TYNDP 2024 Hydrogen Import Infrastructure

Hydrogen imports from outside Europe are a pivotal supply-side lever in long-term decarbonisation scenarios. PyPSA-AT represents these imports using corridor-level data from the ENTSO-E / ENTSOG TYNDP 2024 Scenarios, which provides per-corridor capacity limits, annual energy budgets, and supply costs. This replaces the upstream PyPSA-Eur default of simple uniform-cost generators.

Why Explicit Import Corridors Matter

Europe cannot decarbonise purely through domestic renewable build-out; external hydrogen supply — from North Africa, the Middle East, or Norway — is expected to play a significant role after 2030. Austria is entirely landlocked, so it receives imported hydrogen only via the evolving European H₂ backbone network. Without spatially and economically differentiated import corridors, the model cannot distinguish:

Pipeline imports (low transport loss, infrastructure-intensive) from shipping routes (liquefied H₂ or ammonia carriers, higher reconversion cost)
Low-cost bands representing the most competitive renewable-rich origin regions, which are limited in annual volume
Higher-cost bands that can deliver larger volumes but at greater expense

The TYNDP 2024 data encodes all of this in a single dataset, structured around import corridors with stepped supply bands (sections 14.2.2 and 15.1 of the TYNDP 2024 Scenarios Methodology Report).

TYNDP 2024 Data Structure

Corridors and Supply Bands

Each row in the raw source file (H2 IMPORTS GENERATORS PROPERTIES.xlsx) represents one supply band of one import corridor. A corridor connects a source node outside Europe (NODE FROM) to a European destination node (NODE TO). Multiple rows with the same origin-destination pair but different band suffixes form a stepped supply curve: cheaper tranches are limited in volume; more expensive tranches unlock additional supply.

Source column	PyPSA attribute	Unit	Meaning
`NODE FROM`	`bus0` (country code)	—	Origin country or hub (e.g. `DZ` = Algeria, `NO` = Norway, `Ammonia` = dedicated NH₃ import node)
`NODE TO`	`bus1`	—	European destination network node
`MAX CAPACITY [MW]`	`p_nom`	MW	Peak delivery power of this supply band
`OFFER PRICE [€/MWh]`	`marginal_cost`	€/MWh_H₂	Levelised cost of this supply tranche
`MAX ENERGY YEAR [GWh]`	`e_sum_max`	MWh/year	Maximum annual energy deliverable via this band
`CORRIDOR`	index + `Band`	—	Full corridor identifier; the band suffix encodes the supply tranche

The band is extracted from the last segment of the CORRIDOR string and is used both to create a unique row index and to set the PyPSA Link carrier (e.g., "H2 import pipeline", "H2 import lh2").

Scenarios

The import dataset contains entries for three TYNDP 2024 scenarios:

Scenario code	TYNDP name	Description
`GA`	Global Ambition	High international cooperation; large import volumes at competitive costs
`DE`	Distributed Energy	More self-sufficient Europe; smaller but still significant imports
`NT`	National Trends	Modest ambition; no interzonal H₂ pipeline capacities defined

Rows with Scenario == "All" are included regardless of the selected scenario and represent supply options assumed available in every pathway.

Supply Band Methodology

TYNDP 2024 does not model each import corridor as a single supply option at one price. Instead, each corridor is split into two supply bands — a low-price tranche and a high-price tranche — each with its own annual volume limit and marginal cost. Together they form a stepped supply curve: the model exhausts the cheaper band first before drawing on the more expensive one.

The band structure and prices are derived from base prices (from the European Hydrogen Backbone study, reproduced in Table 26 of the methodology report) and then adjusted by scenario and by band:

Base price (€/MWh)	North Africa	Ukraine	Norway
2030	63	78	48
2040	42	51	48
2050	42	51	48

Scenario adjustments shift the base price up or down to reflect each scenario's storyline on international cooperation and import competitiveness:

Scenario	2040 adjustment	2050 adjustment
DE (Distributed Energy)	+15 %	+15 % (same)
GA (Global Ambition)	−15 %	−20 %
NT+ (National Trends)	fixed quantities only (see below)	—

Band price modifiers are then applied symmetrically around the adjusted base price: the low-price band receives a −25 % discount; the high-price band a +25 % premium. For shipped carriers (LH2, ammonia), the low band always has a price of zero.

The volume split between bands differs by scenario:

Scenario	Low band share	High band share
DE	30 % of corridor potential	70 %
GA	50 % of corridor potential	50 %

This is reflected directly in the e_sum_max values in the Excel file: the low-band row carries 30 % (DE) or 50 % (GA) of the total annual energy budget for that corridor, and the high-band row carries the remainder.

Worked example — North Africa → Europe, DE scenario, 2040

Base price: 42 €/MWh × 1.15 = 48.3 €/MWh (adjusted)

Band	Price	Annual volume
Low	48.3 × 0.75 = 36.2 €/MWh	86.7 TWh/yr (≈ 30 %)
High	48.3 × 1.25 = 60.4 €/MWh	217.1 TWh/yr (≈ 70 %)

These match Table 28 of the methodology report exactly.

National Trends+ (NT+) fixed-quantity logic

The NT+ scenario uses a different approach for 2030 and 2035, grounded in the National Energy and Climate Plans (NECPs). Rather than priced supply bands, a fixed quantity is established for each corridor and these imports carry no price (i.e. marginal_cost = 0). The logic is:

In 2030 and 2035 the fixed quantity is 30 % of the corridor's import potential. Imports cannot exceed this predetermined share.
Exception: the Algeria → Italy corridor uses 51 %, reflecting the PCI SouthH2 Corridor (450 GWh/day, approved 2022).
Liquefied import segments are adjusted to match specific NECP figures.

For 2040 in NT+, a minimum fixed import is again set to 30 % of the total import (zero price, NECP-aligned), while the residual potential (the remaining 70 %) is available at the base price with a +25 % premium.

How this reaches PyPSA-AT: the NT+ fixed-quantity rows in the source Excel have a non-numeric value in the OFFER PRICE [€/MWh] column (the text "Fixed quantity"). clean_tyndp_h2_imports.py coerces non-numeric price values to NaN and fills them with 0.0 (lines 104–107), so these rows enter the model as zero-marginal-cost generators bounded only by e_sum_max.

Ammonia: fixed minimum import

Ammonia has a fixed minimum import quantity in every time horizon to ensure Security of Supply diversity. The fixed low-band volume is set at scenario level (e.g. DE 2040: 135 TWh/yr, GA 2040: 135 TWh/yr per Table 28). The low-price ammonia band always has marginal_cost = 0 regardless of scenario; the high-price ammonia band carries a positive price (e.g. DE 2040: 104 €/MWh, GA 2050: 69 €/MWh).

OFFER QUANTITY [MW] is not currently used

The source Excel also contains an OFFER QUANTITY [MW] column representing a sustainable (non-peak) delivery rate, distinct from MAX CAPACITY [MW]. The cleaning script preserves this as offer_quantity in the prepped CSV, but prepare_sector_network.py:add_import_options does not pass it to n.add(). Only p_nom (max capacity), marginal_cost, and e_sum_max are attached to the network. The offer_quantity field is available for future use.

Ammonia as a Hydrogen Carrier

The Ammonia source node represents imports via an ammonia shipping route. Ammonia (NH₃) is a dense liquid carrier for hydrogen that can be transported by sea without cryogenic infrastructure. At the European port of entry, ammonia is catalytically cracked back into H₂. The TYNDP dataset treats this as an "H₂ import" corridor with higher conversion cost embedded in the offer price. The source coordinates are placed near the Faroe Islands (an ENTSO-E convention for the ammonia hub) in clean_tyndp_h2_imports.py:match_centroids.

Data Processing Pipeline

Four sequential steps transform the raw TYNDP Excel file into network components:

data/tyndp/Hydrogen/H2 IMPORTS GENERATORS PROPERTIES.xlsx
        │
        ▼  clean_tyndp_h2_imports  (scripts/open-tyndp/clean_tyndp_h2_imports.py)
resources/h2_import_potentials_prepped.csv
        │   • rename columns to PyPSA conventions
        │   • convert e_sum_max from GWh → MWh
        │   • coerce marginal_cost to numeric (NaN → 0)
        │   • extract Band from corridor string
        │   • snap source node to country centroid coordinates
        │
        ▼  build_tyndp_h2_imports  (scripts/open-tyndp/build_tyndp_h2_imports.py)
resources/h2_import_potentials_{planning_horizons}.csv
        │   • filter by planning year (wildcards.planning_horizons)
        │   • filter by scenario (Scenario == 'All' or selected scenario)
        │
        ▼  modify_prenetwork  (mods/network/h2.py:add_h2_imports)
Network components attached to n

Step 1 — clean_tyndp_h2_imports Reads the raw Excel file and produces a scenario- and year-agnostic cleaned CSV. Country centroid coordinates are resolved from data/countries_centroids.geojson and assigned to each source node so that import buses have valid geographic positions. The rule requires data/countries_centroids.geojson (retrieved by retrieve_countries_centroids from a public GeoJSON of world country centroids).

Step 2 — build_tyndp_h2_imports Filters the prepped CSV to the specific planning year and TYNDP scenario. The output is resources/h2_import_potentials_{planning_horizons}.csv — one file per planning horizon, fed directly into prepare_sector_network as a conditional Snakemake input.

Step 3 — add_h2_imports Reads the filtered CSV and attaches one Generator per corridor band directly to the domestic H₂ bus. See the PyPSA topology section below.

PyPSA Network Topology

Each supply band is represented as a single Generator:

[import Generator]─►[domestic H2 bus]
  carrier="import H2"
  p_nom (MW)
  e_sum_max (MWh/yr)
  marginal_cost (€/MWh)
  p_nom_extendable=False

Import Generator One Generator per corridor band, attached directly to the domestic H₂ bus of the destination country. Key properties:

bus: "{bus1} H2" (or "{bus1} H2 Z2" when sector.h2_zones_tyndp=True).
p_nom_extendable=False: import capacity is fixed at the TYNDP value, not a decision variable for the optimiser.
p_nom: maximum delivery power from the TYNDP MAX CAPACITY column.
e_sum_max: annual energy cap from the TYNDP MAX ENERGY YEAR column (converted to MWh). This prevents the model from over-using a low-cost band in every hour of the year — the budget constraint is the primary mechanism enforcing supply band logic.
marginal_cost: the TYNDP offer price in €/MWh_H₂.
carrier: "import H2" for all bands.

!!! note experimental, untested: "H2 bus zone suffix" When sector.h2_zones_tyndp=True, the model uses a two-zone H₂ topology (upstream TYNDP convention: Zone 1 for domestic production, Zone 2 for long-distance transport). Import generators connect to the Zone 2 bus (H2 Z2). When the flag is False (default in PyPSA-AT), all H₂ flows on a single H2 bus.

Fallback Mode

When sector.h2_topology_tyndp=False, the TYNDP corridor data is ignored entirely. Instead, H₂ import generators are placed at every gas pipeline node (using gas_input_nodes["pipeline"]), with a uniform marginal cost from sector.imports.price.H2 and no annual energy cap. This is the upstream PyPSA-Eur default — it provides H₂ supply flexibility but gives no information about where imports originate or their infrastructure cost structure.

# Fallback path (prepare_sector_network.py lines 6596–6607)
p_nom = gas_input_nodes["pipeline"].dropna()
p_nom.rename(lambda x: x + " H2", inplace=True)
n.add("Generator", p_nom.index, suffix=" import",
      bus=p_nom.index, carrier="import H2",
      p_nom=p_nom, marginal_cost=import_options["H2"])

Configuration

# config/config.at.yaml
sector:
  imports:
    enable: true
    carriers_tyndp:
      - H2                   # include H2 in the import options
  h2_topology_tyndp: true    # use TYNDP corridor data; false = fallback mode

h2_topology_tyndp is the only switch. When it is true, the rule build_tyndp_h2_imports is included in the DAG and resources/h2_import_potentials_{planning_horizons}.csv is passed to modify_prenetwork_at as a Snakemake input (see rules/pypsa-at/modify.smk). The Snakemake branch() helper passes an empty list when the flag is false, so the fallback path is taken automatically.

The TYNDP scenario used for H₂ imports is controlled by the scenario selection in build_tyndp_h2_imports.py. The AT default is DE (Distributed Energy), consistent with the base scenario lineage from PyPSA-DE.