Choi, Young-Seo; Lis an Be nal, Ma ia; K i oshapkina, Ma ga i a; Yeo, Gi-Tae
A icle
Analyzing isk ac o s in ship- o-ship lique ied na u al gas
bunke ing ope a ions
Asian Jou nal o Shipping and Logis ics (AJSL)
P o ided in Coope a ion wi h:
Ko ean Associa ion o Shipping and Logis ics, Seoul
Sugges ed Ci a ion: Choi, Young-Seo; Lis an Be nal, Ma ia; K i oshapkina, Ma ga i a; Yeo, Gi-Tae
(2025) : Analyzing isk ac o s in ship- o-ship lique ied na u al gas bunke ing ope a ions, Asian
Jou nal o Shipping and Logis ics (AJSL), ISSN 2352-4871, Else ie , Ams e dam, Vol. 41, Iss. 1, pp.
52-60,
h ps://doi.o g/10.1016/j.ajsl.2025.01.003
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/329756
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h ps://c ea i ecommons.o g/licenses/by-nc-nd/4.0/
Analyzing isk ac o s in ship- o-ship lique ied na u al gas
bunke ing ope a ions
Young-Seo Choi, Ma ia Lis an Be nal, Ma ga i a K i oshapkina , Gi-Tae Yeo
*
G adua e School o Logis ics, Incheon Na ional Uni e si y, Incheon 22012, Sou h Ko ea
ARTICLE INFO
Keywo ds:
Risk ac o s
Ship- o-Ship
LNG
Bunke ing Ope a ion
CFPR
ABSTRACT
This s udy e alua es he isk ac o s du ing ship- o-ship lique ied na u al gas bunke ing in Sou h Ko ea using he
consis en uzzy p e e ence ela ions me hod. The s udy showed ha he de ailed ac o “The ope a o ’s inabili y
o ecognize he ope a ional condi ion, sys em s a us, o measu ing ins umen s a us” was anked i s wi h
0.112, alluding o he human e o p incipal ac o . “Con ol equipmen mal unc ion (0.073)” and “Ope a ion o
he Eme gency Shu -Down (ESD) Sys em (0.071)” anked in second and hi d place, espec i ely. Based on his
s udy’s indus ial implica ions, expe opinions we e in eg a ed o p io i ize isk ac o s, enabling indus ial
manage s o p oac i ely p epa e.
1. In oduc ion
In 2018, he In e na ional Ma i ime O ganiza ion (IMO) se a land-
ma k goal o educe g eenhouse gas (GHG) emissions om in e na ional
shipping. This objec i e, aimed a educing emissions by a leas 50 % by
2050 (compa ed wi h 2008 le els), ep esen s a signi ican miles one in
global clima e change go e nance (Ga cia e al., 2021). Achie ing he
objec i es o he Pa is Ag eemen , aimed o s abilize global empe a-
u es a “well below 2◦C, and owa d 1.5◦C,” equi es eaching ne -ze o
CO
2
emissions by 2050–2070 o ea lie and possibly ansi ioning o
ne -nega i e emissions he ea e . These a ge s apply o all sec o s o
he economy (Ba aille, 2020).
Selec ing an op imal solu ion o comply wi h egula ions and add ess
economic p essu es poses a signi ican challenge o shipowne s and
decision-make s. Choosing op ions equi es conside ing en i onmen al,
echnological, and economic ac o s, o en in ol ing con lic ing p io -
i ies and adeo s. In he con ex o shipping, essels can choose be-
ween in eg a ing emission aba emen echnologies (such as sc ubbe s)
using mo e expensi e low-sul u uels, such as ma ine gas oil (MGO) o
ma ine diesel oil (MDO), and adop ing lique ied na u al gas (LNG)
(Dalaklis e al., 2017). Despi e i s ossil o igin, LNG is widely ecognized
as mo e en i onmen ally sus ainable, pa icula ly when compa ed o
coal and oil. Consequen ly, i has become a subs an ial componen o he
ene gy mix in nume ous de eloped economies. I demons a es alue as
a b idge uel, acili a ing a shi om ossil uels o enewable ene gy
sou ces (S ini asan e al., 2024).
Among he LNG bunke ing op ions, Ship- o-ship (STS) is a ype o
di ec ansshipmen in which goods a e ans e ed di ec ly om he
mo he essel o he eede essel while he ships a e moo ed nex o one
ano he in he open sea (Al Sam ou e al., 2024). The e o e, STS
bunke ing o e s ad an ages, such as educed ueling supply ime,
inc eased speed, and la ge capaci y o loading and unloading, making
his compendium o bene i s a signi ican con ibu ion o ope a ional
cos sa ings (Liu e al., 2024). Addi ionally, when cons uc ing ixed
ins alla ions a po loca ions has es ic ions, STS bunke ing se es as a
p ac ical and e icien al e na i e, especially o essels wi h b ie po
s ays (Anezi is e al., 2022).
Thus, iden i ying isk ac o s ensu es sa e STS bunke ing p ac ices.
Analyzing eal STS bunke ing case scena ios unde sco es he c i ical
impo ance o leaking acciden s (Nubli e al., 2022), he designa ion o
he pe inen sa e y zones (Pa k e al., 2018a; Anezi is, 2022; Zhang
e al., 2024; Duong e al., 2023), ship collisions (A ici e al., 2020;
Sokukcu & Saka , 2022), human e o (U laz e al., 2022a), and o he
ac o s. Addi ionally, unce ain ies and inciden s a ising om egula-
ions and guidelines loopholes (Anezi is e al., 2022) should be consid-
e ed simul aneously.
Despi e hese s udies, esea ch on he isk ac o s associa ed wi h STS
ope a ions in Sou h Ko ea is limi ed. This s udy iden i ies isk ac o s
using he consis en uzzy p e e ence ela ion (CFPR) me hodology and
weighs and anks a se ies o isk ac o s ob ained om p e ious
* Co esponding au ho .
E-mail add esses: [email p o ec ed] (Y.-S. Choi), [email p o ec ed] (M. Lis an Be nal), [email p o ec ed] (M. K i oshapkina), k yeo@
incheon.ac.k (G.-T. Yeo).
Con en s lis s a ailable a ScienceDi ec
The Asian Jou nal o Shipping and Logis ics
jou nal homepage: www.else ie .com/loca e/ajsl
h ps://doi.o g/10.1016/j.ajsl.2025.01.003
Recei ed 17 No embe 2024; Accep ed 27 Janua y 2025
The Asian Jou nal o Shipping and Logis ics 41 (2025) 52–60
A ailable online 30 Janua y 2025
2092-5212/© 2025 The Au ho s. Published by Else ie B.V. on behal o The Ko ean Associa ion o Shipping and Logis ics, Inc. This is an open access a icle unde
he CC BY-NC-ND license ( h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/ ).
li e a u e and expe s’ opinions.
2. Cu en s a us o LNG STS bunke ing
2.1. Ca bon emission egula ions
In July 2023, he In e na ional Ma i ime O ganiza ion (IMO) adop-
ed he “2023 S a egy o he Reduc ion o G eenhouse Gas Emissions
om Ships” du ing he 80 h session o he Ma ine En i onmen P o-
ec ion Commi ee (MEPC), amending he a ge imeline o achie ing
ne -ze o g eenhouse gas emissions in in e na ional shipping. The a ge
yea o ne -ze o emissions was b ough o wa d om 2100 o 2050.
Acco dingly, he p e ious educ ion goal o a 50 % cu by 2050 was
e ised o educing ca bon emissions by 30 % by 2030 and by 70 % by
2040. By 2050, as implied by he e m “Ne -ze o,” ca bon emissions
mus con e ge o ze o, as shown in Table 1.
2.2. Response s a egies o ca bon emission egula ions
G adual esponse s a egies ha e been es ablished o decommission
exis ing ships and in oduce new ships. Table 2 shows a p oposed h ee-
s ep coun e measu e s a egy, wi h he i s s ep al eady unde imple-
men a ion. LNG and me hanol a e conside ed b idge uels in he second
s ep o he esponse s a egy. In pa icula , LNG is equen ly classi ied
as a ossil uel owing o i s a ibu ed CO
2
emission le els o 2.7 ons pe
on o uel.
To align wi h he IMO’s ca bon egula ions, changes o ship uel, he
main sou ce o ca bon dioxide emissions, a e essen ial. Cu en ly, LNG is
p io i ized as a b idge uel o achie e he IMO’s sho - e m a ge o
educing emissions by a leas 30 % by 2030. LNG emi s 25 % less ca bon
dioxide compa ed wi h bunke oil and is mo e economically ad an a-
geous han uel oil. Despi e being a ossil uel wi h inhe en limi a ions,
LNG has been chosen as he p ima y b idge uel due o ac o s in ol ing
ship engine compa ibili y, supply chain, in as uc u e, and p ice
compe i i eness.
The e o e, simul aneous LNG bunke ing ope a ions, in ol ing LNG
bunke ing and ca go handling a he same ime, a e being implemen ed
o mo e e icien po ope a ions. Ini ially, when supplying LNG o
ships, all o he asks had o be hal ed as a p ecau ion agains po en ial
acciden s. Howe e , because la ge-scale bunke ing ope a ions can ake
wo o h ee days, and shipping companies p e e ea ly depa u e,
simul aneous LNG bunke ing ope a ions ha e begun o be in oduced.
2.3. Cu en S a us o LNG Bunke ing
Agains his backd op, he expansion o bunke ing in as uc u e is
necessa y o he inc eased use o LNG uel. As o Ma ch 2022, 15 % o
he o al numbe o ship o de s and 33 % o he o al onnage we e o
ships using LNG uel p opulsion. Acco ding o Cla kson Resea ch
Table 1
IMO Ca bon dioxide (CO
2
) emissions educ ion a ge o 2050.
Pe iod Ca bon dioxide (CO
2
) emissions educ ion a ge o 2050
(Rela i e o he ca bon emissions o 2008)
Yea 2015–2019 10 %
Yea 2020–2024 20 %
Yea 2025–2030 30 %
Yea 2030–2050 70 %
Yea 2050 100 %
Sou ce: Compiled by he au ho based on da a e ie ed om he In e na ional
Ma i ime O ganiza ion (n.d.).
Table 2
Response S a egies o ca bon emission egula ions.
S ep 1 S ep 2 S ep 3
Fuel ype Con en ional uel B idge uel Al e na i e uel
Scope o he
measu es
Ope a ional Technical T ansi ion o eco- iendly (Use
o low-ca bon and ca bon- ee
uels)
Ta ge (ships) Decommissioning o he exis ing lee be o e 2030 - Decommissioning o he exis ing lee be o e 2040
- In oduc ion o new lee be o e 2030
In oduc ion o new lee a e
2030
Response
s a egy
Decele a ed ope a ion (op imal ope a ion speed),
op imiza ion o shipping lanes (shipmen ), and applica ion
o sma ship solu ion pla o m
In oduc ion o LNG p opulsion ships, use o bio uels, and
ins alla ion o ene gy-sa ing de ices (e.g., in oduc ion o wind
powe de ices o help p opel ships)
Ammonia o hyd ogen
p opulsion line
(b idge uel no used)
Sou ce: POSCO low
Table 3
In-dep h in e iew o ganiza ions.
O ganiza ion Name Wo k Rela ed
Ko ea LNG bunke ing LNG bunke ing ship ope a o
H Line Shipping Company LNG bulk ship ope a o
Ko ea Resea ch Ins i u e o Ships
& Ocean Enginee ing
LNG bunke ing ship de elopmen o ganiza ion
Ko ean Regis e o Shipping LNG bunke ing essel app o al o ganiza ion
Ulsan Po Au ho i y Po au ho i y ha ope a e LNG bunke ing
acili ies
POSCO Flow Supe iso o e i ica ion wo k o he STS
me hod’s simul aneous LNG bunke ing and
unloading ope a ion
Table 4
Fac o s ex ac ed.
P incipal ac o De ailed ac o
Componen e o Ship pump mal unc ion
P essu e o le el senso mal unc ion in
unc ionali y, audibili y, o isibili y
De ec o ailu e
Con ol equipmen mal unc ion
O ganiza ional e o Lack o a well-documen ed ope a ional p ocedu e
Insu icien p epa a ion be o e he ope a ion
Lack o communica ion be ween in e disciplina y
eams
(including lack o communica ion be ween
mul ina ional sailo s)
Lack o exis ence/implemen a ion o acciden
p e en ion s a egies
(high wa e, wind, igni ion, d opped objec )
Human e o The ope a o ’s inabili y o ecognize he
ope a ional condi ion, sys em s a us, o measu ing
ins umen s a us
Ope a o ’s inadequa e pe o mance due o high
wo kload o wo k dissa is ac ion
Sys em e o o LNG supply &
ecei ing essel
Elec ici y blackou , communica ion e o
Pipeline leakage
O e p essu e/o e hea ing o equipmen
Ope a ion o he ESD Sys em
Wo k p ocess e o Inadequa e pu ging o he loading a m
O e - ueling (only o ueled ships)
BOG (Boil-o gas) emo al mal unc ion
Flame igni ion
Poo ballas o moo ing line ailu e
Y.-S. Choi e al.
The Asian Jou nal o Shipping and Logis ics 41 (2025) 52–60
53
(2023), he majo ypes o essels a e shi ing o LNG p opulsion, and
o e he nex 10 yea s, he a e age con e sion a e o LNG-powe ed o
newly buil ships is p ojec ed o be 31 %. Fu he mo e, by 2043, global
LNG bunke ing demand is expec ed o inc ease o 53 million ons.
In Asia, coun ies such as Japan and China a e de eloping LNG-
powe ed ships. Singapo e de eloped i s i s bunke ing essel and
decla ed i s posi ion as an LNG hub po . Addi ionally, in 2021, he i s
simul aneous ship- o-ship (STS) LNG bunke ing ope a ion was
success ully conduc ed a he Po o Singapo e.
In 2020, Sou h Ko ea’s Minis y o Oceans and Fishe ies enac ed he
“Ac on he P omo ion o he De elopmen and Dis ibu ion o En i-
onmen ally F iendly Ships,” known as he “Eco-F iendly Ship Ac ,” and
es ablished a basic plan se o con inue un il 2030. This legisla ion in-
cludes he manda o y use o eco- iendly uel-powe ed ships o newly
cons uc ed go e nmen public essels, and suppo o R&D ela ed o
eco- iendly ships. Addi ionally, om Sep embe 2020, he i e majo
Table 5
Summa y o expe ’s ques ionnai e.
In e iew
Da e
Feb ua y 2024–Ap il 2024
Expe s
a ea
Shipping
company
Resea ch
Ins i u e
LNG bunke ing ship
app o al au ho i y
Po
Au ho i y
STS-based LNG bunke ing and unloading
simul aneous wo k demons a ion
Academic
Expe s
expe ience
(yea s)
Less han 10
yea s
0 0 0 1 0 0
10–15 yea s 2 1 1 0 0 2
16–20 yea s 0 2 0 0 1 0
Mo e han 21
yea s
0 0 1 0 0 0
To al expe s 11 2 3 2 1 1 2
Fig. 1. P incipal ac o esul .
Fig. 2. Componen e o de ail ac o esul .
Y.-S. Choi e al.
The Asian Jou nal o Shipping and Logis ics 41 (2025) 52–60
54
domes ic po s—Busan, Incheon, Yeosu-Gwangyang, Ulsan, and
Pyeong aek-Dangjin—we e designa ed as Emission Con ol A eas
(ECA), whe e he sul u con en o ship uel is egula ed and main ained
a o below 0.1 %. In 2015, uck- o-ship bunke ing was p ima ily
conduc ed o small LNG-powe ed essels. Howe e , h ough a na ional
R&D p ojec in 2022, Sou h Ko ea is de eloping i s i s LNG bunke ing
essel and ope a ional echnologies o LNG ship bunke ing. Addi ion-
ally, unde he join suppo o he go e nmen , esea ch ins i u ions,
and he indus y, including he Minis y o T ade, Indus y, and Ene gy,
he Minis y o Oceans and Fishe ies, KOGAS, POSCO, and H-line LNG, a
bunke ing demons a ion p ojec was conduc ed. The demons a ion
ope a ion ook place on Oc obe 28, 2023, a Gwangyang Po , whe e
1000 ons o LNG we e simul aneously bunke ed on a essel engaged in
ca go handling.
Ulsan Po is home o he coun y’s i s and la ges LNG bunke ing-
exclusi e e minal, o e ing nume ous expec ed bene i s in line wi h eco-
iendly po policies. Addi ionally, in July 2023, Ulsan Po success ully
became he i s in he wo ld o bunke g een me hanol using he pipe-
o-ship (PTS) me hod.
3. Li e a u e e iew
The STS bunke ing p ocess o LNG is a ela i ely no el concep in
he ma i ime indus y, eme ging alongside he global inc ease in he
u iliza ion o LNG (U laz e al., 2022b). Anezi is e al. (2022) a gued ha
STS bunke ing is an ad an ageous e ueling me hod o po s accom-
moda ing essels o a ying sizes — om small o e y la ge capaci y-
—especially hose wi h b ie po s ays. They emphasized ha STS
bunke ing is a logical subs i u e o po s whe e ixed ins alla ions a e
ei he p ohibi ed o un a o ed. Depending on he policies o he po
au ho i y, STS bunke ing can occu ei he a he pie o ancho age in
open seas.
Al hough a ious s udies unde sco e he bene i s o STS bunke ing,
his p ac ice in oduces po en ial isks. Anezi is e al. (2022) iden i ied
Fig. 3. O ganiza ional e o de ail ac o esul .
Fig. 4. Human e o de ail ac o esul .
Y.-S. Choi e al.
The Asian Jou nal o Shipping and Logis ics 41 (2025) 52–60
55
se e al possible isks associa ed wi h LNG bunke ing. The accu acy o
he bunke ing equipmen , as highligh ed by Tam (2022), also plays a
c ucial ole in mi iga ing isks. Mo eo e , he lack o da a, limi ed un-
de s anding o sa e s o age and bunke ing o LNG, and a sca ci y o
comp ehensi e isk assessmen s udies con ibu e o he o e all un-
ce ain ies in LNG bunke ing (Ge bec & Anezi is, 2022).
Iannacconea e al. (2018) and Anezi is e al. (2022) speci y po en ial
scena ios ha may lead o LNG elease, i es, o explosions du ing
bunke ing. These include mal unc ions in boil-o emo al, o e illing o
ship anks, inad e en al e closu es, ex e nal i es, and addi ional
loads caused by e en s such as collisions, g ounding, poo ballas , and
moo ing line issues. Fu he mo e, na u al e en s such as ea hquakes,
sunamis, o high winds can impose ex a loads on hoses, he eby
inc easing he associa ed isks. Sys em ailu es, encompassing compo-
nen mal unc ions and human e o s, a e also conside ed po en ial
haza ds (Anezi is e al., 2022).
To in es iga e he p obabili y o human e o , Fan e al. (2022)
in es iga ed he ole o human ac o s in LNG bunke ing sa e y. Human
e o s a ise om a combina ion o onsi e condi ions and pe sonal
psychological ac o s. The concep o sa e y philosophical ac o s high-
ligh s he in luence o psychological ac o s on human pe o mance. A
obus sa e y cul u e has been p oposed o os e a posi i e men al s a e
among wo ke s du ing LNG bunke ing asks, he eby educing he
likelihood o human e o (Fan e al., 2022). U laz e al. (2022a)
explained ha enhancing shipboa d sa e y signi ican ly elies on human
eliabili y. As s ingen en i onmen al egula ions a e en o ced, ship
owne s a e explo ing he adop ion o cleane al e na i e uels o mini-
mize ship emissions and ensu e compliance wi h p esc ibed limi s.
U laz e al. (2022b) also emphasized ha bunke ing ope a ions
expe ience undesi able inciden s, including o e low, leakage, and sea
pollu ion. The challenges o he o e all LNG bunke ing we e discussed
by Aymelek e al. (2014), who di ided hem in o i e ca ego ies: sys-
emic, ope a ional, echnical, sa e y, and conjec u al.
Mac o-le el challenges p esen signi ican obs acles o business
en ep eneu s aiming o achie e mic o-le el objec i es. Poli ical in-
s abili ies, he isk o wa , inancial c ises, luc ua ing na u al gas p ices,
and po en ial addi ional egula o y measu es o en i onmen al con-
ce ns ha e eme ged as signi ican sys emic challenges con on ing LNG
Fig. 5. Sys em e o o LNG supply & ecei ing essel de ail ac o esul .
Fig. 6. Wo k p ocess e o de ail ac o esul .
Y.-S. Choi e al.
The Asian Jou nal o Shipping and Logis ics 41 (2025) 52–60
56
bunke ing (Aymelek e al., 2014).
The abili y o handle cold ma e ials wi hin pe inen ship s uc u es
and he po en ial asphyxia ion isk o indi iduals engaged in
bunke ing, con ingen upon ci cums ances du ing an LNG spill, pool
i e, apo cloud i e, explosions, and apid phase ansi ion, pose sig-
ni ican sa e y challenges. These challenges necessi a e p e en i e
measu es h ough enhanced echnological ad ancemen s and aining
p og ams o bo h he c ew and bunke ing pe sonnel (Aymelek e al.,
2014).
Pa k e al. (2018b) highligh ed he lack o guidelines o es ablishing
sa e y zones, poin ing o hei c ucial ole in mi iga ing haza ds asso-
cia ed wi h LNG bunke ing. In he e en o gas leakage, wind speed and
di ec ion, su ounding condi ions, ship d a , and ca go loading plan-
ning we e iden i ied as c i ical in de e mining he need o sa e zones.
Among he s udies add essing he simul aneous ope a ions pe o med
du ing he bunke ing p ocess, Fan e al. (2021) e iewed he li e a u e
on hei ole in sa e y and isk assessmen , namely, people mo emen s,
ca go loading/unloading, supply/c ew ans e , c ane maneu e ing,
po ope a ions, and bunke ing. The au ho s concluded ha he main
limi a ions co esponded o he lack o consensus in sa e y egula ion
guidelines, he di e ence in he unde s anding o sa e y philosophy
conce ning LNG bunke ing, and he absence o a s anda dized isk
analysis me hodology.
As Su ino (2023) s a ed, he sa e y zone o STS LNG bunke ing
mus be mu ually de e mined by he wo in ol ed essels, and hei
compa ibili y mus be e i ied and con i med be o e bunke ing ac i -
i ies begin. The isks associa ed wi h STS bunke ing include moo ing
ailu e, ca go ans e hose ailu e, a igue, a ailabili y o pe sonnel, and
concu en ope a ions.
Sul ana e al. (2019) iden i ied se e al ca ego ies o STS bunke ing
haza d issues, including componen e o (issues ela ed o pump mal-
unc ion, con ol equipmen , p essu e o le el mal unc ion), o ganiza-
ional e o (lack o well- ained human esou ces, lack o in e - eam
communica ion, ope a ional p ocedu e documen a ion), human e o
(lack o condi ion awa eness, missing o w ong ope a ional decision),
so wa e e o (so wa e bug, lagging, o sabo age), sys em o design
e o (o elec ical blackou , equipmen o e p essu e, o o e hea ing
pipeline leaks), and ex e nal e en s (wind, wa es)
Wu e al. (2021) used a Bayesian ne wo k o e alua e isks associa ed
wi h dange ous acciden s du ing LNG bunke ing. Addi ionally, he EPE,
an ene gy-based model, was also in oduced o de i e he causal s uc-
u e. The Bayesian ne wo k-based isk assessmen model p oposed
h ough he analysis iden i ied isks and e alua ed he e olu iona y
p ocess om he cause o he e ec o LNG acciden s. Fu he mo e, a
sensi i i y analysis was conduc ed o quan i y he co ela ion be ween
each ac o conside ed in he LNG acciden s and iden i y he p incipal
isks. The p oposed isk assessmen amewo k can be used o e alua e
acciden sa e y and p e en losses du ing LNG bunke ing in po s.
To e alua e he po en ial isks a ising du ing ship ope a ion in
bunke ing wo k, Goksu and A slan (2023) p oposed a quan i a i e
ma ine sa e y analysis based on he uzzy ailu e mode and e ec
analysis me hodology. As a esul o he analysis, he high- isk ac o s
we e de i ed as a igue/indi idual e o s, e y s ong winds, excessi e
hea , low ide, and inc eased/dec eased ship speed. Addi ionally, he
con ol o he p elimina y ailu e mode was necessa y o he ship’s
sa e y. Thei s udy e ealed ha i was easy o ma ine sa e y inspec o s,
sa e y esea che s, and heal h, sa e y, en i onmen al, and quali y
manage s o iden i y po en ial isks, e ec s, and esul s du ing ancho -
ing/coas al wo k.
LNG bunke ing me hods include uck- o-ship (TTS), STS, and
pipeline- o-ship (PTS). The ypes o p oblems ha a ise may a y
depending on he wo king me hod. Fo example, Duong e al. (2023)
s udied he ou low o LNG and NH3 gene a ed using he PTS me hod. A
ma hema ical model was p esen ed o in es iga e he LNG/NH3 leakage
dispe sion cha ac e is ics and de e mine he sa e y dis ance du ing
bunke ing be ween ships. A nume ical s udy was conduc ed o 26
scena ios conside ing a ious ope a ing and en i onmen al condi ions,
leakage hole diame e s, leakage a es, wind speeds, and wind di ec ions.
Acco ding o he esul s, NH3 exhibi ed a la ge dispe sion ange han
LNG unde he same ope a ing condi ions. The e o e, NH3 equi es a
wide sa e zone dis ance, and i s dispe sion ime is longe han ha o
LNG. The wea he condi ions (wind speed and wind di ec ion) and
leakage cha ac e is ics (leak a e and leakage pe iod) we e iden i ied as
impo an pa ame e s.
Jeong e al. (2018) in oduced a p ac ical new me hod o es ab-
lishing sa e y exclusion zones du ing LNG bunke ing. Thei esul s show
ha p obabilis ic isk assessmen , which ocuses only on
popula ion-independen analyses, is somewha inapp op ia e when
de e mining sa e y exclusion zones. The es ablished hypo hesis showed
ha he ange o sa e y exclusion zones ends o be un ealis ic. The
app oach p oposed in his s udy p o ed use ul o de e mining he zones
mo e ealis ically.
Se e al s udies ha e been conduc ed o es ablish sa e y- ela ed
s anda ds and egula ions o LNG bunke ing. This s udy e alua es he
isk ac o s associa ed wi h STS ope a ions in ol ing LNG bunke ing
me hods. Explo ing esea ch opics no examined in exis ing domes ic
s udies is signi ican , p o iding aluable e e ence ma e ials o STS
Table 6
Global Impo ance Resul .
P incipal Fac o s De ailed Fac o s Global
Impo ance
Rank
Componen e o
(0.232)
Ship pump mal unc ion
(0.226)
0.053 8
P essu e o le el senso
mal unc ion in unc ionali y,
audibili y, o isibili y (0.239)
0.056 6
De ec o ailu e (0.221) 0.051 9
Con ol equipmen
mal unc ion (0.313)
0.073 2
O ganiza ional e o
(0.158)
Lack o a well-documen ed
ope a ional p ocedu e (0.241)
0.038 16
Insu icien p epa a ion be o e
he ope a ion (0.249)
0.039 13
Lack o communica ion
be ween in e disciplina y
eams (including lack o
communica ion be ween
mul ina ional sailo s) (0.242)
0.038 15
Lack o exis ence/
implemen a ion o acciden
p e en ion s a egies (high
wa e, wind, igni ion, d opped
objec ) (0.269)
0.043 12
Human e o (0.172) The ope a o ’s inabili y o
ecognize he ope a ional
condi ion, sys em s a us, o
measu ing ins umen s a us
(0.651)
0.112 1
Ope a o ’s inadequa e
pe o mance due o high
wo kload o wo k
dissa is ac ion (0.349)
0.060 5
Sys em e o o LNG
supply & ecei ing
essel (0.239)
Elec ici y blackou ,
communica ion e o (0.228)
0.055 7
Pipeline leakage (0.272) 0.065 4
O e p essu e/o e hea ing o
equipmen (0.204)
0.049 10
Ope a ion o he ESD Sys em
(0.296)
0.071 3
Wo k p ocess e o
(0.198)
Inadequa e pu ging o he
loading a m (0.192)
0.038 17
O e - ueling (only o ueled
ships) (0.195)
0.039 14
BOG (Boil-o gas) emo al
mal unc ion (0.187)
0.037 18
Flame igni ion (0.245) 0.048 11
Poo ballas o moo ing line
ailu e (0.180)
0.036 19
Y.-S. Choi e al.
The Asian Jou nal o Shipping and Logis ics 41 (2025) 52–60
57
wo ke s, po esea ch o icials, and go e nmen au ho i ies.
4. Resea ch me hodology
This s udy implemen ed he CFPR me hod, based on He e a-Viedma
e al. (2004), o iden i y he isks p oduced in STS bunke ing du ing LNG
ope a ions in Ko ea. He e a-Viedma e al. (2004) p oposed he CFPR
me hod o s uc u e complex decisions by d awing on he ela i e
impo ance o ac o s and hei p io i ies a each hie a chical le el pe
pai wise compa ison p e e ence ma ix.
Acco ding o Chen and Chao (2012), pai wise compa ison p oced-
u es -such as AHP- p esen (n ×(n −1))/2 su ey compa isons o a
g oup (n) c i e ion. This can lead o inconsis encies and expe con usion
due o he ex ensi e numbe o su ey i ems. The e o e, he CFPR
me hod seeks o uphold logical consis ency in decision-making p o-
cesses by implemen ing a educed numbe o simple su ey i ems, which
sh inks (n ×(n −1))/2 compa isons o only an (n −1) compa ison o
g oup (n) c i e ia (Chen & Chao, 2012).
Chao and Chen (2009) and Wang and Lin (2009) u he explo ed he
ounda ional concep s in he CFPR me hodology. In hese s udies, he
CFPR implemen a ion o mul iplica i e p e e ence ela ions is he a io
shown o ep esen he ela i e impo ance o di e en c i e ia in
decision-making. In con as , uzzy p e e ence ela ions enable lexi-
bili y in exp essing p e e ences, he eby educing unce ain y.
In mul iplica i e p e e ence ela ion (1), a se o expe s (E = {e1,
e2,…,em,m≥2}) mus s a e hei choice o e a ini e a ay o
possibili ies (X). (Х= {x1,x2,…,xn,n≥2}), indica ed by a
p e e ence ela ion ma ix А⊂X×X,А=(aij),∀i,j∈ {1,…,n},
aij ∈[1
5,5] ha highligh s he p e e ence a io o he possibili y xi o
he possibili y xj. In he equa ion, aij =1 indica es xi and xj equi -
alency and aij =5 deno es ha xi is he bes decision compa ed wi h
xj. Thus, p e e ence ela ion (A) is a mul iplica i e ecip ocal
aij ×aji =1∀i.j∈ {1,…,n}(1)
Conce ning uzzy p e e ence ela ions (2), in a decision-making
con ex , he expe c i e ia o e a ini e a ay o possibili ies es ablish
he p e e ence a io o he possibili y xi o e xj; X is deno ed by a posi i e
ela ion ma ix P⊂X×X wi h membe ship unc ion
μ
p(xi,xj)=
pij. Addi ionally, pij =1
2 indica es ha he expe s a e unbiased,
showing no eason o p e e xi o xj(xi∼xj). This is e e ed o as
indisc imina ion. Con e sely, pij =0 deno es a p e e ence o xi o e xj
xi>xj, which is e e ed o as p ecedence. The p e e ence ela ion (P) is
an addi i e ecip ocal because he sum o hei ela i e p e e ences al-
ways equals one.
pij +pji =1∀i.j∈ {1,…,n}(2)
The decision ma ixes in CFPR s em om he h ee main
p oposi ions:
Numbe 1. A se o al e na i es abou a mul iplica i e p e e ence
ela ion is adop ed, Х= {x1,x2,…,xn,), A= (aji), wi h
aij ∈[1
5,5]. The ecip ocal addi i e p e e ence is exp essed as
Pij =g(aij)=1
2(1+log5aij)(3)
In he p eceding con ex , g ep esen s a ans o ma ion unc ion, and
log5aij is ela ed o he implemen a ion o aij ∈[1
5,5].
Numbe 2. In he con ex o his ecip ocal uzzy p e e ence ela-
ionship, P =g(A), whe e P =pij co esponds o
Pij +Pjk +Pki =3
2∀i,j,k(4)
Pij +Pjk +Pki =3
2∀i<j<k(5)
Numbe 3. The ecip ocal addi i e uzzy ela ion, P = (pij), is
exp essed as:
Pij +Pjk +Pki =3
2∀i<j<k
Pi(i+1)+Pi(i+1)(i+2)+…+Pj(i−1)+Pji =j−i+1
2∀i<j(6)
In suppo o he o iginal guidelines se by He e a-Viedma e al.
(2004), Chen and Chao (2012) highligh ed he necessi y o a ans o m
unc ion o no malize alues in a decision ma ix ha alls ou side he
in e al [0,1]. This s ep is c ucial o main aining ecip oci y and addi-
i e consis ency. U ilizing a linea ans o ma ion, he alues in he
decision ma ix will be con ained wi hin a b oade in e al [ − a,1+a].
The e o e, :[ − a,1+a]→[0,1]allows o he ollowing ans-
o ma ion unc ion:
(Pk
ij)=(Pk
ij +a
1+2a)(7)
5. Empi ical analysis
5.1. Selec ion o ac o s
This s udy u ilized ou p incipal ac o s and 27 de ailed ac o s
iden i ied by Sul ana e al. (2019) o ex ac he isk ac o s associa ed
wi h STS LNG bunke ing ope a ions. Addi ionally, en de ailed ac o s
we e added unde he p incipal ac o “Wo k p ocess e o ” p esen ed in
Anezi is e al. (2022). These newly added ac o s we e inco po a ed in o
an in-dep h in e iew o selec he mos app op ia e ac o s o he inal
ques ionnai e composi ion.
In-dep h in e iews we e conduc ed be ween Janua y 4 and Janua y
19, 2024, o de e mine he isk ac o s du ing STS LNG bunke ing in
Sou h Ko ea. A isk ac o composi ion able consis ing o i e p incipal
ac o s and 37 de ailed ac o s was deli e ed o he in e iewees in
ad ance. The in e iewees we e asked o iden i y i ems deemed
impo an as isk ac o s. Addi ionally, in e iewees we e asked o eely
sugges any ac o s hey conside ed missing om he lis ed p incipal and
de ailed ac o s.
Six expe s om six STS LNG bunke ing wo k- ela ed o ganiza ions
pa icipa ed in he in-dep h in e iews. The expe p o iles and o gani-
za ional a ilia ions a e lis ed in Table 3.
Eigh een ac o s we e con i med as impo an by a leas h ee o he
six expe s du ing he in-dep h in e iews. Each i em was selec ed and
e ised o de e mine he inal ac o s. Following hese esul s, he
de ailed ac o “Ope a ion o he eme gency shu -down (ESD) sys em”
was added o he de ailed ac o s du ing he in-dep h in e iews.
Fu he mo e, e lec ing he expe s’ opinions, he name o he p incipal
ac o , “Sys em e o ,” was e ised o “Sys em e o o LNG supply &
ecei ing essel.” The de ailed ac o “Lack o communica ion be ween
in e disciplina y eams” was also e ised in o “Lack o communica ion
be ween in e disciplina y eams (including lack o communica ion be-
ween mul ina ional sailo s).” The isk ac o s de e mined h ough in-
dep h in e iews we e i e p incipal ac o s and 19 de ailed ac o s, as
shown in Table 4.
5.2. Resul s
The ac o s de i ed om he li e a u e and in-dep h in e iews we e
classi ied in o i e main ac o s and 19 sub- ac o s and used o he
analysis. The CFPR ques ionnai e was de eloped o compa e he main
ac o s and sub- ac o s in he pai s. Table 5 summa izes he ele en LNG
bunke ing and logis ics expe s who pa icipa ed in he su ey.
Y.-S. Choi e al.
The Asian Jou nal o Shipping and Logis ics 41 (2025) 52–60
58
Focusing on he p incipal ac o ankings, he esul s a e as ollows:
“Sys em e o o LNG supply & ecei ing essel” (0.239), “Componen
e o ” (0.232), “Wo k p ocess e o ” (0.198), “Human e o ” (0.172),
and “O ganiza ional e o ” (0.158). The highes anked ac o s, “Sys em
e o o LNG supply & ecei ing essel” and “Componen e o ,” do no
e e o human de ec s, bu hose de ec s in he sys em di ec ly ela ed o
bunke ing wo k and isk; i is conside ed a high- isk ac o . Con as -
ingly, “Human e o ” and “O ganiza ional e o ” a e di ec human-
caused p oblems ha can be imp o ed h ough sa e y educa ion.
The e is no signi ican di e ence in he weigh alues, as shown in Fig. 1.
The e o e, all elemen s o he p incipal ac o s we e conside ed equally
impo an .
To ully comp ehend he conno a ions o he analysis, i was neces-
sa y o e ise he esul s o he de ailed ac o s pe p incipal ac o .
Rega ding he “Componen e o p incipal ac o ,” he de ailed ac-
o s we e anked as ollows: “Con ol equipmen mal unc ion” (0.313),
“P essu e o le el senso mal unc ion in unc ionali y, audibili y, o
isibili y” (0.239), “Ship pump mal unc ion” (0.226), and “De ec o
ailu e” (0.221). No ably, bo h “Con ol equipmen mal unc ion” and
“P essu e o le el senso mal unc ion in unc ionali y, audibili y, o
isibili y” e e o isk ac o s ha can lead o haza dous acciden s due o
he ola ile LNG uel cha ac e is ics, especially when ela ed o
mal unc ion.
The esul s o he p incipal ac o s o o ganiza ional e o a e closely
aligned in alue, wi h “Lack o exis ence/implemen a ion o acciden
p e en ion s a egies (high wa e, wind, igni ion, d opped objec )’
(0.269), “Insu icien p epa a ion be o e he ope a ion” (0.249), and
“Lack o communica ion be ween in e disciplina y eams (including
lack o communica ion be ween mul ina ional sailo s)” (0.242),
espec i ely. The “Lack o exis ence/implemen a ion o acciden p e-
en ion s a egies (high wa e, wind, igni ion, d opped objec )” ac o ’s
high- anking alue may be ela ed o he e ec i eness o esponse
lexibili y in eme gencies when guidelines on how o p e en na u al
disas e s and unexpec ed acciden s a e p o ided. The e we e no signi -
ican di e ences in he weigh alues. All he ac o s we e deemed o be
equally impo an .
Fo he “Human e o p incipal ac o ,” he esul s showed “The
ope a o ’s inabili y o ecognize he ope a ional condi ion, sys em s a-
us, o measu ing ins umen s a us” (0.651), and he “Ope a o ’s
inadequa e pe o mance due o high wo kload o wo k dissa is ac ion”
(0.349). “The ope a o ’s inabili y o ecognize he ope a ional condi-
ion, sys em s a us, o measu ing ins umen s a us” weigh o he ac o
is nea ly wice as high as he second-place ac o .
The e o e, e o s in sys em-wide si ua ional judgmen s a e di ec ly
ela ed o isk.
A ending o “Sys em e o o LNG supply & ecei ing p incipal
ac o esul s,” “Ope a ion o he ESD Sys em” (0.296), “Pipeline
leakage” (0.272), “Elec ici y blackou , communica ion e o ” (0.228),
and “O e p essu e/o e hea ing o equipmen ” (0.204) anked closely.
Fu he mo e, he ac o weigh alues om 1s o 4 h place showed no
signi ican di e ences. As o he de ailed ac o s co esponding o he
“Sys em e o o LNG supply & ecei ing essel e o ,” he o e all isk is
simila o he o he de ailed ac o s o his ca ego y.
Las ly, he esul s ela ed o he p incipal ac o “Wo k p ocess e o ”
we e in he ollowing o de : “Flame igni ion” (0.245), “O e - ueling
(only o ueled ships)” (0.195), and “Inadequa e pu ging o he loading
a m” (0192). Excep o he highes - anked ac o , “Flame igni ion,” he
weigh alues o he emaining ac o s a e simila . This ac o ’s s and-
ou migh be a ibu ed o LNG’s inhe en isk o i e owing o he na-
u e o gas uels.
Table 6 p esen s he esul s o global impo ance, highligh ing he
o e all analysis alue o he ac o s ex ac ed in his s udy. Global
impo ance can be de i ed by mul iplying he impo ance alue o he
co esponding p incipal ac o by each de ailed ac o and he impo -
ance alue o he de ailed ac o o examine he comp ehensi e esul .
Fo example, he global impo ance alue o he i s de ailed ac o , he
“Ship pump mal unc ion,” is 0.226, and he p incipal ac o o which he
ac o belongs is he componen e o , whose alue is 0.232. The e o e,
he p incipal ac o *de ailed ac o (0.226 *0.232) gene a es a global
impo ance o 0.053 o he “Ship pump mal unc ion” de ailed ac o .
Global impo ance is ega ded as he mos impo an esul o a CFPR
analysis because i iden i ies he o e all anking o all ac o s.
Rega ding he global impo ance analysis esul s, “The ope a o ’s
inabili y o ecognize he ope a ional condi ion, sys em s a us, o
measu ing ins umen s a us” was anked i s a 0.112. The de ailed
ac o alls unde he p incipal ac o o human e o , a opic ex ensi ely
co e ed in logis ics isk- ela ed s udies. Despi e he ise o Logis ics 4.0
and he ac i a ion o au oma ion sys ems in he logis ics sec o , ce ain
aspec s s ill equi e human in e en ion (Cimini e al., 2021). Acco d-
ingly, he occu ence o human e o in he LNG bunke ing a ea needs o
be add essed. Gius i e al. (2019) conduc ed a s udy o educe human
e o s when handling ai ca go. S e ano a (2021) in es iga ed ac o s
conce ning he human e o ha occu ed in he logis ics indus y and
s udied he eason o he occu ence. Thus, despi e ad ancemen s and
echnological changes, we ound ha “human e o ” emains signi ican
in logis ics, highligh ing he necessi y o enhance de ailed logis ics se -
ices in a cons an ly e ol ing ex e nal en i onmen . The i s and i h
ac o s a e human- ela ed isk ac o s. Al hough imp o emen me hods
ha e been applied, he ecu ence o his ac o indica es i s ine ec-
i eness as a solu ion; he e o e, i appea s o be a high- anking isk
ac o .
In second place is “Con ol equipmen mal unc ion” (0.073), which
belongs o he “Componen e o .” This means ha he second-place
ac o indica o s ha con ol equipmen ope a ions occu p incipally.
Kade e al. (2015) men ioned ha i is necessa y o se elemen s o a
sa e y model du ing LNG ope a ion by se ing he componen s equi ed
o LNG unloading. Pio and Salzano (2018) calcula ed he e iciency by
de i ing componen s o conside lammabili y, which is a cha ac e is ic
o LNG. The e o e, se ing he isk zone du ing LNG bunke ing ope a-
ions and con olling he LNG gas ou low is impo an . Mal unc ioning
o con ol equipmen may ha e a isen as an impo an p io i y because
o i s di ec ela ionship wi h acciden s.
In hi d place is he “Ope a ion o he ESD Sys em” (0.071). The ESD
ep esen s he au oma ic shu down o he sys em in he case o an
eme gency du ing LNG ope a ions. As LNG is a combus ible ma e ial,
he e is always he possibili y o i e and gas leakage (Gopalaswami
e al., 2017). This de ailed ac o belongs o he “Sys em e o o LNG
supply & ecei ing essel” (0.239) and is an impo an isk ac o ha
should be p io i ized o LNG bunke ing wo k.
6. Conclusion
Following he IMO and EU en i onmen al egula ions, e o s in he
shipping sec o con inue o mo e owa d achie ing ne ze o (ze o ca -
bon emissions) by 2050. Acco dingly, a ious policies o educe ca bon
emissions ha e been implemen ed, and u he measu es a e being dis-
cussed. Some o hese al e na i es include slowing down he ope a ion
o ships, coa ing he hull, and imp o ing equipmen e iciency. How-
e e , o achie e a comp ehensi e educ ion in ca bon emissions, i is
impe a i e o add ess ship ueling due o i s signi ican con ibu ion o
o e all ca bon emissions. Consequen ly, LNG’s ole as a b idge uel has
gained p ominence, gi en i s e ec i eness in aiding he ansi ion o-
wa d ne -ze o ca bon emissions. The e o e, conside ing LNG’s po en ial
as a b idge uel, in e na ional s udies on his opic a e lou ishing.
Howe e , his opic emains ela i ely unexplo ed in Ko ean domes ic
s udies. Hence, his s udy iden i ies and p io i izes he isk ac o s
associa ed wi h STS ac i i ies du ing LNG bunke ing in Ko ea. Building
on p e ious esea ch and expe su eys, his s udy ga he ed 11 expe
su eys. These su eys we e based on i e p incipal ac o s and 19
de ailed ac o s, which we e hen analyzed using he CFPR
me hodology.
Following he esul s o global impo ance, “The ope a o ’s inabili y
Y.-S. Choi e al.
The Asian Jou nal o Shipping and Logis ics 41 (2025) 52–60
59
