675
1 INTRODUCTION
TheGulfofFinland,inthenorthernpartoftheBaltic
Sea,isoneofthebusiestseaareasintheworld.The
hightraffic density has been identified asincreasing
theriskofamarineoilspill,whichiswhytheFinnish
authorities responsible for oil spill
response in the
region have persistently developed their response
capacity. Under joint agreement between the Baltic
Seastates,thepreparednessisbuiltuponmechanical
recovery,andtheuseofdispersantsorsinkingagents
is avoided, and in‐situ‐burning is applied only ona
verydiscretionarybasis.[1,2.]
As
theapplicabilityofmechanicaloilspillrecovery
technologies is substance‐specific, development of
equipment capability is based on themost likely oil
spill scenarios. Until 2014, a potential oil spill was
likelytohavebeen causedbyheavyfuel oils (HFO)
and,accordingly,theauthoritieshavebeenprocuring
equipmentcustomizedfor
thesetypesofoils.
However, the implementation of the EU Sulphur
Directive in 2015 significantly changed the fuel
profilesoftheshipsoperatingintheBalticSea(Fig.1).
This raises the question of whether the achieved oil
spillresponsecapabilityisstillvalid.
Thispaperexaminestheimpactof
changesinfuel
profilesonoilspillresponseperformance,focusingon
the oil recovery capability of the Finnish side of the
GulfofFinland.Thispaperexaminestheapplicability
of conventional oil recovery equipment for the
recovery of marine distillate fuels. The term
“conventional” is used here in reference to the
equipmentoptimisedforthecollectionofHFOinan
aquatic environment. Marine distillate fuels refer to
marine diesel oil (MDO DMB) and marine gas oil
(MGO DMA). Marine gas oil consists exclusively of
Responding to Spills of Marine Distillate Fuels
J
.Halonen
South‐EasternFinlandUniversityofAppliedSciencesXamk,Kotka,Finland
ABSTRACT:ThecurrentspillresponsecapabilityinFinlandisbuilttorespondtooilspillscausedbyheavyfuel
oils and the most transported oil cargoes. However, the implementation of the Sulphur Directive in 2015
changedthefuelprofilesofthe
ships:priortothenewregulationshipsoperatingintheBalticSeamainlyused
heavyfueloil(HFO),whereasnowshipsusemarinegasoil(MGODMA)ormarinediesel(MDODMB)known
asmarinedistillatefuels.Thispaperreviewstheeffectivenessofthecurrentrecoverytechniquesinresponding
to
spillsof marinedistillatefuelsbasedontheoilrecoveryfieldtests.Theresultsindicatethatconventional
recoverytechniquesareonlypartiallyapplicabletomarinedistillatefuels,whichcallsforareassessmentofthe
marineoilspillresponsecapabilityandfurtherresearch.Theuseandavailabilityoflow‐carbonmarinefuels
will continue to increase as emission regulations become more stringent. This will require a continuous
assessmentoftheoilrecoverycapabilitiesandtheadaptationofspillresponsepreparednessaccordingly.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 17
Number 3
September 2023
DOI:10.12716/1001.17.03.19
676
distillates, while marine diesel oil is a distillate fuel
thatmaycontaintracesofresidualoil.
Figure1.Estimatedquantityofmarinefuelsusedbyships
operating in the Baltic Sea between 2006 and 2020. The
uppermost part of each columnshows the share of heavy
fuel oil (HFO) and the lower parts establish the shares of
distillatefuels(MDO,MGO)andLNG.[3]
No previous research has been conducted on the
recoverabilityofmarinedistillatefuelsandthereare
no incident case reports dealing with the response
methods used. Previous studies have examined the
environmental impact and toxicity of marine diesel
oils, but none of them have addressed how they
should be recovered. Instead,
the challenges of
responding to Low Sulphur Fuel Oils (LSFOs) have
been increasingly studied following recent oil spills,
suchastheMVWakashiogroundinginMauritiusin
2020. However, the properties of these fuels differ
fromdistillates tosuchanextent thattheresults are
not directly applicable. It was
therefore decided to
carry out small‐scale tests to demonstrate the
recoverabilityofmarinedistillatesandtoseeifmore
comprehensiveresearchisneeded.
This paper is structured as follows. First, the
mechanicaloilrecoverymethodsusedinthetestsare
presented, followed by the experimental tests, their
set‐upand
results.Finally,theresultsarediscussedin
thelight ofgeneral research on mechanical recovery
ofoilspills.
2 MEANSOFMECHANICALOILRECOVERY
Mechanicalrecoveryisoneofthemethodstoremove
spilledoilfromwater.Mechanicalremovalisusually
carriedoutbyusingskimmers,themostcommonof
which are of oleophilic type. With oleophilic
skimmers,oilrecoveryisbasedonoiladheringtoan
oleophilic surface of a rotating part of the skimmer,
which can be in the form of a brush, drum or disc
module. [4, 5, 6, 7, 8.] Disc skimmers are mainly
available as portable
units, but brush skimmers are
available both as portable skimmers and as vessel‐
integrated recovery systems. These skimmers vary
greatlyinsizeandcapacity,buttheprincipleoftheoil
recoveryisthesame.
Oleophilic skimmers typically collect very little
watercomparedtotheamountofoilrecovered.This
means
that their oil‐to‐water recovery ratio, also
referredasrecoveryefficiency(RE),isgenerallyhigh,
althoughtheirsuitabilityfordifferentoiltypesvaries
dependingontheskimmerdesignandthetypeofthe
oleophilicsurface[7].
The surface is usually made of steel, aluminium,
fabricorplastic,suchaspolypropylene
andpolyvinyl
chloride[7,8].Theadhesivesurfaceisacriticalfactor
in the oil recovery [9], as the surface material can
affecttherecoveryratebyupto20%[10].Inaddition,
theimmersed area of the rotating surfaceaffectsthe
amount of water entrained: the larger the disc
diameter,the greater the proportion of the adhesion
surfacerotatinginthewaterundertheoillayer[11].
This applies in reverse when the oil layer is or gets
thin,whichhighlightstheimportanceofusinganoil
boom to control the oil layer thickness:this leads to
increasedoil‐wetted
areaofrotatingsurfacereplacing
the equivalent water‐wetted area [11] and results in
higheroilrecoveryefficiency.
Adhesionalso depends on thetypeof oil andits
properties at the time of recovery [5, 6, 8]. These
properties change over time as the oil weathers,
requiring continuous evaluation of
the skimmer
performanceduringtheoperation.Asoilweathersit
becomes more viscous due to the evaporation of
volatile compounds and, in some cases, the
accumulationof water. This water accumulationisa
processknownasemulsification.[5,7,12.]
Increased viscosity enhances the oil recovery to
someextent,afterthat
itmaybecomealimitingfactor
[10,11,12].Inadditiontoviscosity,oilemulsification
can also reduce the effectiveness of oleophilic
skimmers as the high percentage of water inthe oil
preventstheoilfromadheringtotheskimmersurface
[6,8,13].Highviscosityoremulsificationcanfurther
prevent oil from flowing towards the skimmer,
requiringthe oil to be pushed orthe skimmer to be
movedtoprovidecontinuedrecovery[6].
Despitetheselimitations,mechanicalrecoveryhas
beenidentifiedasthemainandmostviablemethodin
the Baltic Sea, as other response options may have
adverseand
unpredictableeffects[1].
Sorbents are used as a secondary recovery
technique to remove the remaining oil film after
mechanical recovery. There are several types of
sorbents,typicallydividedintooil‐onlyandchemical‐
onlyproducts.They alsovaryin physicalshapeand
size. Sorbents usually work by either absorption or
adsorption [4, 5, 12, 14]. They are generally more
effectiveonlighterthanheavier oilproducts[12,14]
but there is limited research on their suitability for
marinedistillatefuels.
As no single recovery method is suitable for all
situations[4,6],itmaybenecessarytobeabletouse
different recovery means and to select the most
appropriate type of skimmer or sorbent for a given
typeof oil and operating conditions at the spill site,
andtoadapttheselectionasnecessaryateachstageof
the response operation as oil characteristics and
conditions change over time [4,
5, 10]. However,
keeping up‐to‐date knowledge of the appropriate
methodsisachallengebecauseoftherapid renewal
anddiversificationofpotentialspillrisksubstances.
677
3 EXPERIMENTALTESTSOFMECHANICAL
RECOVERYOFDISTILLATEFUELS
Thepurposeofthetestswastoexaminethecapability
of the conventional skimmers to recover marine
distillatefuels,and,bymeasuringtheratiosofoilto
water in the total amount of collected fluid, to
comparerecoveryefficienciesofused
skimmers.The
tests were performed with marine diesel oil (MDO
DMB) and Neste Light Fuel Oil (LFO) that is
practically the same fuel as Neste Marine Gas Oil
(MGODMA)withcommonCASnumberandSafety
DataSheet[15].
The experimental tests were carried out at the
outdoor Oil Spill
Response Testing Facility of the
South‐EasternFinlandUniversityofAppliedSciences
(Xamk) in Kotka. The test facility consists of a main
basinwithadiameterof29metresandawaterdepth
of 2–3 metres, and several smaller test tanks. These
testswerecarriedoutinoneofthesmaller
tanksof12
square metres with a maximum water depth of one
metre.
The tests were carried outduringthree days, the
17th,24thand28thofSeptember2022,attheaverage
temperatureof17–23°C(SeeTable1).Thewaterused
was fresh water, originally pumped from a nearby
river,
withawatertemperaturesettledtotheambient
temperature. The difference in salinity between the
river water and the surface water of the Gulf of
Finland brackish water was not expected to have a
significanteffectontheperformanceoftheskimmers.
Table1.Baselinedata.
________________________________________________
Experi‐ TypeofTest Oil ViscosityDensityTemp. Date
ment skimmeroil layer [mm²/s [kg/dm³ [˚C]
module [mm] in40˚C] in15˚C]
________________________________________________
1 Brush MDO10 2...11 0,9 22,5 17.8.22
module DMB
2 Disc MDO10 2…11 0,9 22,5 17.8.22
module DMB
3 Brush Light10 4,5 0,8…0,8518,0 24.8.22
module FuelOil
4 Disc Light10 4,5 0,8…0,8523,0 24.8.22
module FuelOil
5 Brush MDO10 2…11
0,9 17,0 26.8.22
module DMB
________________________________________________
Alltestswere performedusingportableMinimax
12skimmerframewithbrushanddiscmodules, the
electric‐driven Power Pack LPP7.5VXE, hydraulic
linesandhosessuppliedfromthemanufacturer.This
skimmer chosen is the most common skimmer type
usedbytheFinnishrescueservices.
Anareawasboomed atthe surface
of the water‐
filled test tank, in which the available oil volume
formeda10mmlayerofoil.Thesizeoftheareawas
definedtakingintoaccountthatitwouldnotrestrict
the operation of the skimmer. In the first tests
(experiments 1 and 2) the skimmers were
operated
simultaneously (set‐up presented in Fig. 2), but in
subsequent tests, the skimmers were operated one
aftertheother.
The recovered oil was transferred into two 300‐
litre IBC containers, the walls of which were
transparent enough that the accumulation of fluids
wasvisiblefromtheoutside.Therecoveryresults
of
the skimmers were directed into separate storage
containers for comparison (Fig. 3 and 4). Although
estimatingtheliquid fractionsfromthetank column
heights does not have a very high accuracy, it was
consideredappropriateandsufficientforthepurpose.
Figure2. Test tank setup used for skimmer performance
experiments, with a brush skimmer on the left and a disc
skimmerontheright.Photo:J.Halonen.
Figure3. Recovered marine diesel oil (MDO DMB) and
entrained water (free water and emulsion) separated by
gravityinstoragecontainers.Onthelefttherecoveryresult
of the brush skimmer and on the right the corresponding
resultofthediscskimmer.Photos:J.Halonen.
Figure4. Recoveredlight fuel oil andentrained free water
separated by gravity in storage containers. On the left the
recovery result of the brush skimmer, on the right the
corresponding result of the disc skimmer. Photos: J.
Halonen.
Asfarasitwaspractical,thetestprotocolfollowed
the general principles of Standard Test Method for
DeterminingaMeasuredNameplateRecoveryRateof
Stationary Oil Skimmer Systems (F2709‐18) of the
ASTMInternational.Thestandarddefinesacriterion
toquantifythe performanceofa stationaryskimmer
in ideal recovery
conditions allowing a skimmer to
operate at its maximum possible recovery rate [16].
Our test, in contrast, aimed to achieve realistic
recovery efficiency values by mimicking authentic
spillincidentrecoveryconditionsasmuchaspossible.
For this reason, the standard was applied when
appropriate. For example, the initial oil layer
thicknesswassetthinner(10mm)thandefinedbythe
678
standard (>75 mm), and the layer thickness was
allowedtodeclineastherecoveryprogressed.
The skimmers were tested according to the
followingprocedure:
Aknownamountofoilwasaddedtotheboomed
areaofthewater‐filledtesttank.Thevolumewas
calculated to form a 10
mm thick oil layer. The
thicknesswasmeasured.
Theskimmerwasliftedintothetankanditsfree‐
floatingpositioninthemiddleoftheboomedarea
wassecuredwitharope.
Therotationspeedoftheskimmerwasoptimised
bythemeansofvisualobservationtominimise
the
volume of entrained free water. This means that
the water content of the recovered fluid was
assessed at the point where the adhered fluid is
scraped by skimmer’s plastic blade to a recovery
sump. If the rotation speed is too high, water
droplets start to appear in the fluid. As
the
thickness of the oil decreased and the amount of
entrained water began to increase, the rotation
speed was adjusted accordingly to achieve the
optimumrecoveryresult.
When the skimmer could no longer take hold of
theoil,themachineswerestopped.Theremaining
oil was then absorbed by
means of oil‐only
absorbents. The absorption capacity of the
absorbentswasobserved,notmeasured.
Atthe end ofeachtest, the total volume of fluid
(oil, water, water‐in‐oil emulsion) in the storage
containerwasmeasured,andaftertheliquidswere
separatedintotheirownphases,theproportion
of
eachwasassessed.
The main parameter examined was the oil
recoveryefficiency(RE).Recoveryefficiencymeasures
theselectivityoftheskimmer,describingtheabilityof
the skimmer to recover oil in preference to water.
Recovery efficiency is expressed as the ratio of the
quantity of the oil recovered to
the totalquantityof
fluid(oilandwaterandtheiremulsion) collected[6,
16]:
100
oil
total fluid
V
RE
V
where
RE=recoveryefficiency,%
V
oil=volumeofoilrecovered
V
totalfluid=volumeoftotalfluidrecovered
During the tests, the skimmers were operated as
appropriatetoavoiddeliberateunderperformance.As
it is known, the oil recovery rate increases with the
rotational speed up to some extent [11], but the
recovery efficiency suffers: a higher rotational speed
will cause a higher
amount of free water to be
entrained,particularlywhentheoillayerisrelatively
thin.Highrotationspeedwillalsoemulsifytheoilto
a greater extent. [10.] Therefore, the rotation speed
was continually adjusted to maintain optimal
recovery. When the amount of oil in the test tank
begantodecrease,
oilwasdivertedtotheskimmerif
it was found not to flow on its own. This assisted
feeding of skimmer was carried out with the floor
squeegees(Fig.9).
Aftermechanicalrecovery,the restoftheoilwas
removed with absorbents. Their effectiveness was
observed but not measured, as
it was seen that the
amountofoilremainingwasnotsufficienttoachieve
maximumabsorptioncapacity.
At the end of each test, the total amount of
recovered fluid was measured. The recovered fluid
waslettoseparateintolayersbygravityforonehour
after which the proportions of oil,
water‐in‐oil
emulsion and free water was estimated. Hour was
judged to be a suitable duration, as retention times
generally considered adequate range from 15–30
minutesforlightoilsto60minutesforheavyoils[17].
The fluid was left to settle further in the containers
fromoneday
tooneweekbutnosignificantchanges
inthedegreeofseparationwereobserved[18].
3.1 Impactofskimmertypeonthecompositionofthe
recoveredfluid
The results of all five test experiments on the
composition of the recovered fluids are shown in
Figure 5. The composition indicates the recovery
efficiency,i.e.theamountofoilrecoveredinrelation
tothetotalamountofrecoveredfluids.
Acomparisonofthecollectedfluidsshowsaclear
difference between the skimmer types.Based onthe
results, the disc skimmer attained high recovery
efficiencyforbothmarinedieseloilandlightfueloil
(87–92%). With the brush skimmer, the water
entrainmentwassignificant,resultinginlowrecovery
efficiency(8–14%).
Figure5. Compositionof total recovered fluid grouped by
oiltype.Comparingrecoveryresultsoftwoskimmertypes
recovering two distillate fuels under equal recovery
conditions. Tests at 17–23°C with the initial oil layer
thickness of 10 millimetres. The total fluid includes the
liquids recovered by the skimmers; the absorbed liquid
is
excluded.
When recovering MDO, the total water
entrainmentwas+38%greater thaninLFO recovery
(Fig. 5), and the proportions of free water and
emulsionwerequiteequal,38%and62%respectively
(Fig. 6). Conversely, the water entrained in the
recovery of LFO consisted almost entirely of free
water. This occurred, because
the LFO emulsion
produced by the brush skimmer was not stable in
naturebutbrokedown,appearingasfreewaterinthe
finalrecoveryresult.
679
ThehighwaterentrainmentintheMDOrecovery
mainly reflects the low selectivity of the brush
skimmer (Fig. 5). The disc skimmer also took up
slightly more water when recovering MDO (13%)
than when recovering LFO (8%), see Fig. 5. This is
expected to be due to the higher emulsification
tendencyofMDO,asthewaterentrainedwasmainly
informofemulsion(Fig.6).However,thedifference
between the amounts of free water and emulsion in
the disc skimmer’s recovery results is not very
significant,andtheactualquantitieswereverysmall
(Fig.7).
Figure6showsthecomposition
ofthetotalwater
entrainmentdivided intofractions of free water and
emulsion. The amount of the recovered oil itself is
excludedfromthiscomparison.
Figure6. Distribution of total water entrainment into
fractions of free water and emulsion, grouped by the oil
type. Proportion of theoilitselfis excluded.Last columns
peroiltyperepresenttheaveragevalues.
3.2 Impactofskimmertypeonthetotalvolumeofthe
recoveredfluid
Water entrainment affected the total volume of
recovered fluid and the skimmers differed
considerablyinproducingthisvolume:usingabrush
skimmer, the total fluid volume increased by 480–
540% (up to approximately six times the initial oil
volume), while using a disc skimmer the total fluid
volume increased by 5–6%, when all the collected
fluidsweretaken intoaccount.Thetotalvolumes of
fluid relative to the amounts of oil initially spilt is
presentedinFig.7.
On average, entrained free water had a greater
effect on the
total volume of recovered fluid than
emulsification(Fig.5and7).
Figure7. Total recovered fluid volume times the initial
volume of oil discharged, grouped by skimmer type, and
presentingthecompositionofthetotalfluid.
Noneoftheskimmersachieved100%oilremoval,
but the disc skimmer came very close by removing
97% of the light fuel oil. The brush skimmer also
managedtoremove83%ofthelightfueloilalthough
theby‐productwashighproportionsofexcesswater
(Fig.7).
Withbothtypes
ofoils,thediscskimmerleftathin
oilfilmonthewatersurface.Theoilremainingafter
thebrush skimmer recoverywas either afilm(LFO)
oralayerofwater‐in‐oilemulsion(MDO).
3.3 Impactofskimmertypeontheemulsification
The results regarding emulsion formation differ
betweenthetwotypesofskimmers.Thediscskimmer
produced no emulsion (0%) when recovering LFO,
and11%emulsionwhenrecoveringMDO.Whenthe
brushskimmerwasused,thepercentageofemulsions
intheoverallrecoveryresultrangedfrom37%to55%
for MDO but remained below 1% for LFO as
the
emulsionformedwaslargelybrokendown(Fig.5).
As can be seen from Fig. 6, which shows the
proportions of free and emulsified fractions in total
water entrainment, emulsification occurred mainly
during the MDO recovery. There was a clear
differencebetweentheoilstested:MDOhadamuch
higher
tendencytoemulsify.
It was also observed that the movement of the
rotatingbrushesalonegeneratedenergysufficientfor
emulsification(Fig.8)eventhoughtherotationspeed
was kept minimal. Also with the LFO, the skimmer
itselfgenerated emulsionduring the recovery,but it
brokedownvery quickly(inminutes). It
took much
longerforMDOtosettle, andthe emulsionwasstill
clearlypresentafteraweek.
Figure8. Emulsion formation during a brush skimmer
recovery.Skimmerwithabrushmodulerecoveringmarine
680
dieseloil(MDODMB)ontheleftandlightfueloilonthe
right.Photos:J.HalonenandM.Kettunen.
In addition to the mixing caused by the rotating
brushes, the entrained water passing through the
pump and hoses may have contributed to the
emulsification.Itshouldthereforebenotedthatitwas
not possible to assess to what extent the emulsions
were due to brushes and to what extent to
other
contributing factors such as pump‐induced mixing
and turbulence in the discharge hoses, or pressure
pulses and the resulting splashing in the storage
containers.
Thediscskimmerproducedthelowestvolumeof
entrained water in terms of quantity. When the
composition of the water entrainment in MDO
recovery was
compared (Fig. 6), it was found thata
higher proportion of the water was in form of
emulsion (86%) when using a disc skimmer than
whenusingabrush skimmer(41%–60%),suggesting
thatthe adhesive material of thediscsmay be more
hydrophobic.
Bothoils,intheir originalform, weredrawn
into
theskimmerbythepulloftherotatingunit,whilethe
emulsified oils did not flow towards the skimmer
withoutintervention.
Mechanical recovery was continued until the
skimmers could no longer capture the oil. The
remainingoilsconsistedofthinoilfilmsorlayersof
emulsiondependingon the
typeofoiland skimmer
used.First,theoilwascontainedintoasmallerarea
withanabsorbent boom.If theabsorbentboomwas
not sufficient to remove the oil, the remainder was
absorbedwithoil‐only‐sheets.Theabsorbentboom,as
wellasthe sheets, wereeffectivefor LFO. However,
theydidnotwork wellontheemulsifiedMDO.The
oilonlycoatedthesurfaceoftheabsorbentboomand
did not penetrate the material. The absorbent sheets
turned out to be ineffective with the emulsion also.
Thecapabilityofthesheetstoremovetheemulsified
MDO was found to
be based on adsorption rather
thanabsorption.
4 DISCUSSION
Thepurposeoftheexperimentaltestswastoexamine
the capability of two most common types of
conventional skimmers to recover marine distillate
fuels. Recovery efficiency was evaluated by
measuring the ratios of oil to the total amount of
recovered fluid and
comparing the proportions of
emulsified and free water in the total water
entrainment. The oils used were marine diesel oil
(MDO DMB) and light fuel oil, which corresponds
with marine gas oil (MGO DMA). In order to
eliminate the variables introduced, the tests were
conducted with the same skimmer frame
in equal
recoveryconditions.
Although the number of tests carried out was
ratherlimited,andtheresultsmayhavebeensubject
tosomeinaccuracy,asdescribedearlier,theobjective
of assessing the need for further research was
achieved. Some conclusions can also be drawn. The
obtainedresultsindicatethattheskimmer
typehasa
noticeableeffectontheamountofentrainedwateras
wellasonthedegreeofemulsification,andhenceon
thetotalamountofrecoveredfluids.
4.1 Theimpactofemulsificationonrecoveryefficiency
Emulsification is a process by which water droplets
aredispersedintooil.Emulsificationhas
animpacton
the overall recovery operation as entraining water
changesbothdensityandviscosityoftheoil,aswell
ascausesthetotalvolumeofthefluidtoincrease[5,7,
12]. Emulsification is usually associated with a
weathering process and requires external mixing
energy,suchaswaveaction[4,
5,7].Itwastherefore
interesting to note that, in the case of marine
distillates, the rotation of the adhesion surface itself
causedemulsification.
Inthetests,emulsionformationonaveragehada
smallereffectontheincreasedtotalvolumeoffluids
thantheamountoffreewater,butemulsification
led
to a reduction in recoverability. In particularly
oleophilicskimmersareknowntolosetheirefficiency
if the oil emulsifies [6, 7, 13]. Although the brush
skimmer was more problematic in terms of
emulsification,andlittleto no emulsionwas formed
duetothediscskimmer,itisknownthatalso
thedisc
typebecomesineffectiveinsituationswhereawater‐
in‐oil emulsion has already formed. This is mainly
because emulsions can be almost non‐adhesive and,
with high viscous emulsions, the discs are cutting
through the emulsion instead of recovering it [5, 6].
Higher viscosity in general leads to
a slower
spreading rate and results in reduced access to the
skimmer’sadhesionsurface[10].
The inefficiency of disc skimmers in recovering
water‐in‐oil emulsions was also found in the field
tests carried out in Norway. These tests used an
emulsion made from fuels and an emulsifier mixed
with seawater.
Their further studies revealed that
adding an emulsifier reduces the interfacial tension,
whichsignificantlydecreasestheadhesiveproperties
oftheemulsion.[8.]Itisrecognisedhoweverthatalso
theemulsionitselfhaspooradhesionpropertiesand
is therefore difficult to recover with oleophilic
skimmers[6,7,13].On theother
hand,accordingto
Kystverket [19], the emulsion used in their tests
adhered well to the adhesion surface but was
hinderedbyalayerofwaterwhichappearedbetween
theemulsionandtheskimmerpreventingcontinuous
recovery.
Other studies [6, 19] support the finding that
emulsions are less likely to move towards
the
skimmercausingdiscontinuousoilflow.Duetopoor
natural flow of the emulsions, feeding the skimmer
neededtobeassistedwithfloorsqueegees(Fig.9).It
wasfoundoutlater,thatduringtheKystverket’stests,
almostasimilarsolutionwasapplied,namelypaddles
[19].Itispossiblethatthe
assistedfeedingmayhave
contributedto thewater‐uptake,butthe effectcould
notbedistinguishednormeasured.
681
Figure9.Assistanceisneededtofeedemulsifiedoilontothe
skimmer. Photo taken during a training session in which
MDODMBwasrecoveredwithMimimax25skimmerand
feeding was assisted with floor squeegees. Photo: J.
Halonen.
Theexperiencewithassistedfeedingandprevious
studies showed that although the recovery of
emulsified oils is difficult it is not completely
impossiblebut requireseffort andtime – in practice
manual labour – but the problem of excess water
remains unsolved. Emulsified oil can be recovered
whenitisphysically
pushedtowardstheskimmeror
when there is a physical barrier against which the
emulsionisdriven.Thischallengemayonlyapplyto
stationary recovery, as in offshore operations,
recovery vessel and its skimmer being in constant
motion in relation to oil may force the oil onto the
skimmer. [19.]
It therefore seems that dynamic
recoverysystemswouldbepreferableforrecovering
emulsion‐prone oils, although further research is
needed to ensure compatibility. In addition, the
possibilitiesofferedbydecantingshouldbeexplored.
However,decantingcanonlybeappliedtoemulsions
thatbreakdowninarelativelyshorttime.
Theemulsions
formedinthetestsweredifferentin
nature. Based on the categorization of Fingas and
Fieldhouse[20],theemulsionformedwiththeMDO
representedastableormesostableemulsion,andLFO
formedanunstableoil‐watermixture.Thus,theuseof
decanting seems a viable option to facilitate the on‐
siterecoveryforLFOonly.
4.2 Theimportanceofskimmertypeselection
Both conventional skimmer types were capable of
removingdistillatesfromthewater,buttherecovery
results varied. This means that, in the event of an
incident,theremovalofoilispossiblewithcommonly
available equipment, but different levels
of
intervention,storagecapacityandtimeareneededto
manage the quantities recovered to complete the
responseoperation.
There are considerable differences between the
skimmer results obtained in terms of efficiency and
totalfluid.Ofthetypesofskimmerstested,thebrush
skimmer appeared to be less efficient. Further, the
combinationofthebrushskimmerandMDOproved
to be the least favourable in terms of emulsification
due to the high volume and stable nature of the
emulsion formed. This finding is significant in
relation to the skimmers commonly available in
Finland.Basedonasmall‐scalesurveycarriedoutin
March 2023 in the context of this study, 80% of all
skimmersoperatedbytherescueservicesonthecoast
of the Gulf of Finland are of the brush type. More
precisely, 65% of stationary skimmers are of the
brush‐typeandtheremaining35%ofthediscorweir
type of skimmers, while 100% of the ship mounted
recoverysystemsareofthebrushtype.
Therecoveryefficiencyofthebrushskimmerwas
relatively similar for both light fuel oil (LFO/MGO
DMA)andmarinedieseloil(MDODMB).Theuseof
thebrushskimmers resultedinhigh totalvolume of
therecoveredfluidthatreacheditsgreatestvalue,six
times the initial oil volume, during the MDO
recovery.Inadditiontothehighamountofentrained
free water, the skimmer’s tendency to produce an
emulsion affected the total volumes recovered.
Emulsification also affected the recovery itself,
impairingtheeffectiveuse
ofsorbentsandrequiring
manual intervention. Consequently, emulsified oils
seem more time‐consuming to recover potentially
delaying the recovery process and resulting in a
prolongedimpacttimeontheenvironment.
Thetotalvolumeoffluidisanimportantfactoras
storage capacity is usually limited. Both temporary
storage capacity onboard and
adequate intermediate
storing ashore can be potential bottlenecks for a
timely and continuous spill response [12, 21]. The
increase in the total fluid volume will also cause
logistical concerns and raise the costs related to
transportanddisposalofrecoveredfluids[22].
Managing the total volume of recovered fluid is
particularly important inthe Gulf of Finland, where
anoilspillisestimatedtogeneratearelativelylarge
amountofoilywastesduetothecharacteristicsofthe
operatingenvironmentalone[23].Toprepareforthis,
the Finnish national recommendations [24] direct to
scaletheoilspillpreparednessandstoragecapacity
to
1.5 times the expected spill volume. Now that the
risksassociatedwithmarineoilspillshavechangedto
involvemarinedistillates,thiswouldseemachievable
onlybyusingdisctypeofskimmers.
It should be noted that the recovery volumes of
thesetestsmaynotbedirectlyscalableto
thesizeof
anoccurringspillevent,sincethethinnertheoillayer
becomesthegreatertheexpectedvolumeofentrained
water [10]. In contrast to the tests, in an actual
incident situation the recovery phase with a thin oil
layercanberelativelyshortcomparedtotheduration
of the
entire recovery operation. Furthermore, the
effect of evaporation becomes more apparent in
longer‐termoperations.Ontheotherhand,theresults
obtained under controlled test‐tank‐conditions may
notnecessarilyreflecttheoperatingconditionsatsea,
as most skimmers operate at lower efficiency in
moderatetoheavyseasorin
thepresenceoffloating
debrisorice.
Nevertheless,thetestresultsgiveanindicationof
thesignificanceoftheskimmertypeselectionandthe
findings should be considered in contingency
planning. Response plans will need to be updated
more frequently in the future as the risk base for
preparedness evolves and
new substances posing a
potential spill risk are introduced at an increasing
pace. Contingency planning could be supported by
the use of probabilistic models analysing response
efficiency[see 21, 25, 26and 27] by updatingmodel
682
parameters, such as variables related to recovery
capabilities.As Parviainenet al.[28] point out,even
thoughthesemodelsarecurrentlyinstrumentaltools
thatsupportpredefinedpoliciesratherthanexploring
alternative response options, they could be applied
more comprehensively. A systematic use of the
models,supportedbyup‐to‐date
valuesfromthefield
tests, would facilitate both the assessment of the
widerconsequencesoftheselectedmeasuresandthe
regularevaluationrequiredbyHelcomofwhetherthe
operational capacity corresponds the total spill
volumesexpected[2].
Theimportanceofthereassessmentsisunderlined
bythe fact thatalthoughdischarges
of oiland other
harmfulsubstancesintotheBalticSeahavedecreased
over the last 20 years, spills are still being detected
and, more importantly, are increasingly caused by
non‐mineral substances [29]. Optimal recovery of
these substances requires further research, as this
studyonly consideredcommonly usedmarinefuels.
Additionally,
the scope of this paper was limited to
theperformanceofstationaryskimmers,focusingon
twotypesofoleophilicskimmerstypicalfortheBaltic
Searegion. Further studiestestingthe characteristics
of other types of skimmers as well as onboard
recovery systems aretherefore necessary and would
benefitfrombeing
extendedtocovernotonlyawider
range of substances spilt but also the effects of
recoveryratesandpotentialdecantingsolutions.
5 CONCLUSIONS
This paper presented the oil spill response tests
investigating the recovery of marine distillate fuels.
The results indicate that conventional recovery
equipment are only partially applicable to
marine
distillates and thus the capability to respond to
marine oil spills needs to be reassessed and further
researchisneeded.
ACKNOWLEDGEMENTS
AuthorwouldliketothankhercolleaguesatXamk,Mr.A.
Myrén and Mr. M. Kettunen, for their contribution to the
technical implementation of the tests and Neste Oyj for
supplyingthemarinedieseloilfortestingpurposes.
The Oil Spill Response Testing Facility was established in
2019–2022 with the main
funding from the European
RegionalDevelopmentFundtroughtheRegionalCouncilof
Kymenlaakso(projectnoA75152),andthesetestsprovided
comparative data for the studies of the project Response
DemonstrationAreasforSpillsofRenewableandBio‐Based
Liquids funded by the same main financier (project no
A78380).
REFERENCES
[1]Helcom Recommendation 22/2. Restricted use of
chemicalagentsandothernon‐mechanicalmeansinoil
combattingoperationsintheBalticSeaarea.Adopted21
March2001.
[2]Helcom Recommendation 19/17. Measures in order to
combatpollutionfromoffshoreunits.Adopted24March
1998.
[3]Jalkanen J.‐P., Majamäki, E. &
Johansson, L. 2020.
Emissions from Baltic Sea shipping in 2006–2020.
Maritime Working Group Meeting document
MARITIME 21‐2021 of Baltic Marine Environment
ProtectionCommission.
[4]IMO 2005. Manual on Oil Pollution. Combating Oil
Spills. Section IV. International Maritime Organization,
London.
[5]ITOPF 2012. Response to Marine Oil Spills. Second
edition.
The International Tanker Owner’s Pollution
FederationLimited.
[6]ITOPF2014.Use ofskimmersinoilpollutionresponse.
Technical information paper, no. 5. The International
TankerOwnersPollutionFederationLimited.
[7]Fingas, M. 2013. The Basics of Oil Spill Cleanup. CRS
Press.
[8]Farooq,U.,Taban,I.&&Daling,P.2018.
Studyoftheoil
interactiontowardsoilspillrecoveryskimmermaterial:
Effect of the oil weathering and emulsification
properties. Marine Pollution Bulletin, 135 (2018), 119–
128.
[9]Hollebone, B.P. 2015. Oil Physical Properties:
Measurement and Correlation. Handbook of Oil Spill
ScienceandTechnology.Fingas,M.(ed.)JohnWiley&
Sons,Inc.,
39–50.
[10]Broje, V. & Keller, A. 2007. Effect of operational
parameters on the recovery rate of an oleophilic drum
skimmer. Journal of Hazardous Materials, vol. 148
(2007),136–143.
[11]El‐Zahaby, AM., Kabecl, A., Bakry, A. & Khaira, A.
2008. Investigation of Rotating Disk Skimmer
HydrodynamicPerformanceduringOil
SpillsRecovery.
MansouraEngineeringJournal,vol.33.no.1.
[12]Koops, W., Zeinstra, M. & Heins, S. 2014. Oil Spill
ResponseManual.NHLUniversityofAppliedSciences.
[13]IPIECA‐IOGP 2015. At‐sea containment and recovery.
Goodpracticeguidelinesforincidentmanagementand
emergencyresponsepersonnel.IOGPReport522.
[14]IMO
2016. Use of Sorbents for Spill Response. An
Operational Guide. IMO Publication. International
MaritimeOrganization,London.
[15]Engman, A. 2023.Techical Account Manager, Neste
Corporation.Writtennotice6.2.2023.
[16]ASTM International. Standard Test Method for
Determining a Measured Nameplate Recovery Rate of
StationaryOilSkimmerSystems(F2709‐18).
[17]IPIECA‐IOGP
2013. The use of decanting during
offshore oil spill recovery operations. Oil & Gas
Producers and International Petroleum Industry
EnvironmentalConservationAssociation.
[18]Halonen, J. & Kettunen, M. 2022. Keräystuoton
ennakointi öljyntorjunnassa. Xamk READ, Research,
EducationandRegionalDevelopment,vol.3/2022.
[19]Kystverket2022.WorkPackage4.Task4.1Mechanical
Recovery,
ReportsfromIMAROS.
[20]Fingas, M. & Fieldhouse, B. 2015. Water‐in‐oil
emulsions:Formation and Prediction.HandbookofOil
Spill Science and Technology. Fingas, M. (ed.) John
Wiley&Sons,Inc.,225–270.
[21]Lehikoinen,A.,Luoma,E.,Mäntyniemi,S.&Kuikka,S.
2013. Optimizing the recovery efficiency of Finnish
oil
combatingvesselsintheGulfofFinlandusingBayesian
networks.EnvironmentalScience&Technology,vol.47
(4),1792–1799.
[22]Etkin,D.2015.Riskanalysisandprevention.Handbook
of Oil Spill Science and Technology. Fingas, M. (ed.)
JohnWiley&Sons,Inc.,15–35.
[23]Halonen,J.,Altarriba,E.&Kuosa,M.
2021.Toolsforoil
spill response waste management and logistic support.
A field exercise testing the RFID technology and QR
codes.IOSC2021ConferenceProceedings.
[24]Hietala, M. & Lampela, K. 2007. Öljyntorjuntavalmius
merellä‐työryhmän loppuraportti. Suomen ympäristö
41/2007.FinnishEnvironmentInstitute,Helsinki.
683
[25]Aps, R., Sawano, N.,Hamada, S. & Fetissov,M. 2010.
Bayesian inference in oil spill response management.
WIT Transactions on Information and Communication
Technologies,vol.43.
[26]Helle, I., Lecklin, T., Jolma, A. & Kuikka, S. 2011.
Modelling the effectiveness of oil combating from an
ecologicalperspective‐aBayesiannetwork
fortheGulf
of Finland; The Baltic Sea. Journal of Hazardous
Materials,vol.185(1),182–192.
[27]Lu, L., Goerlandt, F., Valdez Banda, O., Kujala, P.,
Höglund,A.&Arneborg,L.2019.ABayesianNetwork
riskmodel for assessingoil spillrecovery effectiveness
intheice‐coveredNorthernBalticSea.
MarinePollution
Bulletin,vol.139,440–458.
[28]Parviainen, T., Kuikka, S. & Haapasaari, P. 2022.
Enhancing science‐policy interface in marine
environmentalgovernance:Oilspillresponsemodelsas
boundary objects in the Gulf of Finland, Baltic Sea.
MarinePolicy,vol.135(2022)104863.
[29]HELCOM2022.HELCOMAnnualreportondischarges
observed during aerial surveillance in the Baltic Sea
2021.HelsinkiCommission.
