663

1 INTRODUCTION

Marine activities in the Arctic have long traditions.

The first ships entering arctic areas were just for

exploration purposes, but from the early 17th century

to present time, various whale-catching, fishing and

seal catch-based activities evolved in the area. The

first whale-catching station was established in

Smeerenburg on Spitzbergen in 1614.

The marine activities often had to deal with major

safety challenges, like encountering sea ice, getting

frozen into the ice, remoteness, darkness, poor

visibility, long distances, etc. However, one of the

major safety challenges these ships had to deal with,

and still have to deal with is ship icing (Dyrcz, 2019).

This phenomenon occurs in temperatures below

subfreezing of seawater where sea spray may, in

combination with snow and different atmospheric

conditions, freeze on the ship's hull and

superstructures, followed by both reduced ship

stability and lowering freeboard. During the last 45

years, there has been a motivation for different

maritime research to develop better forecasting

models for accretion of sea spray icing. Several

researchers (Kachurin et al., 1974; Stallabrass, 1980;

Overland et al., 1986; Makkonen, 1987; Horjen, 1990;

Henry, 1995; Lozowski et al., 2000), all base their

calculations on spray flux ice accumulation rate (RW)

in kg/m²s, by solving the heat flux equation (Eq. 1) for

a freezing surface. The heat fluxes on the right side of

the equation contribute to freezing when positive and

melting when negative, thus contribute to the heat

release due to freezing flux of water (Qf), i.e. ice

Stability of Vessels in an Ice-free Arctic

K. Johansen, M.P. Sollid &

O.T. Gudmestad

UiT The Arctic University of Norway

, Tromsø, Norway

ABSTRACT: One consequence of the declining ice cover in the Arctic is increased areas of open seas. These new

open sea areas lead to some challenging aspects related to ship stability. Longer fetch lengths, associated with

build-up of larger waves followed by increased conditions for sea spray icing on vessels is one aspect. Open

seas in combination with cold atmospheric temperatures is a prerequisite for polar low pressures to occur. Polar

lows may represent an additional aspect of increased icing on vessels by heavy snow in addition to extensive

sea spray ice accretion.

Over the last decades, different formulas for prediction of sea spray ice accretion rate on ships were developed

to form basis for ice accretion warnings. Some of these formulas seem to have certain limitations and appear to

be conservative. Important limitations of some formulas are considerations regarding heat flux, relationship

between wind and waves, and ice accretion related to Polar lows.

This paper will take a closer look at the accuracy and the realism of different ice accretion formulas and, related

to this aspect, we will also discuss whether ship officer candidates receive sufficient maritime education and

training (MET) related to realistic ice accretion and ship icing aspects.

http://

www.transnav.eu

the

International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 14

Number 3

September 2020

DOI:

10.12716/1001.14.03.19

664

accumulation, as given on the left side of Eq. 1,

(Samuelsen et al., 2015).

f = Qc +Qe + Qd (1)

where

f - Heat flux released by freezing (W/m²)

c - Convective heat flux (W/m²)

e - Evaporative heat flux (W/m²)

d – Heat flux from incoming water droplets (W/m²)

According to e.g. Samuelsen et al. (2015), this

equation has limitations by just taking into account

few important heat flux aspects. The importance of

implementing relevant and realistic knowledge and

understanding about ship icing accumulation aspects

in MET is essential for executing proper seamanship

in the Arctic. More advanced heat flux equations will

be discussed in this paper.

Problems concerning ship icing related to reducing

ship stability and lowering freeboard are largely

considered for smaller ships like ordinary fishing

vessels, crab fishing and seal catcher ships from

approximately 10m to 50m length, but there are

several examples of ships larger than that, which have

been exposed to reduction of stability due to ice

accretion. One specific example is the Norwegian

coastguard vessel, KV Nordkapp, which went into

serious icing conditions between Bjørnøya (Bear

Island) and Hopen in February 26-27, 1987, followed

by an estimated accumulation of approximately 110

tons of ice in just 17 hours.

This case shows that ice could accumulate rapidly

if conditions for icing are present, like sailing into a

Polar low. Figure 1 shows pictures of icing of the

superstructure of KV Nordkapp (Samuelsen et al.,

2015).

Figure 1. Icing on Coast Guard vessel KV Nordkapp on 27

February 1987, Barents Sea (Photo by Prof. S. Løset).

A Polar low is defined as: “Small-scale, intense,

short-lived atmospheric low-pressure system

(depression) found over the ocean areas poleward of

the main polar front in the northern and southern

hemispheres. Polar lows can be difficult to detect

using conventional weather reports and are a hazard

to high-latitude operations”, (PWOM, 2019).

An estimation of the stability aspects and an

assessment of observed and estimated ice rate of this

specific example of a Polar low related ship icing

event will be given later in this paper.

Additional ship icing aspects related to Polar lows,

individually or in combination with ordinary sea

spray ship icing, will also be highlighted.

The consequences of Polar lows could be a

severely increased probability of ship icing by: heavy

snow, strong wind gusts, thunderstorms, radically

dropping temperatures, rising waves, waterspouts,

unpredictable conditions etc.

The goal of this paper is to set focus on better and

more adequate understanding of state-of-the-art

research of different meteorological aspects of ice

accretion on ships, and also set focus on to what

degree STCW (IMO STCW, 2011) can be used as

guidelines for MET schools, with respect to ship

stability related to marine icing.

Use of “state-of-the-art” theory in MET schools is

one of the prerequisites for executing proper

seamanship. Overland et al. (1986) define “potential

icing rate” as: “a sustained icing rate by a vessel that

is not actively avoiding icing through heading

downwind, moving at slow speeds, or avoiding open

seas”.

Note, however, that this paper will not focus on

the underlying causes of how polar lows occur, and

in-depth explanations of how ship icing forecasting

accumulation algorithms and heat flux equations are

structured.

The analytical part of the study is approached by

an evaluation of “state-of-the-art” theory about

forecasting and accumulation models for ice

accretion, then by a discussion of Polar lows influence

on ship icing, followed by a case study of how KV

Nordkapp was influenced by a Polar low on 27

February 1987. In addition, a document study of the

icing related content of STCW (IMO, 2011) and the

Polar Code (IMO, 2015) is given.

2 BACKGROUND

Over the last decades sea ice retreats in Arctic have

led to opening of new operation areas for increasing

maritime activities like cod fishing, shrimp trawling,

crab catching, seal hunting, energy harvesting and

tourist related activities, etc. This sea ice retreat has

also opened for new areas of occurrence of Polar lows

due to increasingly open seas along the southern ice

edge of the Arctic Ocean, including the Barents Sea,

the Bering Sea and the Beaufort Sea (Kolstad and

Bracegirdle, 2008). Figure 2 shows the first Polar low

satellite image in open sea north of Svalbard.

MET schools offering sea officer education and

certification related to the Polar Code must strive to

use “state-of-the-art” theory regarding Polar lows,

related to where they can occur and the impact these

lows have on ship icing. In addition, there are new

research-based theories referred to forecasting of

accumulation on ship icing that should be included in

MET. The importance of having proper relevant

understanding of these aspects forms the basis for

preventing accidents related to human elements.

665

Figure 2. NOAA satellite image, Svalbard 11.26 UTC, 8.

January 2010. Image: NOAA/met.no

According to EMSA‘s analysis of 1170 ship

accident events, 60.5 percent were directly attributed

to erroneous human actions (EMSA, 2016). Other

research shows that 75-96 percent of maritime

accidents are directly or indirectly caused by some

form of human error (Hanzu-Pazara et al., 2008).

However, human elements represent a significant

contribution in all ship accidents.

MET schools must therefore strive to provide the

students with the best possible basis for exercising

their practice. In-depth understanding and

knowledge of any subject is a prerequisite for

exercising well-executed actions. (Johansen and

Batalden, 2018).

Ship icing has caused many accidents and near

accidents in polar waters. Raise of the centre of

gravity by ship ice accumulated above deck level of

vessels has led to reduction of stability followed by

severe scenarios like, capsizing, submerging, and the

loss of lives (Shellard, 1974).

Examples of recent documented incidents

associated with reduction of stability and buoyancy

are the following: January 26, 2007, the 75-feet fishing

vessel Lady of Grace sank in Nantucket Sound at the

east coast of the USA (United States Coast Guard

2008), and January 3, 1999 at the coast of Northern

Norway another ship sank, costing the lives of 4 and 3

persons, respectively. The latter incident is revealed

by investigating the database of ship accidents from

the Norwegian Maritime Authority (2014).

Another concrete recent accident arose on 11

February 2017 offshore Alaska. A crab-fishing vessel

with a crew of six suddenly vanished north of St.

George Island in the Bering Sea. From the reports of

other vessels, it is believed that icing may have been

the cause of this accident that led to capsizing,

submerging, and the disappearance of the vessel (The

Seattle Times, 2017).

From the book “Alaska Shipwrecks 1750-2015” by

Captain W. Good and M. Burwell (2018), some

relevant icing related accidents are described, like:

− The 58-foot fishing vessel Hunter capsized and

sank January 2007 in the Shelikof Strait, location:

South Central Alaska 57º 26̍’ N 156º 01’ W. High

winds and freezing sea spray caused heavy icing

followed by instability and starboard list. The ship

sank in ten minutes. (Good and Burwell, 2018, p.

213).

− The 80-foot steel crab-fishing vessel Rosie G took

on water from the stern and sank January 30, 1997

northwest of Cape Cheerful, location Southwest

Alaska 53º 52̍’N 166º 32̍’W. The Rosie G developed

severe ship icing in -40°C followed by a 45°

starboard aft list causing serious water ingress.

The Rosie G sank by the stern because of severe

ship icing (Good and Burwell, 2018, p. 460).

− The 96-foot Lin-J capsized in heavy icing

conditions northwest of Saint Paul Island March

18, 1999. There is no further information of the

accident, probably because the entire crew was lost

(Good and Burwell, 2018, p. 717).

All these accidents are located in Polar waters, but

there are examples of severe ship icing further south

e.g. on 14 February 1979, North Sea, and west of

Thorsminde, Denmark, 6 fishing vessels sank due to

ship icing, followed by reduction of stability and

buoyancy. The weather conditions for these cases

were, eastern breeze, strength 8-10 Beaufort, and the

temperature approx. - 8 ° C i.e. perfect conditions for

ship icing (sbib.dk, 1979). 15 fishermen were lost on

that day.

The accidents presented are just some few

documented examples related to ship icing, which

shows there may be a potential in strengthen the

understanding of this phenomena. In addition, there

are near accidents like the KV Nordkapp case, where

ships have experienced reduction of stability, without

serious consequences. For these cases, accumulated

sea ice lead to a rise of the vertical centre of gravity

and thus reduced metacentric height (GM) followed

by reduced GZ levers.

3 MARINE ICING MODEL

In the physical models for wave-ship-interaction icing

the source of water is expressed in terms of a spray

flux (Rw). This flux provides an upper boundary on

the amount of ice that is accumulated per unit time. In

order to derive the rate of ice accretion (Ri) at a certain

position of the ship, a surface energy balance is

assumed between the heat released from the freezing

flux of water (Qf) and the most important heat fluxes

from the atmosphere, ocean, and underlying surface

(Samuelsen, 2018).

A general full set of heat fluxes per unit area of

spray flux ice accumulation (RW) in kg/m²s acting on

a plate area is written as (Eq. 2).

f = - Qv – Qk + Qc +Qe + Qd + Qr +Qs +Qcond (2)

where

f - Latent heat released during freezing (W/m²)

v - Viscous/frictional/aerodynamic heating from the

air (W/m²) (Makkonen, 1984)

k - Kinetic energy of spray converted to heat in the

interaction process (W/m²) (Lozowski et al., 1983)

666

c - Convective or sensible heat flux from the air

(W/m²) (Makkonen, 1987, 2010)

e - Evaporative or latent heat flux from the air

(W/m²) (Makkonen, 1987, 2010)

r - Heating or cooling from radiation (W/m²)

(Horjen, 1990), (Lozowski et al. 2000)

d - Heating or cooling from the spray (W/m²)

(Henry, 1995)

s - Heating or cooling from snow (W/m²) (Horjen,

1990)

cond -The conductive heat flux through the ice (W/m²)

(Kulyakhtin et al., 2016)

An explanatory illustration of how ice accretion is

set in context with the heat flux balance is shown in

figure 3.

Figure 3. Schematic illustration of the relationship between

sea spray icing and heat flux balance (Dehghani-Sanij et al.,

2017)

Different researchers take into consideration

different combinations of these heat fluxes in their ice

rate prediction models. According to Samuelsen

(2018), the Modified Stallabrass, Overland and the

Marine-icing model developed for the Norwegian

Coast Guard (MINCOG) represent the only physically

based spray flux icing models applied in operational

weather forecasting. The Overland icing rate

predictor model (Eq.3) is based on the heat flux

equation (Eq. 4) and ends up in a schematic icing class

rate, table 1 (Samuelsen, 2018).

( )

1 Φ 

VT T

SST T

−

+−

(3)

where

r = Icing predictor (m/s °C)

V = Wind speed (m/s)

f = Freezing point of sea water (°C)

a = Air temperature (°C)

SST = Sea Surface Temperature (°C)

Φ = 0.3 °C-1 (Overland, 1990), and 0.4 ºC-1

(Overland et al. 1986)

f cd

QQQ= +

(4)

Table 1. Icing class and rate based on (Overland, 1990)

_______________________________________________

PR <0 0 -22.4 22.4 -53.3 53.3-83.0 >83.0

_______________________________________________

Icing None Light Moderate Heavy Extreme

Class

Icing 0 <0.7 0.7-2.0 2.0-4.0 >4.0

Rates

(cm/hour)

_______________________________________________

This model is applied as basis for operational

weather forecasting (Ekeberg, 2010). A comparison of

the applied physically spray flux icing models shows

a gap between the Overland and the other models

This comparison will be presented later in the paper,

(see Figure 9).

4 POLAR LOW CHALLENGES

Ordinary weather is predictable in that both low- and

high-pressure systems and different atmospheric

conditions can easily be detected by satellite images.

This predictability makes it possible for meteorology

authorities to predict weather conditions with

sufficient accuracy, but this is not applicable for

predicting Polar lows. These lows are unpredictable

i.e. when and where they occur, what direction they

take, strength, duration and atmospheric conditions

(Nordeng & Rasmussen 1992). As Polar lows affect

the stability of ships, those who sail ships into Polar

low areas must have competence and knowledge of

these aspects. Figure 4 shows a satellite image of the

Polar low that hit KV Nordkapp at February 27, 1987.

Polar lows like the one in Figure 4 will raise to

strong northerly winds in combination with heavy

snowfall, west of the epicentre. According to

Samuelsen (2017), this snowfall may lead to an

additional negative effect on ship stability.

Figure 4. Satellite image of the Polar low over the Barents

Sea on 27 February 1987. Image imported from the French

Wikipedia.

667

5 THE KV NORDKAPP CASE

The KV Nordkapp case was chosen because of icing

caused by the ship sailing into a Polar low, Figure 4,

and because of a documented comparison between

observed ice rate and icing rates calculated by

different icing rate models as presented by Samuelsen

et al. (2015). See Figure 9 in the discussion.

Figure 5. Synoptic situation at 26 and 27 February 1987, and

the position of KV Nordkapp (red cross) during the trip

when Met-ocean observations were taken. The time of the

observation is in UTC/Z (Universal time center/Zulu time).

The green line is an approximate position of the ice edge at

26th February (Initial time 1987-02-26 06Z). The red dots

mark the approximate position of the Polar low according to

satellite image information from 0428z, 0853z, 1243z and

1702z, Samuelsen et al (2015).

Figure 5 shows KV Nordkapp’s positions during

the voyage, influenced by the weather conditions

west of the Polar low epicentre. The air temperature

from 06z to 21z was between -12°C to - 20°C. There

was wind from north-west between 20 - 30 m/s

followed by maximum wave height 7.5m and sea

surface temperature (SST) between +3 to -4°C; thus,

conditions for ship icing were optimal (Samuelsen et

al., 2015).

Samuelsen et al. (2015) present a comparison

between Overland, Stallabrass and own test models

against observed ice accumulation rate for this

specific case, and this will be highlighted in the

discussion.

5.1 Coast Guard vessel KV Nordkapp’s characteristics

The main particulars of the Coast Guard vessel KV

Nordkapp are reported in Table 3, and the ship

profile is shown in Figure 6.

Table 3. KV Nordkapp`s main particulars

_______________________________________________

Length over all 105.05 m

Length between perpendiculars 97.50 m

Beam 14.60 m

Depth 7.50 m

Lightweight (∆_ls) 2785.4 ton

Max operating(∆max) 3579 ton

Min operating (∆min) 3239 ton

_______________________________________________

Figure 6. KV Nordkapp’s profile from the GA plan.

5.2 KV Nordkapp’s stability reduction

The calculation of the stability reduction of KV

Nordkapp due to icing is based on an estimated

weight of 110 ton accumulated ice, Samuelsen et al

(2015), at a vertical position above the baseline

(VCGice = 11.557m), as shown in Figure 7. VCGice is

the average vertical centre of gravity for accumulated

ice, as reported by the stability manual of the ship.

The stability manual was originally developed

when the ship was built in 1982, but it was

redeveloped in 2017 according to the same criteria

that can presently be found in the IMO requirements

regarding intact and damage stability (IMO, 2008).

Based on information by the Norwegian Navy, KV

Nordkapp was stated with loading condition code 2,

fully loaded, at departure 26 February 1986. This

condition is equivalent to “max operating condition”

(∆max). According to the KV Nordkapp stability

booklet, this condition corresponded with the data

reported in Table 2 and visualized in Figure 7.

Table 2. KV Norkapp`s loading condition 2 particulars

_______________________________________________

Displacement (∆0) 3579 ton

Mean draught (T0) 5.120 m

KG0 (incl. FSC) 6.395 m

KM0 7.532 m

G0M0 (incl. FSC) 1.137 m

KGmax 6.691 m

_______________________________________________

KGmax (Table 2) corresponds to the maximum KG

according to both intact and damage stability

requirements.

Figure 7. KV Nordkapp from the GA plan of the ship,

showing the center of gravity of accumulated ice (VCGice),

initial vertical center of gravity (KG0), initial depth (T0),

668

initial metacentre (KM0) and initial meta-center height

(G0M0).

Departure displacement, including the 110-ton

accreted ice, became 3689 ton. According to the

stability booklet, this increased displacement was

followed by a corresponding mean draught Tice

5.273m, KMice 7.464 m and KGmax-ice 6.687m. Based

on the following equation (Eq. 5), the center of gravity

rises by the accreted icing of 110 ton to:

0 ice ice

ice

KG M VCG

KG 6.594

∆× +∑ ×

∆+

(5)

KGice ended below KGmax-ice, thus within the

stability booklet requirements by (6.687 – 6.549) m =

0.138 m. However, considering also the variation from

KM0 to KMice, the ice accretion led to a variation of

metacentric height (GM) from 1.137 m to (7.464 –

6.549) m = 0.915m, i.e. a reduction of 1.137 – 0.915 =

0.222 m. The corresponding reduction of the KV

Nordkapp stability is visualized by the GZ curve,

Figure 8.

Figure 8. KV Nordkapp`s GZ curve based on departure and

ice accreted conditions, February 26 and 27 1986.

6 MINCOG METHODOLOGY

As mentioned in the introduction, several different

researchers have tried, for decades, to develop

predictions of sea spray ship icing.

Samuelsen (2017) presents a review of the most

central research done in this area, as well as a

presentation of a new model, the Marine Icing model

for the Norwegian Coast Guard (MINCOG). This

model is developed on the basis of different data sets

derived from observation of 37 ship icing events

obtained from the Norwegian Coast Guard. In his

thesis, Samuelsen presents the uniqueness of this

model by a high level of accuracy compared to

currently applied methods.

One of the findings in Samuelsen’s research is the

importance of not relating wave information with

wind parameters in the model because there could be

strong winds close to the ice edge, but short fetch

lengths for waves to develop (Samuelsen, 2017).

Regarding waves, in his research he also found that

the nature dictates an upper limit of ship icing from

interaction between wave and ship, because high

waves rarely coexist with very low temperatures

(Samuelsen, 2017). The amount of sea spray related

icing also depends on ship type (bow height and

shape), and wave characteristics.

Samuelsen (2017) also highlights that inclusion of

snow may be an important contribution for ship icing

to accumulate, and that this contribution ensues most

frequently during cold air outbreaks from the ice and

in areas influenced by Polar lows.

The MINCOG heat flux equation (Eq. 6) does not

take into consideration the heat fluxes flux

components - Qv – Qk +Qs +Qcond (see Eq.2). The

reason for this is partly to simplify the equation, but

also that the data set showed that the neglected heat

fluxes had little impact compared to empirical results.

f = Qc +Qe + Qd +Qr (6)

The effect of snow, Qs, is neglected because of

uncertainties related to amount of water accumulation

by the snow (Samuelsen and Naseri, 2018). Other

research shows that wet snow has low density and

weak adhesion during forming. Therefore, wet snow

may have limited impact on icing. However, Table 4

shows that the adhesion is strong when the snow is

frozen.

Table 4. Typical properties of accreted atmospheric ice

(Fikke et al., 2006)

_______________________________________________

Type Density Adhesion & General Appearance

of ice (kg/m³) Cohesion Colour Shape

_______________________________________________

Glaze 900 strong transparent evenly

distributed/

icicles

Wet 300-600 Weak white evenly

snow (forming) distributed/

Strong eccentric

(frozen)

Hard 600-900 strong opaque eccentric,

rime pointing

windward

Soft 200-600 Low to white eccentric,

rime medium pointing

windward

_______________________________________________

7 THE STCW’S AND POLAR CODE’S

REQUIREMENTS TO COMPETENCE

The STCW Standard of competence (IMO, 2011)

represents the compulsory minimum requirements

for the acquisition of a certificate as an on-duty deck

officer on ships with a gross tonnage of 500 or more.

Until the Polar Code (IMO, 2015) had its entry into

force in 2017, the requirements of competence

regarding ship icing was supposed to exclusively be

given by the International Convention on Standards

of Training, Certification and Watch keeping for

Seafarers, 1978 (IMO STCW, 2011). Tables A/II-1 and

A-II/2 in the STCW convention, describe minimum

standards of competence for master and officers

operating a ship. The only STCW requirement related

to ice was a general knowledge about sea ice

conditions when planning a voyage i.e. no

requirements related to ship icing.

669

When the Polar Code entered into force, the

content was added into the governing STCW

Convention followed by additional minimum

requirements for training and qualifications of

masters and deck officers operating in polar waters,

as given in chapter V/section A-V/4. The additional

Polar Code related competence requirements became

divided into basic training for all officers on-board

and an additional advanced training for master and

chief mates. The convention then stated that

candidates for basic as well as advanced training

should have knowledge and understanding of the

implications of spray-icing, danger of icing up,

precautions to avoid icing up and options during

icing up (IMO, 2011).

There are some aspects regarding these

requirements that are challenging for the acquisition

of competence with respect to Ship icing:

1 The Polar Code is only mandatory for ships

operating in polar waters and the definition of

“Polar waters” is more a political definition than a

meteorological definition and therefore icing

events are also likely to occur outside the area of

polar waters.

2 The Polar Code is implemented into the STCW

convention, but this code is only applicable to

ships larger than 24 meters in length. This means

the major part of fishing vessels is not regulated

under the Polar Code, thus, the Polar Code

requirements with regard to icing do not have

implications for the training requirements for crew

on-board these vessels.

8 DISCUSSION

This paper shows there are still a lot of challenges

regarding knowledge and understanding both with

respect to the amount of and the consequences of ship

icing. To give ship officer students’ relevant

understanding of proper prediction of ice accretion,

presupposes that the MET schools base their learning

on “state-of-the-art” theory, both for the matter of

design and operation.

Samuelsen et al. (2015) comparison between real

icing rate using the Overland, Stallabrass and own

icing rate models show some gaps. According to

Figure 9, one can notice that Overland`s estimated

icing rates (blue lines) shows large differences with

respect to the real icing rate (pink line). It is, thus,

questionable that the Overland`s model is applied as

both basis for operational weather forecasts and as

theory for MET. The T1 model by Samuelsen (2018)

and the ModStall model M3 by Stallabrass (1980)

correlate well with the real icing rate.

Figure 9. Observed icing rate (pink color) and icing rates

from nine different algorithms related to the KV Nordkapp

case. Three Overland algorithms (O1, O2, O3) in blue

colour, three ModStall algorithms (M1, M2, M3) in green

colour and three Test model algorithms (T1, T2, T3) in black

colour (Samuelsen et al., 2015).

This outcome is an important reminder for

weather forecasters, STCW and for MET schools in

developing requirements and curriculum for ship

officers’ education. Samuelsen and Naseri (2018) have

also revealed other limitations of the Overland model

e.g. limited correlation between wind strength and

wave height; the upper limit of ship icing caused by

high waves rarely coexists with very low

temperatures and conditions related to Polar lows.

One challenge regarding the data collection that

forms the basis for the MINCOG model is that the

data sets are only collected from one type of vessels.

Droplets from sea spray that form ship icing will

more likely hit the superstructure and upper part of

smaller vessels as a consequence of bow height and

shape. Because of this aspect, the MINCOG and other

ice rating models should be tested and verified for

several types and sizes of ships.

With reference to Figure 9, the limitation of the

Overland prediction icing rate model results in

inaccurate estimates. Bearing in mind the variability

in accuracy among the presented prediction icing rate

models is important for both weather forecasts and

MET. Using the correct model for icing rate prediction

is a prerequisite for issuing an accurate and realistic

decision supportive ship icing forecasts.

The occurrence of Polar lows in new areas of polar

waters is challenging and demands that ship officers

be aware of this aspect with reference to Figure 2.

The estimated stability reduction in the KV

Nordkapp case was within the maximum required

vertical center of gravity (KGmax), regulated by the

stability manual. If the ship had continued to sail

without action to remove ice, the KGice had exceeded

the requirement of the stability manual.

However, an ice accretion of 110 ton, like in the KV

Nordkapp case, could easily be actual for smaller

ships followed by more severe consequences.

In addition, IMO and other authorities also have

developed and implemented the ice accretion

allowance regulations discussed by Mintu et al.

(2016), see Table 5.

670

Table 5. Ice Accretion Allowance regulations discussed by

Mintu et al. (2016)

_______________________________________________

IMO Regulations IMO Polar Code (2015)

IMO Intact Stability Code (2015)

Classification ABS (2015)

Society RMRS Rules (2016)

Rules DNVGL (2015)

Lloyd`s Register (2015)

Offshore ISO 19906 for Arctic Offshore Structures

Structures (2010)

Codes NORSOK (2007)

Canadian Std. CSA S471 (2004)

_______________________________________________

For references mentioned in Table 5, see Mintu et

al. (2016).

Ship stability is affected by accretion of sea spray

icing, atmospheric icing and snow (freezing to ice).

Norsok N-003 (Standards Norway, 2017) discusses

the accumulation of sea spray ice, atmospheric icing

and snow on stationary objects. We should note that

the recommendations of Norsok N-003 are as follows:

− North of 70° north on the Norwegian Continental

Shelf, a nominal value for thickness of the

accumulated icing caused by precipitation may be

selected as 20 mm. If the nominal values are

applied, an ice density of 900 kg/m3 shall be used.

This thickness can be assumed constant from a

height of 5 m above sea level to the top of the

facilities (Norsok N-003 paragraph 6.7.3).Note that

atmospheric icing applies at all surfaces on a

vessel, in particular on masts and pipes etc.

− According to Norsok N-003 (Table A5), the

expected snow accumulation per day represent 0.8

– 1.0 kPa with an annual probability of exceedance

of 10-2 over a horizontal area of 1m2.

Using an approximate horizontal area of KV

Nordkapp of 1500 m2, the weight of the accumulated

snow using a value of e.g. 0.5 kPa, would be 750 kN

(76tf), a substantial weight.

Whether these values should be communicated to

the maritime industry as design basis for ultimate

stability checks in case of getting into a Polar low

situation, should be debated.

The STCW convention requests just an

introduction of the theme “Ship icing” for ship

officers according to the Polar Code, and the code

only applies for larger ships operating in polar

waters. This is challenging because all the highlighted

shipwreck cases show that hardly none of these ships

are regulated by this code.

The loss of 6 fishing vessels west of the coast of

Denmark in 1979 also shows the limitations of the

code related to occurrence of ship icing outside polar

waters.

9 CONCLUDING REMARKS AND

RECOMMENDATIONS

The objective of the present work was to set focus on

ship icing, and may provide a guideline or a

foundation for recommendations of measures for

preventing or limiting ship icing. The analysis

presented has revealed that the phenomenon of ship

icing is one important cause of stability-related

accidents due to the reduction of reserve buoyancy

and stability. In this respect, the important objective

for ship officers is to have competence to enable

precautionary measures to avoid such situations.

Focus on what MET schools should teach ship officer

students related to ship icing aspects is therefore a

direct consequence of the objective of this work.

Sufficiently accurate icing rate prediction models

have still to be developed and verified to obtain

realistic icing forecasts. Some icing rate models seem

to be accurate, but they may have to be customized to

various types of ships. Prediction of Polar lows is still

challenging, and this is probably the most challenging

aspect because it is a perquisite for shipmasters to

secure their sailing routes in advance to avoid icing

conditions. The KV Nordkapp case shows that even

larger ships can be affected by severe ship icing,

challenging their stability limits in such conditions.

Different technological de-icing solutions have

been developed during the last decades. Some of

these technologies have been developed for other

industries and are later adapted and identified for

marine use (Ryerson, 2011). Advanced ships have

installed different types of these technologies to

prevent ice to accumulate and for removing already

accumulated ice like e.g. the Roll to Roll CNT Coating

for Electro Thermal Heating (Rashid et al., 2018). Such

equipment is expensive, both in procurement and in

use, and it is not currently relevant for e.g. fishing

vessels, which represent most of the losses due to ship

icing.

For fishing vessels, introducing operational

measures to minimize ice accretion could be an

option. In such cases, avoiding the sea spray impact

by manoeuvring the ship for downwind heading

leads to a reduction in wave impacts with consequent

reduction in the amount of sea spray icing (Guest and

Luke 2005).

Based on the number of stability related accidents

caused by ship icing, we would recommend that

STCW should implement stricter standards for

minimum required competence for all ship officers,

related to ship icing. These standards should be

mandatory in MET schools’ curriculums.

REFERENCES

Dehghani-Sanij, A, R., Dehghani, S, R., Naterer, G, F.,

Myzychka, Y, S., 2017, “Sea spray icing phenomena on

marine vessels and offshore structures: Review and

formulation.https://doi.org/10.1016/j.oceaneng.2017.01.01

Dyrcz, C., 2019, “Phenomenon of icing and a review of ice

Accretion nomograms, models and chart in navigation”,

Scientific Journal of Polish Naval Academy, Number 4,

pp. 22, 2019.

Ekeberg, O, C., 2010, “State-of-the-art on the marine icing

models and observations”. Technical report

(confidential) 2010 - 745, DNV – Det Norske Veritas,

Høvik, Norway

EMSA, 2016, “Annual Overview of Marine Causalities and

Incidents 2016”. http://www.emsa.europa.eu/news-a-

671

press-centre/external-news/item/2903-annual-overview-

of-marine-casualties-and-incidents-2016.html

Fikke, S., Ronsten, G., Heimo, A., Kunz, S, Ostrozlik, M.,

Persson, P.-E., Sabata, J, Wareing, B, Wichure, B, Chum

J., Laakso, T., Säntti, K. and Makkonen, M., 2006. “Cost

727: atmospheric icing on structures. Measurements and

data collection on icing: state of the art”,

Veröffentlichung Meteo Schweiz Nr. 75.

https://www.wmo.int/pages/prog/www/IMOP/meetings

/Surface/ET-STMT-2/COST-727-report_MCH-V75.pdf

Good, W., Burwell, M., 2018,” Alaska Shipwrecks 1750-

2015”. ISBN 978-1-387-98114-4

Guest, P., Luke, R., 2005, “Mariners Weather Log, Vol 49,

No.3, Dec. 2005”. https://vos.noaa.gov/MWL/

dec_05/ves.shtml

Hanzu-Pazara, R. Barsan, E. Arsenie, P. Chiotoroiu, L.

Raicu, G., 2008., “Reducing of Maritime Accidents

Caused by Human Factors using simulators in training

process”. Journal of Marine Research, Vol 5 No 1, pp. 3-

18, 2008. https://www.jmr.unican.es/index.php/jmr/

article/view/32/3

Henry, N. L., 1995, “Forecasting vessel icing due to freezing

spray in Canadian east coast waters”. Part I: Model

physics. Technical report, Environment Canada,

Newfoundland Weather Centre.

Horjen, I., 1990, “Numerical modelling of time-dependent

Marine icing, anti-icing and de-icing”. PhD thesis, NTH

- Norges Tekniske Høgskole, Trondheim, Norway.

IMO, 2008. Int. Code of Intact Stability, IMO, London, UK.

IMO STCW, 2011. International Convention on Standards of

Training, Certification and Watch keeping for Seafarers,

1978, IMO, London, UK

IMO Polar Code, 2015, IMO, London. UK.

http://library.arcticportal.org/1908/1/POLAR%20CODE

%20TEXT%20AS%20ADOPTED.pdf

Johansen, K. and Batalden, B., 2018, “Active learning for

enhanced understanding of ship damage stability”.

https://www.iises.net/

Kachurin, L. G., Gashin, L. I., Smirnov, I. A., 1974, “Icing

rate of small displacement fishing boats under various

hydrometeorological conditions”. Meteorologiya I

Gidrologiya 3: 50–60. Moscow.

Kolstad, E. W., Bracegirdle, T.J., 2008, “Marine cold-air

outbreaks in the future: an assessmentof IPCC AR4

model results for the Northern Hemisphere”. Clim Dyn

30, 871–885 (2008). https://doi.org/10.1007/s00382-007-

0331-0

Kulyakhtin, A., Kulyakhtin, S., Løset, S., 2016, “The role of

the ice heat conduction in the ice growth caused by

periodic sea spray”. Cold Regions Science and

Technology 127, DOI: 10.1016/j.coldregions.2016.04.001

Lozowski, E. P., Szilder, K., Makkonen, L., 2000, “Computer

simulation of marine ice accretion”. Philosophical

Transactions of the Royal Society of London. Series A:

Mathematical, Physical and Engineering Sciences

358(1776): 2811–2845, doi:10.1098/rsta.2000.0687.

Makkonen, L., 1987, “Salinity and growth rate of ice formed

by sea spray”. Cold Regions Science and Technology

14(2): 163–171, doi: 10.1016/0165-232X (87) 90032-2,

http://www.sciencedirect.com/science/article/pii/0165232

Mintu, S., Molyneux, D., Oldford, D., 2016, “State-of-the-Art

Review of Research on ice Accretion Measurements and

Modelling”. Offshore Technology Conference, OTC-

27422-MS. https://doi.org/10.4043/27422-MS

Nordeng, T. E. and Rasmussen, E. A. “A most beautiful

polar low, 1992, A case study of a polar low

development in the, Bear Island region”. Tellus A,

44(2):81–99, 1992

Norwegian Maritime Authority, 2014, “Ulykkesdatabase

(Database of ship incidents)”.

https://www.sdir.no/sjofart/ulykker-og-sikkerhet/

Overland, J. E., Pease, C. H., Preisendorfer, R. W. and

Comisky, A.L., 1986, ”Prediction of vessel icing”. Journal

of Climate and Applied Meteorology, Vol. 25, No. 12,

pp. 1793-1806

Overland, J. E., 1990, “Prediction of Vessel Icing for Near-

Freezing Sea Temperatures”, Weather and Forecasting,

March, pp. 62-77.

PWOM, 2019, “Guidelines for development of a Polar Water

Operational Manual” Issued by the International

Chamber of Shipping. https://www.ics-

shipping.org/docs/default-source/resources/guidelines-

for-the-development-of-a-polar-water-operational-

manual.pdf?sfvrsn=4

Rashid, T., Khawaja, H., Edvartsen, K., 2018, “Measuring

Thickness of Marine Ice Using IR Thermography”. Cold

Regions Science and Technology 2018. ISSN 0165-232X.s

doi: 10.1016/j.coldregions.2018.08.025.

Ryerson, C., 2011, “Ice protection of offshore platforms”

https://doi.org/10.1016/j.coldregions.2010.02.006

Samuelsen, E. M., Løset, S., Edvartsen, K., 2015, “Marine

icing observed on KV Nordkapp during a cold air

outbreak with a developing polar low in the Barents

Sea”, Proc. of the 23rd Int. Conf. on Port and Ocean

Engineering under Arctic Conditions, Trondheim,

Norway, 2015. https://munin.uit.no/bitstream/handle/

10037/8583/article.pdf;sequence=1

Samuelsen, E.M., 2017, “Prediction of ship icing in Arctic

waters, Observations and modelling for application in

operational weather forecasting”. PhD thesis, University

of Tromsø, Norway https://hdl.handle.net/10037/11801

Samuelsen, E.M., 2018, “Ship-icing prediction methods

applied in operational weather forecasting”. Q. J. R.

Meteorol. Soc. 144: 13–33, January 2018

https://doi.org/10.1002/qj.3174

Samuelsen, E. M. and Naseri, M., 2018, “Anti-icing expected

heat loss as a risk indicator for Arctic offshore

operations”, Safety and Reliability – Safe Societies in a

Changing World, Proceedings of ESREL 2018, June 17-

21, 2018, Trondheim, Norway, pp. 1403-1409.

Sbib.dk.,1979,

http://www.sbib.dk/files/bibliotek/statistik/1979.pdf

https://library.wmo.int/doc_num.php?explnum_id=8211

Shellard, H. C., 1974, “The meteorological aspects of ice

accretion on ships”. Technical Report 10, World

Meteorological Organization, 1974. Marine Science

Affairs Report.

Stallabrass, J. R., 1980, “Trawler icing” - A compilation of

work done at N.R.C. Mechanical Engineering Report

MD-56, National Research Council Canada.

Standards Norway, 2017, “Norsok standard N003, N-003-

2017. Actions and action effects”. Standards Norway,

Lysaker, Norway.

The Seattle Times., 2017, “Memorial fund for crew lost in

Bering Sea gets help from ’Deadliest Catch’ captain”.

Electronic. Journalist Hal Bernton. Accessed 02 June

2017. http://bit.ly/2nNa9Vp.

United States Coast Guard., 2008, “Report of investigation

into the sinking and loss of four crewmembers aboard

the commercial fishing vessel Lady of Grace”. Technical

report, United States Coast Guard, Washington D.C. 29

August 2016. http://www.uscg.mil/hq/cg5/cg545/docs/

documents/