There is an estimated 1 in 10k - 100k years chance that all 8 of the Nov 2024 - Jan 2025 Baltic Sea cable breaks are coincidental. Background data is hard to get by but no matter which real data or reasonable assumption one picks it is extremely unlikely all 8 of them are explained by chance. 🧵
Illustration of a 1 in 108 908 years chance that all 8 of the Nov 2024 - Jan 2025 Baltic Sea cable breaks happens with present background levels. Above is 1 in 10 609 years illustrated.
Two alternative ways of illustrating the extremely low probability of the Baltic Cable breaks being normal. Red dots in each, promise!
This is the data I have based my various estimates on. I chose to use the only two real data series I have found. Thanks @dahlgrenf.bsky.social for help! Ill describe the two I picked in more detail below. C: UK example and E: Bornholm example.
C: 1 in 108 908 years. During 2007-2010 13 incidents were recorded in UK waters. I extrapolate this without adjustment to be true for Baltic Sea. One could argue this makes the estimate low, but one could also argue it makes it high. Figures are from this paper, rather poor on details:
cil.nus.edu.sg
A 2011 paper by Mick Green & Keith Brooks: The Threat of Damage to Submarine Cables by the Anchors of Ships Underway Unfortunately it is not clear on how many ships or cables is counted, nor is the area clear. But, best we got. These are some figures on UK waters and Baltic Sea. Comparable enough
E: 1 in 10 609 years. A Baltic Sea one. The Bornholm power cable broke by anchor from a ship underway 5 times in 20 years as shown by TV2 Denmark. This rate is extrapolated to be true for all relevant Baltic cables (n=18), meaning with this assumption they all break every 4th year.
www.tv2bornholm.dk
If we go back to my table you see I listed some other made up assumptions. To reach even a 1 in 100 year event we need to double the Bornholm estimate, ie all the cables break on average every 2nd year. But they don't. And the cable breaks Yi Peng 3, Eagle S and Vezhen caused are abnormal.
In my opinion most of the cable breaks are likely intentional, but that we can not prove this way. If not intentional, they are extremely negligent, which leads to the same conclusion : these ships needs to be prosecuted. Further, I think the normal background rate is likely somewhere between C and E but given the bad access to background data we can not say for sure. Besides the references in table I have read various other papers including very expensive industry ones by CIGRE. I have not found any summary with a high scientific quality and the estimates vary wildly, even within papers. They are not precise in what they are describing nor how data was collected. Alas, these estimates are very imprecise. But again, we know Baltic Sea cables do not break every 4th year. One can also surely critique that I switch between quarters and years, but it is a compromise to make it understandable. Fact is, we had 8 breaks in one quarter and for the stat calculations I have used quarters. That is what is tested, how likely is a 8 event quarter. Oh, I did not include the maybe damaged Russian cable. I have only counted confirmed and complete failures and I do not believe the Russian Baltika cable was completely severed. Power cable NordBalt, which also received damage is not included either as it stayed operational.