883

1 INTRODUCTION

The total volume of oil lost to the environment from

tanker spills in 2022 was approximately 15,000 tonnes;

more than 14,000 tonnes of which was lost in the three

large incidents. This figure is higher than the previous

three years but remains a fraction of the 2.95 billion

tonnes of crude oil and petroleum products that are

transported by sea each year. [11]

The tendency of number of oil spill gradually

decreased and last 10 years seems to stabilize on the

same level – average 6. See figure 1.

ITOPF has recorded tanker spill statistics over the

last 50 years and in this time, despite some annual

fluctuations, the number and volume of oil spills from

tankers has dropped dramatically. These numbers are

stabilising at a low level, with the reduction being

driven by positive change from the shipping industry

and supported by governments. Accidents, other than

shipping sources, such as pipeline spills and oil

industry activities, as well as natural seepage, also

contribute towards the global input of oil into the

marine environment. [11]

Figure 1. Quantity of Oil spill Tanker 1970-2021

The Baltic Sea is almost surrounded by land and

therefore sources of marine pollution are located

mainly on the land. Another paradox relates to

Environmental Impact of the Oil Spill Caused by the

Leakage of the Exemplary Pipeline in the South Baltic

Sea

P. Wilczyński

Gdynia Maritime University, Gdynia, Poland

ABSTRACT: A significant spillage of oil-derived cargo or fuel in the port areas causes serious threats to the

natural environment and to the ship traffic. Hydrometeorological conditions and the availability of means to

limit such spillage have a significant influence on the way the oil spill propagates. In the article, the authors

presented a simulation of the distribution of oil spills taking place in Port Polnocny in various

hydrometeorological conditions and the impact of the spill on areas located near the port. For simulation

process was used GNOME an interactive environmental simulation system designed for the rapid modelling of

pollutant trajectories in the marine environment.

http://www.transnav.eu

the International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 18

Number 4

December 2024

DOI: 10.12716/1001.18.04.14

884

shipping traffic intensity and maritime accidents. The

Baltic Sea is characterised by the large shipping traffic,

but the last decades’ data show only more than 100,

usually insignificant and minor, accidents and

incidents at the Baltic Sea every year. [7]

In the last few years, the development of port

terminals in the North Port of Gdansk – Deepwater

Container Terminal & Oil Terminal has resulted an

increase ship traffic in area of the Gulf of Gdansk.

Due to present the geopolitical situation of the

recent months, the delivery of the raw material -

crude oil for fuel production in refinery - is carried

out basically only through the Oil Terminal. This year

will be record-breaking in terms of the number of

crude oil tankers calling at the Oil Terminal as well as

transhipments and the amount of transhipped crude

oil.

Figure 2. Oil Terminal in Gdansk Port Polnocny - Transas

Simulator.

The location of the Oil Terminal and the safety of

crude oil handling operations mean that it has a

significant impact on the safety of other port terminals

as well as tourist areas located on the Vistula Spit –

Figure 2.

During cargo handling operations between the

crude oil tanker and the oil terminal, an oil spill may

occur. The weakest link in this process seems to be the

ship-to-terminal connection or the failure of the

pipeline and valve system at the oil terminal. [1]

The article assumes one of the most probable

scenarios related to the interruption of the pipeline

continuity and the leakage of the pipeline contents

after an emergency stop of handling operations.

Figure 3. Place of the incident – GNOME [2]

Figure 3 shows the location of the failure causing

the most unfavourable situation, a pipeline rupture,

which will cause leakage of the contents of practically

the entire transfer pipeline with a length of about 1

km.

2 MATERIALS AND METHODS

The amount of crude oil that could be released into

the sea depends mainly on the pressure of the

reloaded raw material and on the transhipment rate,

which can reach even 10,000 cbm/h, which gives 2,5

cbm/s. [1]

Performed 60 hours of the simulation with wind

force 4°B variable from WSW to WNW (with

uncertainty 3%), wind’s direction changed every 12

hours.

For the simulation scenario, it was assumed that an

amount of crude oil corresponding to the content of

the unprotected pipeline for 10 minutes continuously

discharging was released into the Gulf of Gdansk.

Simulations were carried out based on the

GNOME application. GNOME, version 1.3.5, is an

interactive environmental simulation system designed

for the rapid modelling of pollutant trajectories in the

marine environment. [3]

The overall design is that of modular and

integrated software. Inputs to GNOME include:

− maps of areas,

− bathymetry,

− numerical circulation models,

− location and type of the spilled oil,

− oceanographic and meteorological observations,

and

− other environmental data.

2.1 Maps of area

GNOME is not specific for any particular region and

no specific shoreline is built in. Application accepts

two different types of maps: one with shoreline data

in the form of boundary file atlas (BNA) maps (such

as those produced by GOODS, the other with grid

boundary data and bathymetry to create pseudo

three- dimensional (3-D) maps. [3]

The user must download a map via GOODS’ “Map

generator” tool or create a map to run a scenario. Each

map is rasterized into a land/water bitmap for the

purposes of tracking the oil beaching. The land/water

bitmap is of finite resolution, so it doesn’t exactly

match the map outline.

2.2 Movers in simulation area

Movers are any physics that cause movement of the

oil parcel in the water – generally currents, winds, and

diffusion. Universal movers apply everywhere, and

usually consist of wind and diffusion. To get the

overall movement the u (east-west) and v (north-

south) velocity components from currents, wind,

diffusion, and any other movers are added together at

each time-step, i, using a forward Euler scheme. The

885

movers are given a point (x, y, z, t) and return a

displacement (Δx, Δy, Δz) at t - Equation 1-3.

Calculation of zonal, meridional, and vertical

displacement by movers. [3]

cos

111120,00024

 

=

u t y

(1)

111120,00024



=

(2)

0 =z

(3)

where:

Δt = t - t1 is the time elapsed between time-steps i;

y is the latitude in radians;

111,120.00024 is the number of meters per degree of

latitude (assumes 1' latitude = 1 nautical mile

everywhere); and

Δx, Δy are the 2-D longitude and latitude

displacement, respectively, at the given depth layer z.

At present, movement in GNOME cannot occur

between depth layers (thus the vertical displacement,

Δz, is held at zero).

The calculation of total movement is a simple

vector addition of the displacement of a given

pollutant particle by each mover over the time-step.

2.2.1 Winds in simulation area

GNOME allows several different wind movers –

constant, time-dependent, and time/space dependent.

Table 1 includes constant wind speed in range 12.9-

15.3 kn (force 4°B scale) and directions which are

changed every 12 hours according to track of most

probably the low-pressure over the simulation area.

[3]

Table 1. Wind parameters during simulation.

________________________________________________

Date & time Wind speed Wind direction

[kn] [°]

________________________________________________

2022/05/29 00:00:00 12.9-15.3 247.50

2022/05/29 12:00:00 12.9-15.3 258.75

2022/05/30 00:00:00 12.9-15.3 270.00

2022/05/30 12:00:00 12.9-15.3 281.25

2022/05/31 00:00:00 12.9-15.3 292.50

________________________________________________

2.2.2 Currents in simulation area

On simulation area doesn’t exist any constant

ocean or tidal currents. In situation when the wind

direction is constant for long time wind-driven

current is created. Its parameters mainly depend on

the force and direction of the wind.

The current components can be scaled linearly by

wind speed or by the square of wind speed (i.e., wind

stress) or if no historical winds are available (or the

wind record is insufficient to satisfy the selected time

over which to average) the model can be set to

extrapolate and will use whatever wind is available

until enough data are accumulated.

Currents move the oil around and are defined on a

particular grid. GNOME recognizes finite element,

rectangular, and curvilinear grid circulation models.

For standard mode locations generally in USA,

files are used from CATS hydrodynamic model

(Current Analysis for Trajectory Simulations). The

CATS model is a 2-D depth-averaged steady-state

finite element circulation model. [3]

2.2.3 Diffusion of crude oil

For simulation purpose was used crude oil, typical

to this one which is delivered to Oil Terminal in

Gdansk. Medium Crude Oil properties contain the

Table 2 below.

Table 2. Generic Medium Crude Oil properties.

________________________________________________

Properties Value & units

________________________________________________

API 29.70

Standard Density at 15°C 878.0 [kg/m3]

Flash Point +8.60°C, (-13.00°F)

Pour Point +14.00°C, (-10.00°F)

Emulsification Constant 0.04

________________________________________________

Random spreading - diffusion, is done by a simple

random walk with a square unit probability. The

random walk is based on the diffusion value D, set in

the model which represents the horizontal eddy

diffusivity in the sea water. The D value could be

changed in range 1,000 – 1 000,000 cm2/s. In GNOME

diffusion and spreading are treated as stochastic

processes.[3]

  

=  + 





C C C

(4)

where:

C is the concentration of a material;

D are the scalar diffusion coefficients in the x and y

directions;

Diffusion can be simulated with a random walk

with any distribution, with the resulting diffusion

coefficient being one-half the variance of the

distribution of each step divided by the time-step -

Equation 5. [3]

0,5



=



(5)

For simulation Diffusion Coefficient set 200 cm²/s,

uncertainty factor 2, marked with red colours spots on

the Figures 7-10 in Chapter 3.

2.2.4 Evaporation of crude oil

Evaporation of crude oil in GNOME is not treated

by the more complete theories available. GNOME

uses a simplistic 3-phase evaporation algorithm

(Equation 6) where the pollutant is treated as a three-

component substance with independent half-lives.[4]

The pollutant type selected for the spill determines

the parameters chosen for the weathering simulation

and there is evaporation if the oil type requires. [3]

886

1 1 2 2

1 2 3

2 2 2 2 2 2

2 2 2

−

−−

− − 

− −  − − 

−

−−



   



   

− + − + −



   

   



 +  + 

i i i

i i i i i i

t t t

t t t t t t

H H H H

prob

P P P

(6)

where:

t & t1 are the time elapsed (age; in hours) at time-step i

& the previous time-step i-1, respectively;

H are the half-lives of each constituent (in hours) from

Table 3 for the pollutant;

P are the percentages of each constituent (as decimals)

from Table 3 for the pollutant.

Table 3. Crude oil pollution example type, their default

composition, and half-lives parameters. [3]

________________________________________________

Pollutant Percent each Half-Life each Observational

Type constituent Constituent Threshold Time

[h] [h]

________________________________________________

Medium 22.0 14.4 170.1

Crude Oil 26.0 48.6 170.1

52.0 10x109 170.1

________________________________________________

2.2.5 Spills of crude oil

Spilled substances are modeled as point masses

(up to 10,0004) called LE’s (Lagrangian elements) or

“splots” (from “spill dots”). Pollutants are not treated

as blobs with variable volume and thickness.

Parameters assigned to each point mass include

location (latitude/longitude), release time, age,

pollution type, and status – floating, beached,

evaporated – see Table 4.

Table 4. Information carried by each LE. [3]

________________________________________________

Parameters Units Descriptions

________________________________________________

leKey indexed LE identity

leCustomData space for custom LE data;

currently = 0

position [decimal latitude, longitude position

degrees]

z [m] depth

releaseTime time of release in spill

age [sec] time since release

clockRef [sec] time offset

pollutantType LE pollutant type for weathering

(Table 1)

mass [g] amount of pollutant in LE

density [g/cm

] pollutant density

statusCode code to indicate whether or not LE

is released, floating, beached, etc.

bVisible flag to indicate whether or not LE

is drawn

lastWaterPt last on-water location of LE before

beaching

beachTime time when LE was beached

________________________________________________

In simulation considered continuous 10 minutes

crude oil release from the one-point pipeline system

after pipeline rupture.

2.2.6 Trajectories of crude oil

Once the map, movers, and spills are set, the

model is run, and the solution is produced in the form

of trajectories. GNOME provides two solutions to an

oil spill scenario:

− a “best estimate,” or forecast, trajectory,

− a “minimum regret,” or uncertainty, trajectory,

The “best estimate” solution shows the model

result with all the input data assumed to be correct.

However, models, observations, and forecasts are

rarely perfect. Consequently, the GNOME introduced

understanding of the uncertainties (such as variations

in the wind or currents) that can occur.

This second solution a “minimum regret” allows

the model to predict other possible trajectories that are

less likely to occur, but which may have higher

associated risks.

GNOME introduced the trajectory that

incorporates these uncertainties the “minimum

regret” solution because it gives you information

about areas that could be impacted if, for example, the

wind blows from a somewhat different direction than

you have specified, or if the currents in the area flow

somewhat faster or slower than expected. In some

cases, the areas within the minimum regret solutions

might be especially valuable or sensitive to oiling. [3]

Both trajectories are represented by LE’s, which are

statistically independent pieces of the modelled

pollutant. They appear as small “pollutant particles”

on a map of simulation area when spill is running.

The “best guess” trajectory is represented by black

“splots” (from “spill dots”); the “minimum regret”

trajectory is represented by red “splots”.

2.2.7 Windage

Windage is the movement of oil by the wind. This

is typically about 3% of the wind speed based on

analytical derivation and empirical observation that

oil tends to spread out in the direction of the wind. [5]

Experience and observation have led us to use a

factor in the range 1 - 4%, adjusted based on overflight

reports.[6]

This range is used as the default in GNOME with a

uniform distribution. A given oil droplet will move

differently depending on how close it is to the wind

effects at the surface. The windage is lower as the oil

weathers and spends more time below the surface.

3 RESULTS

The results of simulation are presented below on

Figures 4 - 6 without the uncertainty factor. The oil

spot is moving according to wind direction slowly

increasing the surface due to the diffusion. Dominant

wind directions caused that the oil spot is moving

towards to the land - Vistula Spit, where are located

touristic places & beautiful beaches.

Figure 4. Oil spot moved due to the wind - GNOME. [2]

887

Figure 5. Further drift of oil spot eastward - GNOME. [2]

Figure 6. Oil spot drift to land - Vistula Spit - GNOME. [2]

The Figure 7 illustrated the trajectory of oil spot

with 4-hour interval from the place of origin the spill -

Oil Terminal to the beach located on Vistula Spit.

Figure 7. Trajectory of oil spill with interval every 4 hr -

Google Earth.

Figures 8 - 11 below presented the same trajectory

of oil spot with the uncertainty factor. Where the

black spots mean the “best guess” trajectory and the

red spots represented the “minimum regret”

trajectory.

Figure 8. Oil spot moved due to the wind with uncertainty

factor - GNOME. [2]

Figure 9. Oil spot moved with random walk - GNOME. [2]

Figure 10. Oil spot with random walk - GNOME. [2]

Figure 11. Oil spot ashore - GNOME. [2]

Figures 12 below contain the estimated dimensions

of the pollution track without the uncertainty factor,

where:

− d is diameter of pollution,

− A is area of the pollution,

− P is perimeter of the pollution,

Figure 12. Estimated dimension of the oil spill trajectory

without the uncertainty factor - Google Earth.

Figures 13 below presented the estimated

dimensions of the pollution track including the

Minimum Regret (the uncertainty factor), where:

− A is area of the pollution,

− P is perimeter of the pollution,

888

Figure 13. Estimated dimensions of the pollution including

minimum regret.

4 DISCUSSION

The possibility of presenting and predicting the

trajectory of a drifting oil spill on the sea surface, as

well as determining its size, will allow for the proper

use of emergency services to fight with oil spills.

Properly determined oil spill size requires the use

of relevant methods appropriate for the hydro-

meteorological conditions and the place of spillage.

Use the GNOME application allow the operators of

Oil Terminal develop the most likely scenarios of

events and based on them, assess the ability and

readiness of emergency services to take antipollution

action. [8]

In the case of insufficient forces and means to fight

with oil spills, investments will be necessary to ensure

the safety of the port and surrounding region.

This applies primarily to the beaches located on

Vistula spit and Natura 2000 areas where oil spillage

can reach. Collecting the oil residues will be laborious

and costly, and the region will lose its tourist

attractiveness for several years.

Successful shoreline clean-up depends on the

timely availability of personnel, equipment, and

materials and upon the quality of the organization

established to manage and conduct the operation.

Also, the type of shoreline determines the most

appropriate clean-up technique to be used. [9,10]

Future research should be focused on specialized

ship, equipment, and manpower especially on their

capacities to combat with the amount of oil, which

could be released during cargo handling in Oil

Terminal or due to ship’s accidents on fairway to

protected environment in south part of Gdansk Bay.

Az recommendation ITOPF manpower and

equipment should be identified in the local

contingency plan and regularly mobilised in practical

exercise to test their effectiveness. [9,10]

5 CONCLUSIONS

GNOME application is very useful multitool to

predict trajectory of oil spill with all factors affecting

the direction and speed of the oil spot on sea surface.

It could be used to support decision making during

planning and developing oil spill plans for the oil

terminals. Every oil terminal should ensure that all

means (ship and material) to combat with oil spill on

the sea surface is sufficient and the personnel is

properly trained.

Many years ago, only TRANSAS issued Oil Spill

Simulator, which allow to assess the trajectory of oil

spot with all assisted factors. Nowadays generally

available application GNOME looks like the properly

tools to make assessment during oil spillage. GNOME

use many additional functions which are prepared for

many locations USA together with all movers, but for

other location there are many options too to apply this

solution.

REFERENCES

[1] Blokus A., Kwiatuszewska-Sarnecka B., Wilczyński P.,

Wolny P. Crude oil transfer safety analysis and oil spills

prevention in port Oil Terminal Journal of Polish Safety

and Reliability Association Summer Safety and

Reliability Seminars, Volume 10, Number 1, 2019

[2] https://gnome.orr.noaa.gov

[3] NOAA Technical Memorandum NOS OR&R 40, General

NOAA Operational Modelling Environment (GNOME)

Technical Documentation. Seattle, Washington October

2012 Department of Commerce, National Oceanic and

Atmospheric Administration (NOAA) National Ocean

Service, Office of Response and Restoration.

[4] Boehm, P. D., Feist, D. L., Mackay, D., & Paterson, S.

(1982). Physical-Chemical Weathering of Petroleum

Hydrocarbons from the Ixtoc I Blowout: Chemical

Measurements and a Weathering Model. Environmental

Science & Technology16 (8), pp 498-505.

[5] Stolzenbach, K. D., Madsen, O. S., Adams, E. E., Pollack,

A. M., & Cooper, C. K. (1977). A Review and Evaluation

of Basic Techniques for Predicting the Behaviour of

Surface Oil Slicks. Cambridge: Rep. 22, Department of

Civil Engineering, Massachusetts Institute of

Technology.

[6] Lehr, W. J., Barker, C. H., & Simecek-Beatty, D. (1999).

New Developments in the Use of Uncertainty in Oil Spill

Forecasts. Proceedings of the 22nd Arctic & Marine Oil

spill Program (AMOP) Technical Seminar (pp. 271-844).

Ottawa, Ontario: Environment Canada.

[7] A. Dobrzycka-Krahel, M. Bogalecka The Baltic Sea under

Anthropopressure—The Sea of Paradoxes, Journal

Water, Vol. 14, Issue 22

[8] Weintrit A., Neumann T., Marine Navigation and Safety

of Sea Transportation: Advances in Marine Navigation,

pp 1-313, CRC Press, 2013.

[9] ITOPF Recognized of oil on shorelines, Technical

Information Paper 6.

[10] ITOPF Clean-up of oil from shoreline, Technical

Information Paper 7.

[11] https://www.itopf.org/news-events/news/tanker-spill-

statistics-2022/