Baltic Energy Transit Risk
How exposed the new Baltic gas import terminals are
What this is
This started as a student project in international affairs. I wanted a map of the terminals the Baltic region built after Russian pipeline gas stopped, and a way to say how exposed each one looks from public data alone.
In 2021 the eight countries on that sea took about 70 bcm from Russia by pipe. By 2023 the same Eurostat series was down to almost nothing. Five working coastal LNG terminals, plus a sixth planned at Gdańsk, can land about 23 bcm a year between them, which is roughly a third of what stopped arriving. The rest of the gap closed because demand fell and because gas still comes in by pipe from Norway and the west.
So the region is not short of gas on a normal year. What I wanted to see is how concentrated those terminals are, how close they sit to Russian territory, and how much of the supply run depends on ships coming through the Danish Straits. I score each terminal on those three things. The weights are a choice I made, not a result, and you can move them on the Analysis page.
Nord Stream was blown in 2022, Balticconnector was damaged in 2023, and Estlink 2 was cut in 2024. I am mapping the geography those events keep landing on, not scoring the events themselves.
Terminal capacities are operator figures from 2024 and 2025. Import dependency is Eurostat for 2023. Everything I used is on the Data page.
The question
Before 2022 most of this gas came by pipe from Russia. After the invasion of Ukraine and the Nord Stream sabotage that route was shut or abandoned, and the region started landing chilled gas from ships at Klaipėda, Świnoujście, Inkoo and a handful of other coastal sites. That put a lot of eight countries’ import capacity onto a few points you can actually put on a map. The question I am asking is:
After 2022, Baltic energy imports shifted onto a small set of coastal LNG terminals. How concentrated is that infrastructure, and how exposed is it, if you measure by capacity, by distance to Russian territory, and by how far a cargo has to run past the Danish Straits?
Hover a point for the name, click for the figures. Size is capacity.
What I found
The numbers sit on the Analysis page. Capacities move, so I would read them as a picture of the pattern rather than as a precise measurement.
Świnoujście in Poland and Mukran in Germany hold about three fifths of the operating import capacity. The northeast runs on one or two terminals each: Inkoo for the Gulf of Finland, Klaipėda for Lithuania. Five working sites replaced a continental pipe network, with Gdańsk still planned.
The terminals closest to Russian territory are the small ones. Hamina sits at the eastern end of the Gulf of Finland; Klaipėda and Gdańsk sit next to Kaliningrad. The two largest terminals are among the farthest from Russia. That mismatch is why I did not rank on size alone.
Every cargo still comes through the Danish Straits. The Kiel Canal is the other door, and it is too small for LNG carriers, so terminals up in the Gulf of Finland have the longest run once they are inside the sea.
Finland and Lithuania built well beyond their own demand. Germany and Poland, with much larger consumption, cover only part of theirs by sea. The northeast built a buffer that neighbours draw on. That is the finding I would actually take into a conversation, because the terminals sitting closest to Russia are also the ones carrying spare capacity for the rest of the region.
How the score is built
Each terminal gets a number from three ingredients:
- how close it is to Russian territory (I gave this 0.4, the largest share),
- how much of the region’s import capacity it holds (0.3),
- how long its supply route past the Danish Straits runs (0.3).
I weighted proximity most because the question is about exposure to one neighbour, and distance is the most direct thing the public data will support. Capacity is in there because a larger terminal does more damage if it stops. Transit is in there because a site deep in the Gulf of Finland has a longer voyage to be restocked.
Those weights are a judgement. They are on the Analysis page if you want to drag them and see how much the order moves.
Disruptions since 2022
Baltic energy and data links have been damaged several times since 2022, mostly by ships dragging anchors across the seabed.
| Date | Asset | What happened |
|---|---|---|
| Sep 2022 | Nord Stream 1 and 2 | Subsea gas pipelines ruptured by underwater explosions off Bornholm. |
| Oct 2023 | Balticconnector | The Finland-Estonia gas pipeline and a telecoms cable were damaged by a ship’s anchor. The pipeline was shut for six months and returned to commercial use on 22 April 2024. |
| Dec 2024 | Estlink 2 | The Estonia-Finland power cable and four telecoms cables were cut on Christmas Day; a shadow-fleet tanker is the prime suspect. Repairs ran into 2025 and the link returned to commercial use on 20 June 2025. |
I left those events as context. The Analysis page maps the corridors and which have been hit.
Sources: Balticconnector return to service (Gasgrid), Estlink 2 return to service (Fingrid), Estlink 2 incident (Wikipedia).
What I would take from this
If you borrow the EU’s N-1 idea and knock out each country’s largest terminal, Finland is left with Hamina, covering about a fifth of demand. Lithuania, Poland and Germany are left with no seaborne capacity at all. They are not failing the real N-1 test, which also counts pipes and storage. The fallback is the same links that have already been damaged.
Estonia, Latvia and Sweden have no terminal of their own. Their gas comes through neighbours, over pipes and cables such as the Balticconnector. That layer has already been hit.
The Danish Straits are one entrance for all eight countries, so anything that watches that stretch (the kind of patrol NATO stood up after the December 2024 cable cuts) covers the whole region at once.
FSRUs were quick to bring in and can be moved, but they are few and chartered. Mukran lost about half its capacity when a second ship came off charter, which is why I count it at about 6 bcm rather than the 13.5 bcm still quoted in a lot of headlines.
Limitations
This is geographic exposure, not a probability that anything will happen. A terminal close to a maritime boundary is a fact about a map.
- The score uses distance and capacity share as proxies. Change the weights and the ranking moves.
- It is one snapshot. Capacities, routes and terminals change.
- Capacities are compiled from operator reporting and Global Energy Monitor, current to 2024 and 2025. Dependence figures are Eurostat for 2023. See Data.
- The 2023 to 2025 cable and pipeline incidents are context, not events in the model.
I would treat this as a way to see where the exposure sits, not as a forecast.
Terms
- LNG (liquefied natural gas): natural gas chilled to a liquid so it can be carried by ship rather than a pipeline.
- LNG terminal: the coastal site where those ships unload and the cargo is turned back into gas for the pipeline network. Some are built on land; some are a moored ship, known as an FSRU (a floating storage and regasification unit).
- Capacity (bcm): how much gas a terminal can bring in over a year, measured in billion cubic metres.
- Import dependency: the share of a country’s gas that comes from abroad rather than from its own production.
The rest of the site
- Data lists every source, all public, and says what is solid and what is still rough.
- Analysis has the maps, the distances, and the ranking you can re-weight.
- Slides is the same argument as a short deck.
- The printable brief is two pages.