legacy German, pre-2017
  feed_in_premium,                # some early EU schemes
  merchant_only,                  # pure market revenue
  merchant_plus_concession_fee,   # DE N-12.1 style
  merchant_plus_quality_tender,   # DE 2024 central model style
}

UK AR6 = cfd_two_way. DK Thor = cfd_two_way_with_caps. DE N-12.1 = merchant_plus_concession_fee.

9.2 Auction mechanism as a multi-stage concept

The auction_stages structure needs to be a list, not a single value. Each stage has its own mechanic enum. For N-12.1:

auction_stages:
  - stage_number: 1
    name: "Sealed bid on price per kWh"
    mechanism: sealed_bid_single_price
    competitive_dimension: subsidy_price_per_kwh
    triggers_next_stage_if: "multiple bids at floor"
    outcome_if_triggered: single_winner
    outcome_if_not_triggered: advance_to_stage_2
  - stage_number: 2
    name: "Dynamic ascending auction on concession fee per MW"
    mechanism: dynamic_ascending_multi_round
    competitive_dimension: concession_fee_per_mw
    triggers_next_stage_if: null
    outcome_if_triggered: null
    outcome_if_not_triggered: single_winner

UK AR6's auction stages list would have a single entry for the sealed-bid pay-as-clear pot auction. Thor's would have a single entry with a lottery fallback. Germany's has two sequential stages with different mechanics.

9.3 Dynamic bidding fields

New fields that only Germany uses so far:

dynamic_bidding_format: enum {sealed_multi_round_ascending_clock_with_interim_bids, dutch_descending, anglo_dutch, hybrid, open_english_ascending}
dynamic_bidding_rounds_count: int  # actual rounds run
dynamic_bidding_platform: enum {online_software, in_person, telephonic, mixed}
dynamic_bidding_parallel_sites: int  # how many parallel auctions in one round

For N-12.1: sealed_multi_round_ascending_clock_with_interim_bids (per WindSeeG §21(4) which requires verdeckte Gebotsabgabe — sealed round submissions — with bidders informed only of the next Gebotsstufe and the number of eligible bidders, NOT of a visible standing high bid as a classic open English ascending auction would show), 65 rounds, online_software, 4 parallel sites, lottery fallback per §21(6) where equal Zwischenrunden-Gebote or empty rounds trigger BNetzA drawing lots.

9.4 Proceeds allocation — hypothecated splits

New sub-object on auctions:

proceeds_allocation: list

ProceedsAllocation = {
  recipient: enum {federal_budget, tso, state_general, lccc, developer_escrow},
  recipient_name: str,
  percentage: float,
  purpose: str,
  payment_schedule: enum {lump_sum_within_year, equal_annual_instalments, pro_rata_on_generation},
  schedule_years: int | None,
  schedule_start_event: enum {award, first_payment_due, commissioning, construction_complete},
}

For N-12.1 the list would have four entries per WindSeeG §23(1a) (the four-way split specific to 2023 auctions — NOT the 5/5/90 three-way split from pre-amendment §23(1) which my iter1 draft incorrectly listed):

1. 3.125% federal budget (Bundeshaushalt), Meeresnaturschutzkomponente, lump sum within 12 months of award per §58(1)
2. 1% federal budget, Fischereikomponente, lump sum within 12 months of award per §58(2)
3. 5.875% federal budget, Transformationskomponente, lump sum within 12 months of award per §58(4)
4. 90% TenneT (TSO), Stromkostensenkungskomponente used to reduce the Offshore-Netzumlage under the Energiefinanzierungsgesetz (NOT generic "grid connection charges"), paid in equal annual instalments over 20 years per §59(2), starting from the proof under §81(2) no. 4 (at least one turbine technically ready for operation including internal cabling — NOT at "commissioning" in the generic sense)

UK AR6 would have no prize_consideration entries (winning a CfD has no upfront fee). Thor would have the bidirectional CfD payment caps — DKK 6.5bn state cap + DKK 2.8bn concessionaire cap per §9.9 of the draft Concession Agreement — reflecting the bidirectional flow.

9.5 Award scope — what exactly the winner receives

New enum on auctions:

award_confers: enum {
  cfd_contract_only,                      # UK AR6 — you already had the DCO
  cfd_plus_seabed_lease,                  # hypothetical
  seabed_concession_with_cfd,             # DK Thor
  seabed_concession_merchant_only,        # DE N-12.1
  seabed_lease_only,                      # US BOEM (coming pilot)
  planning_rights_only,                   # DE 2023 non-central (technically)
  right_to_apply_for_planning,            # more precise for DE
}

For Germany specifically, winning the auction confers the right to apply for planning approval (and to build and operate the wind farm if approval is granted), not the planning approval itself. The BNetzA press release is explicit: "The successful bidders now have a right to a planning approval procedure for constructing and operating offshore wind installations on their site and to grid connection and the necessary capacity." This is weaker than UK AR6's award structure (which presumes the DCO is already in hand as a qualification gate) and weaker than Thor's (which comes with a signed concession agreement and an SEA-approved site).

9.6 Site investigation split

New field:

site_investigation_model: enum {
  state_pre_investigation_published,       # DE central track (2024+)
  developer_post_award,                    # DE non-central track (2023)
  developer_pre_qualification,             # UK AR6 — DCO/EIA complete before application
  hybrid_sea_pre_tender_eia_post_award,    # DK Thor
  no_site_investigation_required,          # hypothetical / very early
}

N-12.1 = developer_post_award. This is a distinctive German design choice — the 2023 non-central round explicitly asks developers to absorb the investigation risk, in exchange for which the bid price reflects a larger uncertainty discount. By 2024 Germany had pivoted to state_pre_investigation_published for at least some sites (the central track), seeking to reduce developer risk premiums.

9.7 Winners as SPVs with specific naming convention

Germany's winners were named bp OFW Management 1 GmbH, North Sea OFW N12-1 GmbH & Co. KG, etc. — SPVs whose names encode the site designation directly. The schema can accommodate this via the existing applicant_legal_form: enum (include site_specific_spv as a value) and by ensuring the applicant_organisation_id points to the SPV while ultimate_parent_organisation_id points to BP or TotalEnergies. This is already supported in principle from the AR6 writeup; N-12.1 confirms the need.

9.8 Ministry / political authority separate from awarding body

Germany's BMWK (Federal Ministry for Economic Affairs and Climate Action) is distinct from BNetzA (the awarding body). Same pattern as UK (DESNZ + NESO + LCCC) and Denmark (Ministry of Climate, Energy and Utilities + DEA). Schema already supports this via the AuctionAuthority list from AR6 §20; no new field needed.