ID:32 Spray Icing: A major marine operational barrier in the Arctic

21 February 2023 | 08:30 - 10:00 (GMT+1)
21 February 2023 | 14:00 - 15:30 (GMT+1)

Open Session - HYBRID


Room: Hörsaal 3


Session Conveners: Sushmit Dhar (University of Tromsø – The Arctic University of Norway, Norway); Masoud Naseri (University of Tromsø – The Arctic University of Norway, Norway)


Session Description

A catastrophic ship icing incident in the Arctic waters was the sinking of fishing vessel ONEGA, which led to the loss of 17 lives on December 28, 2020, while fishing west of Yuzhny Island in the Novaya Zemlya Archipelago. There were 19 crew members onboard, out of which only two were rescued, one found dead, and 16 were not found. Apparently, the vessel accumulated heavy sea-spray icing, eventually leading to the capsizing. Similar damaging incidents have been reported on ships and offshore structures throughout time in the Arctic waters - 81 vessels were reported to be lost from the winter of 1942 to December 1970 due to icing. Two English steam trawlers, the Loretta and Roderigo, capsized and sank with their crews on January 26, 1955, in the northwest of Iceland - this accident presumably provoked investigations on the ship icing problem. There are also well-documented reports of severe icing events on offshore platforms such as “Ocean Bounty” semisubmersible during the winter of 1979 in the Lower Cook Inlet, “Sedco 708” semisubmersible on the North Aleutian Shelf during the winter of 1982, and “Sedneth II” semisubmersible around Feb 1970.

Many researchers have contributed to developing icing models, and with improvements in weather forecasting, it is possible to provide operators with prior warnings. Nevertheless, icing still possesses a safety hazard for the crew working, can affect structural stability and may damage communication and safety equipment or other critical and essential machinery. Also, the icing models are built on empirical formulas based on limited field observations and may only perform satisfactorily in distinct areas and for certain shaped structures. Meteorological, oceanographic, and sea-spray field data collection for a more extended period and on different shaped structures are desirable to improve and make the present models more robust.

The aim of this Special Session is to provide an opportunity for the researchers to share and exchange their knowledge, experience, and ongoing works in fields relevant to marine icing. 

Related topics are listed as, but not limited to:

  • Recent development in sea-spray-icing estimation models, related data collection, and icing weather forecasting models
  • Experimental and laboratory study of spray icing
  • Anti-icing and de-icing tools and techniques applicable across various maritime sectors
  • Spatial/temporal modelling and simulation of future trends
  • Probabilistic framework for studying spray icing climatology
  • Decision analyses for spray-icing-concerned risks applicable across various maritime sectors
  • Decision support system for risk-informed decisions related to spray icing
  • Spray icing risk management.



Session 1 (08:30 - 10:00 GMT+1):

  • unfold_more08:30-08:45: Spray-icing forecasting in Northern Norway

    Eirik Mikal Samuelsen
    Norwegian Meteorological Institute


    The hazardousness and risk associated with spray-icing on ships are well known and documented. Regardless of this fact, there has just recently been an increased attention on spray-icing forecasting in Norway in the last couple of years. There are mainly two reasons for this: Firstly, the methodology for forecasting spray-icing on ships has improved, i.e. the icing model MINCOG (Marine Icing model for the Norwegian COast Guard) is now routinely run twice a day and the results are easily available for operational forecasters. Secondly, there have been several periods of cold weather in the last couple of years, allowing for both moderate and severe icing to occur at the coast and in the fjords of Northern Norway. In this study, an overview of the operational ship-icing forecasting system at MET Norway is presented. Examples of reported ship-icing events are presented and compared with the calculated icing rate from the operational modelling system. In addition, for some of these cases, the atmospheric model HARMONIE-AROME with hectometric horizontal resolution is utilized in order to calculate icing rate, and compared with the icing rate calculated from the operational HARMONIE-AROME model with 2.5 km between each grid point. Ensemble prediction of MINCOG utilizing atmospheric input variables at 2.5 km resolution is also tested out. The results illuminate that there is a potential in further improving spray-icing forecasting by using input from models with hectometric resolution, in addition to ensemble prediction systems for spray-icing on ships.

  • unfold_more08:45-09:00: A Geographic Perspective on Prediction of Vessel Icing

    James Overland


    When sea temperatures are <2-3° C above the saltwater freezing point there is a likelihood of supercooling of spray during its trajectory and extreme ice accretion on vessel topside structures. Awareness of the danger is complicated by icing intensity being a hazard based on simultaneous multiple environmental parameters of strong winds and low air temperatures; warm sea temperatures and short wave fetch are mitigating factors. Therefore, vessels often depend on weather services for warnings of potential icing. Severe icing is primarily caused during strong cold-air advection events. Some regions are most dependent on cold air temperatures as in the Bering Sea, the Sea of Okhotsk and the Sea of Japan, and others are dependent on extreme winds as in the Gulf of Alaska. Low sea temperatures contribute to severe icing in the Labrador Sea, Denmark Strait, and eastern Barents Sea. Weather services have the limitation that they must consider regional differences and cannot be vessel specific. Forecasts should be easily communicated and understood, and should reflect the uncertainty in absolute rates from both the input data and the variation between vessels. For these reasons, categorical forecasts are to be preferred. For example, “heavy icing” should imply the likelihood that all but the largest vessels should take evasive action. The operational NOAA categorical vessel icing algorithm is in use for Alaska since the late 1980s. Ever improving atmospheric models are capable of providing reliable 48-h forecasts of cold-air advection, as noted in the recent Year of Polar Prediction (YOPP) evaluations.

  • unfold_more09:00-09:15: A study of the MINCOG and Overland algorithms for Subseasonal Forecasting of Freezing Spray Conditions in the Northern Hemisphere

    Todd Spindler1; Eirik Mikal Samuelsen2; Joseph Sienkiewicz1
    1NOAA/NWS/National Centers for Environmental Prediction; 2Norwegian Meteorological Institute


    The NOAA Ocean Prediction Center (OPC) issues marine warnings and forecasts for Metarea IV and XII covering most of the North Atlantic and the North Pacific. OPC weather forecasts and warnings for these areas primarily ensure the safety of ocean-crossing commercial ships and other vessels on the high seas. Hazardous maritime conditions caused by freezing spray and ice accumulation on vessel and platform superstructures in high latitudes is highlighted by tragic events such as the loss of the Scandies Rose in the western Gulf of Alaska in late December 2019. Timely warnings of impending dangerous conditions is a priority at OPC, with forecast guidance of freezing spray conditions currently available to the forecasters using a modified Overland algorithm with Global Forecast System (GFS) model forecasts of up to 17 days. Beginning with GFS v16, forecasts are now generated out to 35 days in extended runs on a once-per-day basis, This is sufficient for a subseasonal study of weeks 3-4 comparing freezing spray forecasts computed using the Overland algorithm and a simplified version of the Marine Icing model for the Norwegian Coast Guard (MINCOG) algorithm. This talk will discuss verification results for the fall and early winter of 2022 of the model icing rates against rates computed from satellite observations of 10m winds from METOP-B scatterometer measurements, Sea Surface Temperature from OSPO GHRSST analyses from NOAA NESDIS, and 2m air temperatures from Gridded NUCAPS soundings.

  • unfold_more09:15-09:30: An interdisciplinary and stakeholder-informed approach to understand sea spray icing risks and decision-making

    Dina Abdel-Fattah1; Truls Bakkejord Ræder2
    1UiT - The Arctic University of Norway; 2SINTEF Nord


    Sea spray icing presents a unique challenge to marine operations in Arctic waters and other cold environments. Understanding this phenomenon and the challenges it poses is important, since it contributes to increased risk in a wide range of marine sectors. Therefore, a better understanding of sea spray icing decision-making is critical, to ensure that tools and information developed to support sea spray icing-decisions result in better decision-making, risk perception, and operational safety. This work is part of a broader scientific project, SPRICE, in which we seek to understand what thresholds exist in sea spray icing-decision-making, and whether they differ by different stakeholders. In this session, we discuss the various spray icing-rated decisions and challenges that we have identified. We collect information on sea spray icing decision-making via interviews with relevant stakeholders, to understand not only sea spray icing-related decisions and challenges, but also how they differ according to different sea spray icing situations. We highlight therefore different decision thresholds for spray icing-concerned stakeholders, as well as what spatial, temporal, or situational factors impact these thresholds. Furthermore, we discuss the various risks associated with sea spray icing, particularly in contexts where there are extreme and hazardous conditions. Lastly, we present the novelty of this work, as an interdisciplinary project, where our findings are built upon new innovative scientific models for sea spray icing, as well as help to inform the presentation and dissemination of these findings to a broader audience.

  • unfold_more09:30-09:45: Field spray flux measurements to improve icing models for fish farms

    Sushmit Dhar1; Hassan Khawaja1; Masoud Naseri1; Kåre Edvardsen1
    1UiT - Arctic University of Norway


    One key parameter in a sea-spray-icing model for marine structures and vessels is the magnitude of liquid water impinging on the structure. However, though some field studies have been carried out for the measurement of sea spray on vessels, there have been even fewer attempts made for stationary structures, in particular fish farms. The fish farm structures installed in cold regions are vulnerable to spray icing that may lead to heavy load on the net, resulting in structural failure, and may also affect the safety of the workers.

    To calibrate icing models to suit structures such as fish farms, field measurements are required on similar structures. Under a given set of met-ocean conditions, the structure shape, size, and its relative heading play a significant role in the pattern of spray produced, which subsequently impinges to form icing. Such factors make it challenging to transfer theoretical and observational results among vessels, platforms, and different marine structures. However, apart from the logistical challenges and cost considerations associated, there are also no standard instruments available to carry out such field observations, especially for reliable spray measurements. In this study, we analyzed the feasibility of past attempts to measure sea-spray on ships and offshore platforms. Subsequently, we designed a novel spray collector system that can carry out spray measurements autonomously. The collector works on the principle of cyclone separator. CFD simulations of the collector reveal higher catchment efficiency compared to the methods attempted by previous researchers. Finally, after carrying out lab tests, the spray collector system is deployed on fish farms in complex terrains in Northern Norway.


Session 2 (14:00 - 15:30):

  • unfold_more14:00-14:15: Multidisciplinary crowdsourcing of real-time user information on sea spray icing incidents and near-misses

    Truls Bakkejord Ræder1; Dina Abdel-Fattah2
    1SINTEF Nord; 2UiT - The Arctic University of Norway


    Sea spray icing is a problem for maritime operators in northern and Arctic waters. Sea spray icing is not only a dangerous safety hazard, which, at worst, can result in loss of life, but it also impedes normal operations and can result in significant economic losses. Therefore, understanding how often sea spray icing incidents happen, as well as do not happen (what we refer to as near-misses), is important in understanding what decisions need to be made by maritime operators regarding this hazard. As a part of analysing decisions relating to sea spray icing-concerned risk, the SPRICE project is engaging in crowdsourcing information from maritime operators on sea spray icing incidents and near-misses, particularly from those who operate primarily in the Norwegian and Barents Seas. We do so via a mobile phone application with an existing user base of fishing vessels in the region, where fishing vessel staff can provide pertinent meteorological and icing-related data. We also present forecasts of existing icing models from the Norwegian Meteorological Institute in the app. Importantly, we analyse the usage of this app, to see how and when people use it most. To complement the interdisciplinarity of our project, we employ a user centric design approach, to elicit continuous user requirements, needs, and other inputs from relevant stakeholders regarding our app. This way, we can ensure the app meets and supports actual user needs, and can therefore be a useful contribution to their decision-making needs.

  • unfold_more14:15-14:30: Intelligent decisions for safe navigation in polar waters under uncertain spray icing events

    Masoud Naseri
    UiT - The Arctic University of Norway


    Accretion of sea spray ice on vessels can lead to vessel capsizing and loss of life. The reports on loss of ONEGA fishing trawler in 2020 in the Barents Sea suggested that the capsizing of the vessel, although triggered by heaving icing events and rough weather conditions, was also attributed some decisions made on board by the vessel crew. This highlights the importance of risk-informed decision-making for safe navigation in the polar waters under harsh conditions. However, the research on managing the spray icing risks for sailing in Arctic waters and research on making informed decisions prior to sailing, and while sailing in the sea, with respect to sea spray icing are scarce. Furthermore, there is no database of near-misses and vessel encounters with icing storms that can be used for performing risk assessments and generating risk-informed decision-making insights. The main proposes a modelling framework for making informed decisions with regards to spray icing risks in the Norwegian Arctic waters. This is achieved by employing machine learning algorithms to analyse historical Automatic Identification System (AIS) data. The extracted underlying patters of vessel operational parameters are further integrated with the historical met-ocean parameters and spray icing model outputs for the corresponding junctures and locations. The model outputs will be confirmed by integrating the field knowledge from fishery industry in the Barents Sea.

  • unfold_more14:30-14:45: Sea spray freezing with Magnetic Resonance Imaging (MRI) and portable Nuclear Magnetic Resonance (NMR)

    Igor Mastikhin1; Shahla Ahmadi1; Duncan Osmond1; Andres Ramires Aguilera1; Grant Wilbur1
    1University of New Brunswick


    Seawater spray and precipitation cause ice accumulation in cold ocean regions, presenting a major challenge for shipping and operating maritime equipment [1]. A development of analytical techniques to study the spray icing is important for better understanding of freezing phenomena. MRI is known for its non-invasive capabilities in measurements of solid ice [2,3]. We performed studies of sea spray freezing with an MRI scanner [4] and a portable NMR [5]. The 1H NMR signal comes from brine pockets in ice, decreasing with temperature and further brine freezing. 3D MRI showed freezing patterns and temperature gradients depending on temperature and surface geometry. NMR signal lifetimes, expressed as NMR relaxation parameters, changed with the freezing, indicating changing environment for the brine in the growing ice [4]. 3D scanning was done with a big MRI scanner. To test NMR for freezing studies in a more open environment, we also used a mobile, portable device to characterize periodic sea spray freezing on a cold surface. The device consisted of a flat 3-magnet array [6] with the detection region (~2 mm x 15 x 15 mm size) 1 cm away from the magnet surface. 1D-resolved data provided information on the brine concentration, growth rate, and NMR relaxation vs temperature at two orientations [5]The results provide information on brine parameters in freezing sea sprays, with a potential for field measurements with portable NMR.
    [1]A.R. Dehghani-Sanij et al, Ocean Eng. 143 (2017) 1–23.
    [2]M.W. Hunter et al, Appl. Magn. Reson. 36 (2009) 1–8.
    [3]J.R. Brown et al, J. Magn. Reson. 225 (2012) 17–24.
    [4] G.Wilbur et al, J. Magn. Reson. 310 (2020) 106647.
    [5] S.Ahmadi et al, J.Magn.Reson. (2022) 107109
    [6] J.C.Garcia-Naranjo et al, J. Magn.Reson. 207 (2010) 337-344

  • unfold_more14:45-15:00: Microstructure of sea spray ice

    Sönke Maus1; Paul Rübsamen-von Döhren1
    1Norwegian University of Science and Technology


    Sea spray icing on ships and offshore structures is a major safety concern for operations in polar regions. During a storm an ice layer of several centimeter thickness can grow from sea spray within an hour, which can lead to dysfunction and overloading of the structure. The prediction of this spray ice growth is still unsatisfying. One considerable uncertainty factor is the freezing process of spray ice. This ice is porous, containing pure ice, air, liquid brine, and to some degree precipitated salts. The amount of brine within the ice affects the freezing rate and determines the total load and the adhesion of the ice to surfaces. First studies of the microstructure of sea spray ice were based on magnetic resonance imaging with a spatial resolution of 0.2 to 1 mm, which is insufficient for a quantitative description of the pore space. In the MICROSPRAY project at NTNU we observed, for the first time, the microstructure of saline spray ice by X-ray microtomography, with a spatial resolution of down to 10 micrometers. We observe brine drainage channels within saline spray ice that are similar to those in sea ice (yet formed much faster, in less than an hour compared to days). Here we present a first analysis of these 3D images, reporting detailed statistics of the location of salt in sea spray ice as well as pore size distributions of brine inclusions and drained brine channels in the micro- to millimeter range. Observations are presented for spray ice growth under air temperatures from -7 to -15 °C. We discuss some implications of these new data for modelling of the growth rate and properties (in paricular salinity) of sea spray ice.

  • unfold_more15:10-15:15: Adhesion of sea spray ice

    Paul Rübsamen-von Döhren1; Sönke Maus1
    1Norwegian University of Science and Technology


    Sea spray ice can grow several centimeters an hour during sea spray events and can potentially lead to the inability to navigate, and submerging of small vessels. Thus, it is a major concern for safety of operation in polar regions. There is limited knowledge about the growth process and the properties of ice grown in a sea spray event. Of particular concern is the adhesion strength of the ice, as it may limit the amount of sea spray ice growing before dropping due to its own weight. So far data about the adhesion strength of spray ice is very limited. On the one hand it is difficult to measure; on the other hand, the focus during a spray event is the deicing to ensure operation of the structures.

    In the MICROSPRAY project at NTNU we perform, for the first time, systematic ice adhesion measurements, growing spray ice in the laboratory, with air temperatures from -7 to -15 °C. Besides the impact of the atmospheric conditions on the freezing process and adhesion strength of the resulting spray ice we also study the effect of water salinity and of the size of the icing surface. In the experimental setup we have the possibility to measure the adhesion of the ice in situ, shortly after the growth process.

    Here we present a first analysis of the growth temperature's impact on the sea spray ice adhesion strength.

  • unfold_more15:15-15:30: Understanding Physical Properties of Fresh Water and Marine Ice using Multiphysics Modelling and Infrared Thermography

    Hassan Khawaja1
    1UiT - The Arctic University of Norway


    Ice has unique physical properties, which tend to vary depending on how, where, and when it forms. In the presented work, we looked deeper into various physical properties of ice. In one of the studies, we focused on studying the ice adhesion on two types of polymers; polyurethane (PU) and polyvinyl-chloride (PVC). It revealed interesting behaviors with a viable potential for roads/highways coated with respective materials. The study was validated using Multiphysics modeling techniques involving the finite element method. In another study, we examined the thermal diffusion characteristics of fresh water and saline ice cubes using infrared thermography. A Multiphysics based method on finite difference method is employed to study the thermal behaviors revealing thermal characteristics of ice; i) thermal conductivities and ii) heat transfer coefficients. The results further led to the development of ice detection and mitigation solution for marine operations in cold regions.

Every presenter will get 10 minutes to present followed by 3-4 minutes of Q & A session.