At 09:47 on the morning of 10 March 2025, the 141m container vessel Solong struck the anchored 183m oil/chemical tanker Stena Immaculate at 16 knots approximately 10 nautical miles off the Humber. The collision resulted in a breach of one of Stena Immaculate’s cargo tanks, resulting in ignition of the aviation fuel it was carrying and a significant fire.
Miraculously, both crews were able to abandon ship, however, unfortunately a crew member from the Solong was unable to be found and is believed to have died in the accident1.
A detailed and thorough investigation by the UK’s Marine Accident Investigation Branch has commenced and, once complete, will shed light on the circumstances which led to this incident. See MAIB interim report.
In this article, we make several observations pertinent to management of navigation safety in the UK following the incident.
TL:DR?
1. UK waters are some of the busiest and most complex waterways in the world.
2. Serious collisions between commercial ships in the UK are rare.
3. Analysis of historical accidents can be a tool for assessing risk, but their infrequency can make this challenging.
4. Modelling can be a tool to overcome these deficiencies
5. As the waters around the UK coast get busier, and more offshore windfarms are developed, there is a greater need for strategic and evidence-based management of navigational safety.
Context
Location
The incident occurred outside of the statutory harbour area of the Port of Humber. Grimsby and Immingham are the UK’s second busiest port by tonnage accounting for 46 million tonnes in 20232.
It is a busy and complex area, with multiple offshore wind farms, sandbanks, a busy anchorage, pilot boarding stations and ship routeing measures (Gallery Image 2). Located outside of the port’s limits, vessels navigate neither under pilotage of control by Vessel Traffic Services, following the maritime rules of the road, the Collision Regulations.
The Solong had navigated this route numerous times. Analysis of ship tracking information undertaken by NASH Maritime (Gallery Image 3) shows the vessel had taken an almost identical route on 20 occasions prior to the incident, eight of which were in the previous month. The Stena Immaculate meanwhile had anchored in a busy anchorage alongside multiple other cargo ships and tankers.
Incident rates
Fatal collisions between commercial vessels in the UK are rare. Data published by the MAIB in their annual reports shows that there had been no loss of life reported to the MAIB in UK waters between 2020 and 2023 for merchant vessels over 100 GT.
Similarly, between 2011 and 2023, only two UK registered merchant vessels were lost (a sailboat and a tug)3. One of the last collision incidents between two large commercial vessels which resulted in loss of life within the MAIB dataset was in 2001 in the English Channel.
That is not to say such incidents do not occur.
The Chief Inspector of the MAIB recently pointed to several recent examples the MAIB has been involved in. The Scot Carrier and Karin Hoj, resulting in two fatalities, and the Polesie and Verity, resulting in the loss of five lives, albeit neither of these incidents occurred within UK waters4. However, it should be recognised that most incidents involving ships result in relatively minor consequences such as minor damage, no pollution or fatalities.
Analysis of MAIB incident data undertaken by NASH Maritime shows that across all vessel types, only 2.6% of reported collisions result in injuries and 0.7% result in fatalities. This is significantly less than other incident types such as capsize or flooding.
Furthermore, a breakdown of collisions by vessel type shows that commercial cargo and tanker vessel types have a significantly lower rate of injuries and fatalities than fishing, recreational or tug and service vessels. This is likely a function of both the greater structural integrity of commerical shipping and the underreporting of minor incidents of small craft, biasing the MAIB dataset to more serious incidents.
Managing navigational safety
Maritime risk management is challenged to some degree by the relative infrequency at which incidents occur relative to the volume of traffic within an area.
With such a small sample size of serious incidents, decisions made on expensive mitigation measures can be hard to justify. If we were to focus on specific waterways or areas, rather than nationally, that sample size diminishes further.
Use of Incident Data
One means to overcome this might be to look at less serious incidents and near misses, which occur more frequently and use this to predict more serious events.
This concept has parallels to the Heinrich’s accident pyramid5, similar accident chains have varying levels of severity. By identifying and preventing lesser incidents, such as unsafe acts, more serious incidents such as fatalities could be avoided. He postulated that for each 300 unsafe acts, 29 minor injuries and one major injury or fatality would occur.
The MAIB’s annual report for 20236 provides an equivalent breakdown of incident severity for UK merchant vessels of over 100GT and show a pyramidal shape.
However, given this accounts for only 146 incidents across the UK, focusing on specific regions or areas, such as the outer Humber estuary, would hugely limit the sample further, making robust decisions on navigational safety untenable.
An analysis of the MAIB’s dataset from 1992 to 2024 shows that there had been only 11 collisions reported within 10nm of the Solong-Stena Immaculate incident site, of which only two had been in the last 15 years.
Modelling
A key challenge of historical incident data is its infrequency, and many studies have noted that minor incidents are frequently underreported7,8.
Furthermore, a reliance on incident data results in a reactive approach to maritime risk. For example, the Safety of Life at Sea (SOLAS) convention followed Titanic, MARPOL followed Torrey Canyon and the UK’s Port Marine Safety Code followed the Sea Empress9.
Yet, it is better to be proactive to prevent incidents before they occur rather than reactive to each new tragedy.
Modelling of vessel traffic, undertaking risk modelling and consultation, are well regarded methods of understanding maritime safety and overcoming limitations with accident data alone.
The use of tools such as encounter modelling can highlight hot spots where vessels meet more frequently, or models such as IWRAP will use mathematical formulas to understand the potential likelihood of collisions, allisions and groundings.
These methods give invaluable intelligence to support robust and targeted application of appropriate mitigation or the design of new developments, such as offshore wind farms.
Conclusions
Reactive responses, whereby mitigation is only implemented once a major incident has occurred, is clearly an ineffective approach to accident prevention.
A proactive approach to risk management, relying on evidence-based analysis can be a tool to identify areas of higher risk before incidents occur, and ensure that the risks of collision at sea are reduced as far as possible.
As the waters around the UK coast get busier, and more offshore wind farms are developed, there is a greater need for strategic and evidence-based management of navigational safety.
References
1. https://assets.publishing.service.gov.uk/media/67ebb6a1632d0f88e8248ada/Solong-StenaImmaculate-InterimReport.pdf
2. https://www.gov.uk/government/statistics/port-freight-annual-statistics-2023/port-freight-annual-statistics-2023-overview-of-port-freight-statistics-and-useful-information
3. https://assets.publishing.service.gov.uk/media/67110acc9cd657734653d72b/MAIBAnnualReport2023.pdf
4. https://assets.publishing.service.gov.uk/media/67ebba71e9c76fa33048c561/2025-1-SafetyDigest.pdf
5. Heinrich, H. (1931). Industrial Accident Prevention. A Scientific Approach.McGraw-Hill Book Company, Inc. New York.
6. https://assets.publishing.service.gov.uk/media/67110acc9cd657734653d72b/MAIBAnnualReport2023.pdf
7. Qu, X. Meng, Q. and Li, S. (2012). Analyses and Implications of Accidents in Singapore Strait. Journal of Transportation Research Board, 2273, pp. 106-111.
8. Hassel, M. Asbjornslett, B. and Hole, L. (2011). Underreporting of maritime accidents to vessel accident databases. Accident Analysis and Prevention, 43,pp. 2053-2063.
9. Awal, Z. and Hasegawa, K. (2017). A Study on Accident Theories and Application to Maritime Accidents. Procedia Engineering, 194, pp.298-306.
