SOLAS 2020

Damage Stability Regulations

Keith W Hutchinson, Babcock International Group, UK

Andrew L Scott, Marine and Coastguard Agency, UK

SOLAS 2020 Damage Stability Regulatrons Hutchinson and Scott

The Western Europe Sect·on of t11e Soaety of Naval Architects and Manne Engineers 1
Lloyds Register of Sh1pp1ng London United Kingdoll' 6 Februaiy 2018
- - - - -

Disclaimer

The views expressed in this presentation are

those of the presenters and do not
necessarily represent those of the
organisations to which they belong or the
professional institutions of which they are
members

SOLAS 2020 Damage Stab1l1ty Regulauons Hutcrinson and Scott

rre Western Europe Seeton of the Society of Naval Architects and Manne Engineers 2
Lloyds Register of Shipping London United Kingdom 6th February 2018
 -

- c='" - - - = = - - - ' --

Presentation
• Some nomenclature etc. and issues with SOLAS 90 and SOLAS 2009
• A century of incidents regarding loss of I uncertainty in stability
• Development of Statutory Rules and Regulations, current IMO Damage
• Development of Passenger Ship Residual Damage Stability Standards
• Issues regarding SOLAS 2009 and those relating to Passenger Ships
• Recent incidents
• Issues with RO-RO Passenger (RO-PAX) Ships
• Time-line Amendments to SOLAS 2009 RO-PAX Damage Stability Regs
• Proposed changes to R index and s finat; factor as they effect RO-PAXs
'
• Likely effect of potential regulation changes on RO-PAX design
• Stockholm Agreement
• Conclusions
SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scott
ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 3
Lloyds Register of Shipping Londo/"\ United Kingdom 6th February 2018
 Some nomenclature etc.
and issues with
SOLAS 90 (deterministic)
and
SOLAS 2009 (probabilistic)

SOLAS 2020 Damage Stab ty Reg t ons H :ison and Scott

West n Europe SectJon of the Soaety of ava Atchlt cts and Mannf' fng neer 4
Lloyd Reg ster of Sh ilP ig London Ur t d Kmgdom ~ February 2018
 -

- c='" - - - = = - - - ' --

Frequently used Terms and Expressions

• SOLAS - Safety of Life at Sea Convention. We will focus only on

Chapter 11-1 Parts 8 , 8-1 , 8-2, 8-3 and 84 (subdivision and stability)
• SOLAS 2009 - the current version applicable to new ships constructed
on or after 1st January 2009, and based on probabilistic damage
• SOLAS 2020 - the amended version which will be applicable to new
ships constructed on or after 1st January 2020, on probabilistic damage
• SOLAS 90 - the last version of SOLAS based on deterministic damage
stability applicable to ships constructed prior to 1st January 2009
• StockholmAgreement - specifiicallyfor EU-flagged RO-PAX ships

SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scott

ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers
Lloyds Register of Shipping London United Kingdom 6th February 2018
 -

- c='" - - - = = - - - ' --

Some Probabilistic Terminology (1)

• A -Attained subdivision index= 0.4As + 0.4Ap + 0.2A/where As,

Ap and Al are partial indices calculated at 3 draughts at assumed
KG(s (GMr's)
• Each partial index is a summation of contributions from all damage
cases taken into consideration, using the formula A =L Pi * si
i represents each compartment or group of compartments being
considered
P; accounts for the probability that only the compartment or group of
compartments under consideration may be flooded , disregarding any
horizontal subdivision
s; accounts for the probability of survival after flooding the compartment
or group of compartments under consideration , including any horizontal
subdivision
SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scott
ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 6
Lloyds Register of Shipping London United Kingdom 6th February 2018
 -

- c='" - - - = = - - - ' --

Some Probabilistic Terminology (2)

• R - Required subdivision index; the subdivision of a ship is

considered sufficient if A > Rand if, in addition , the partial indices
are not less that 0.9R for passenger ships and 0.5R for cargo ships.
• R is obtained from empirical formulae with separate values for:
cargo ships greater than 100 metres in length (Ls)
cargo ships with 80 < Ls < 100 metres in length
passenger ships > 24 metres Ls in length

• SOLAS 11-1 only covers above passenger ships and dry cargo ships.
Tankers, SPS ships, OSV's have their own damage regulations

SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scott

ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 7
Lloyds Register of Shipping London United Kingdom 6th February 2018
- .=-- - - - =- -

Some Probabilistic Terminology (3)

• Minor damage - refers to Part 8-1 Regulation 8 which applies to

passenger ships carrying more than 36 passengers and requires
that they comply with specified s values (depending on number of
persons carried) after suffering damage with limited penetration and
longitudinal extent. Designed to prevent loss after minor damage
• Safe return to Port (SrtP) in Part 8-1 Regulation 8-1 requires that
passenger ships with Ls > 120 metres or MVZ > 3 shall maintain
functionality of certain key operating systems (propulsion, navigation
etc.) following damage to any single watertight compartment

SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scott

ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 8
Lloyds Register of Shipping London United Kingdom 6th February 2018
 -

- c='" - - - = = - - - ' --

What was wrong with SOLAS 90?

• Damage extents were limited to 2 or 3 compartments longitudinally

and 8/ 5 transversely leading to SOLAS 90 'designs' which were built
to these limits. The result of exceeding these limits- unknown
• The subdivision regulations, with terms such as criterion of
service, margin line, floodable length and permissible length
were regarded as old-fashioned and rather obscure, having been in
place since the first SOLAS convention following the loss of the
RMS Titanic

SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scott

ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers q
Lloyds Register of Shipping London United Kingdom 6th February 2018
What was wrong with SOLAS 2009?

• Equivalence with SOLAS 90 did not include Water on Deck for

RO-PAX
• RO-PAX optimised to meet SOLAS 2009 were found to be vulnerable to
capsize following plausible damage scenarios, even in calm water
• R index was too low -too many damage cases were permitted to
result in loss
• The application in Part B to Tankers etc . was confined to a footnote - not
legally enforceable
• The Explanatory Notes required enhancement

SOLAS 2020 Damage Stability Regula!lons Hutchinson and Scott

rile Western Europe SE"c!lon of the Sooety of Naval Architects and M ne Eng neers 10
Lloyd Reg ster of Shlppmg London United Kingdom Februarv 2018
A century of incidents regarding
loss of I uncertainty in stability

SOLAS 2020 Damage Stab ty Reg t ons H :ison and Scott

West n Europe SectJon of the Soaety of ava Atchlt cts and Mannf' fng neer 11
Lloyd Reg ster of Sh ilP ig London Ur t d Kmgdom ~ February 2018
 -

- c='" - - - = = - - - ' --

SS Principessa Jo/andia capsizing directly following

launch from Genoa, 22nd September 1907
141 m, 9,21 OGRT transatlantic ocean liner built in Genoa for Navigazone
Generale ltaliana (NGI ). Upon launch she was almost fully completed ,
even with all fittings and
furniture installed! Due to this, •
launch condition had
insufficient stability. She
heeled sharply to port,
movable fittings shifted
increasing her list further and
she began to down-flood
within 20 minutes and shortly
after capsized , fortunately
without loss of life
SOLAS 2020 Damage Stability Regulations Hutchinson and Scott
ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 12
Lloyds Register of Shipping London United Kingdom 6th February 2018
- -- - -- -

USS Lafayette, ex SS Normandie, capsized in New

York harbour following fire, 1Qth February 1942
313.6m 83,423GRT transatlantic ocean liner
built in 1935 for Compagnie Generale
Transatlantique (CGT). Whilst undergoing
conversion to a troopship in New York
harbour she caught fire and fireboats and
shore appliances poured water on and she
began to heel to port and consequently
down-flood and capsized in the early hours
of the morning with the death of 1 worker.
She was finally re-floated in August 1943
and was finally broken up between 1946
and 1948

SOLAS 2020 Damage Stab1I ty Regula!Jons Hutchinson and S ott

rile Western Eu ope SE"cilon of the Sooety of Nava Architects and M e Eng neers 13
Lloyd Reg ster of Shipp ng London u ted Kingdom F bruarv 2018
SS Flying Enterprise sinking in Western
Approaches, January 1952
WWII C1-B, 120.8m, 6,711GT, general cargo ship built in 1944. On night
of 25th December 1951 she encountered a storm which caused the cargo
to shift and on 28 December she
was listing 45deg to port On 29th
December all passengers and crew,
bar the captain , were evacuated ,
with the loss of 1 life. A tow line was
attached on 5th January but on 1Qth
the line parted , with increasing list
her captain and tug crewman
abandoned ship mid-afternoon and
she sank an hour later only 41
nautical miles from Falmouth
;::, LA'> l020 Damage Stab 1ty Regu tions Hutchinson and S ott
rile Western Eu ope SE"ction of the Sooety of Nava Architects and M e Eng neers 14
Lloyd Reg ster of Shipp ng London u ted Kingdom F bruarv 2018
MV Lairdsfield sank off the Tees, 6 th February
1970, with the loss of all 10 crew
53.2m, 522GRT single hold coaster built in 1953. She had loaded 373t of
hollow hexagonal steel columns, tiered between dunnage, and 354t of
steel plates on top, at Middlesbrough . As a result of these loading
arrangements the deadweight KG,
was higher than the ship was
designed for and hence, whilst she
did have a positive GMr , her
stability was inadequate for sailing ~~~~
in open water. Hence, when
leaving the River Tees in moderate
weather, possibly whilst executing
a turn , she suddenly capsized
without any distress signal
.., S 2020 Damage Stab 1ty Regu !Ions Hutchinson and S ott
rile Western Eu ope SE"cilon of the Sooety of Nava Architects and M e Eng neers 15
Lloyd Reg ster of Shipp ng London u ted Kingdom F bruarv 2018
- .=-- - - - =- -

MS Herald of Free Enterprise capsized off Zeebrugge

6th March 1987, with loss of 193 passengers & crew

131.9m, 13,601GT, RO-RO car and passengerferry built in 1980. Just

after leaving the harbour at Zeebrugge, Belgium, en-route to Dover,
United Kingdom, with 539
passengers and crew
on-board , water began to
enter the car deck due to
her bow door still being
open . As a consequence
of this Water-on-Deck
(WoD) and resultant free
surface, her GMr was
diminished and she
capsized within minutes
SOLAS 2020 Damage Stability Regulations Hutchinson and Scott
ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 16
Lloyds Register of Shipping London United Kingdom 6th February 2018
 -

- c='" - - - = = - - - ' --

MS Estonia sank in the Baltic Sea, 28th September

1994, with the loss of all 989 passengers and crew
157m,
15,566GT,
RO-RO car
and
passenger
ferry built
in 1980,
which sank
en-route ..._-.;
Tallinn ,
Estonia, to
Stockholm, =:;
Sweden
SOLAS 2020 Damage Stability Regulations Hutchinson and Scott
ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 17
Lloyds Register of Shipping London United Kingdom 6th February 2018
- .=-- - - - =- -

MS al-Salam Boccaccio 98 sank in the Red Sea, 3 rd

February 2006 , with the loss of 1,031 lives
131 m, 11 ,799GT, RO-RO car and passenger ferry built in 1970. En-route
from Duba , Saudi Arabia, to Safaga, Egypt, with 1,418 passengers and
crew on-board , an uncontrollable fire broke out. As a consequence,
exacerbated by the scuppers not
functioning correctly, firewater
accumulated , hence resulting
in the ship becoming unstable. ___,. ~
Due to this, and possibly also
the weather conditions, she
began to list excessively and
as a result began to down-flood
and subsequently sank

SOLAS 2020 Damage Stability Regulatmns Hutchinson and Scott

ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers
18
Lloyds Register of Shipping Londo/"\ United Kingdom 6th February 2018
 -

- c='" - - - = = - - - ' --

AHTS MV Bourbon Dolphin capsized off the west

coast of Shetland, 12th April 2007, with loss of 8 crew
75.2m, 2,985GT, 2006 builtAHTS. Whilst anchoring the semi-submersible
drilling platform Transocean
Rather, in deep water, a
heavy anchor chain , that
it was pulling , suddenly slid
across the side of the deck
and began to drag the ship
over. She capsized within
seconds, with the loss of 8
of the 15 crew on-board .
The ship sank 3 days later

SOLAS 2020 Damage Stability Regulatrons Hutchinson and Scott

ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 1Q
Lloyds Register of Shipping London United Kingdom 6th February 2018
MV Deneb capsized alongside the quay in
Algeciras, Spain, 11th June 2011
101 .1m, 3,992GT, 1992 built feeder container ship which capsized whilst
loading cargo. Upon loading a 40' container at height, the ship began to
list to starboard . The ship initially heeled slowly, but progressively more
rapidly, until at 45 degs the
containers contacted with the
quayside. Harbour tugs were
employed to push the ship in
to the quay in order to prevent
total capsize. She began to
down-flood and within hours
the ship was resting on the
bottom at 54 degs, finally
settling at 75 degs
SOLAS 2020 Damage Stability Regulallons Hutchinson and Scott
rile Western Europe SE"cilon of the Sooety of Naval Architects and M ne Eng neers 20
Lloyd Reg ster of Shipping London United Kingdom Februarv 2018
 -

- c='" - - - = = - - - ' --

MS Sewol capsized and sinking off Jeju, Korea, 16th

June 2014, with the loss of 304 passengers and crew
145.6m, 6,835GT, 1994 built ferry. With 476 passengers on-board , she
rapidly developed a 22 deg port list after executing a starboard turn due
to insufficient GMr. Cargo shifted , she began to down-flood and within 10
minutes she was at 30 degs and
sank 3 hours later within a mile of
land . It was subsequently estimated
-
that she was loaded with twice the
legal amount of cargo and hence
without sufficient water ballast to
counteract the deadweight KG,
which exacerbated the increases
in lightweight and KG,

SOLAS 2020 Damage Stab1l1ty Regulations Hutchinson and Scott

The Western Europe Sect·on of the Society of Naval Architects and Manne Engineers 21
Lloyds Register of Sh1pp1ng London United Kingdoll' 6 Februaiy 2018
Car Carrier MV Hoegh Osaka grounded in the
Solent, 3 rd January 2015
179.9m, 51 ,770GT, 2000 built PCTC. After departing Southampton, she
developed a severe list after executing a port turn at 12 knots due to her
GMr whilst being inadequate. The upper vehicle decks were full whilst
the lower decks were lightly
loaded which , together with
the state of the bunker oil
low down in the ship being
light, resulted in a high KG,.
She lost steerage and
propulsion and grounded on
- "

Bramble Bank, preventing

capsizing and settled at
52 degs to starboard
SOLAS 2020 Damage Stab Illy Regu tions Hutchinson and Scott
rile Western Eu ope Sf'cilon of the Soaety of Nava Architects and M e Eng neers 22
Lloyd Reg ster of Shipp ng London u ted Kingdom F bruarv 2018
Development of Statutory
Rules and Regulations,
current IMO Damage

SOLAS 2020 Damage Stab ty R g I ons Hut :ison and Scott

West n Europe SectJon of the Soaety of ava Atchlt cts and Mannf' fng neer 23
Lloyd Reg ster of Sh ilP ig London Ur t d Kmgdom ~ February 2018
- - - _- - ~

Outline structure International Maritime Organization

AS~DIJILT
C..U..t .OW.•••f-IUl9
W..•~-.,,,._ .. ti.. b ........_,,.. _ _ .. -fliCT
.\J9C1wt . . Wark~"'*' •"""'1 • ..._ a...s.i - ......
tloc> .. _ . .

..,,'1._..,.,...,i.,.... .._.....,._.....i-...... ,,..,,.,

COOMGIL

...... _...,...._....._ .........

~"'o.c-oru.r>. •JIClldlil"Gt11pdNVI(
.......n1..__ ~~
,~
---11
.....-_._..
,___..,...., .... ~ ....... £~
~a..wo.....-.17_.w...,.s ....

n.w.aa-~~C.--.(W!X)
,..,..,.... ... Co~ollbn.o.

CQNSAI.
,.,
SM~.,._~s....,
W•C.---M•
UllS:...di.•t.-.. ...ir..
..,.v-~Sdhr

SOLAS 2020 Damage Stability Regulauons Hutchinson and Scott

ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 24
Lloyds Register of Shipping London United Kingdom 6th February 2018
 -

- c='" - - - = = - - - ' --

International Maritime Organisation and Stability

• Special Agency of UN , established 1948, first met in 1959

• 172 Member States, 3 Associate Members
• MSC responsible for safety issues, meets for 8 days twice a year
• Technical matters devolved to Sub-Committees, stability SOC (SLF)
• SOC meets annually for 1 week discussing papers and issues
• Plenary Session and Specialized Working I Drafting Groups
• Draft Amendments or Proposals for further work, research etc.
• SOC reports considered by MSC , proposals new legislation
• IMO Regulations implemented by Flag States into their law
SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scott
ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 25
Lloyds Register of Shipping London United Kingdom 6th February 2018
 -

- c='" - - - = = - - - ' --

IMO Damage Stability Regulations

• Safety of Life at Sea (SOLAS 2009)

• Prevention of Pollution from Ships (Annex 1 to MARPOL 73/78)
• International Bulk Chemical Code (IBC 2007)
• International Gas Carrier Code (IGC 1993 I 2016)
• Offshore Supply Vessel Code (OSV 2006)
• Special Purpose Ship Code (SPS 2008)
• International Convention on Load Lines (ICLL 1966 I 1988)
• Mobile Offshore Drilling Unit Code (MODU 2001)
• Stockholm Agreement (1997 - EU countries only)
SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scott
ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 26
Lloyds Register of Shipping London United Kingdom 6th February 2018
Development of Passenger Ship Residual
Damage Stability Standards since 1960

SOLAS 2020 Damage Stab ty Reg t ons H :ison and Scott

West n Europe SectJon of the Soaety of ava Atchlt cts and Mannf' fng neer 27
Lloyd Reg ster of Sh ilP ig London Ur t d Kmgdom ~ February 2018
- ---- - -
c '"
""'
~=
Comparison of : :'IQ

Damage ....,..
: !')O

01-e

Residual i!
0 ,.,

.. 0 •9"0

Stability i 0110
~ 01))
"
standards i~ c0 ,.:
'"'

~ a •:c
•
from 1960 ~,,.

• :io:

to date ',.., " •

OllO
0,., -
S~IC

(Final Stage 004'

fi~ -· ·!.'

0 °"'
of Flooding) 0 <:!')
0010
------
0

,.
S t ons H ch :ison and Scott
West n Europe SectJon of the Soae of a At t cts and Mann fng n er 28
lJ yd Reg ste of Sh ilP ig Lond n U t d Kmgdom February 2018
 -

- c='" - - - = = - - - ' --

Comparison of Damage Residual Stability standards

from 1960 to date (Final Stage of Flooding)

DETERMINISTIC PROBABI LISTI C

IMO IMO IMO IMO SOlAS2020
IMO UK 0
CRI TE RION UNI TS n Sa.ASOO SOLAS2009 SOLAS2009 RO-RO
SOLAS60/7a STABBO Sa.AS
CARGO CARGO PASSENGER PASSEtJGER
~:s >.·: us..:. L~-r""-~ -E\:S'i. ~;:i
,: _q;. E 'l. .~E ~.
.:. E~ ,.YT :;;:a: • :g .. "'
1 0:~,.2!~
:!' ~a;-.·ca-.s;.::~
3Y1!:.,.... " • "
'ft:.o\ .t..::Eli --.::: :l::C-"tl."£_ ~
·~
"t\E
>;;~ ~· ~ · :g ..:r :.~SC'"-D r ei :!.;hl .. tl :,..
!~C1 n: (t•('
~ -•t "-£:: .:>-~~ --~ _£>Uta:....... llE,-Ui c f ;t c :.-:-c
. "~· ~ lh: ~" -·
' -<
·~ '"' - -=---'"- ,_
'"-.C.~ :: ' : ¥ra::;:£::_•_~,.:.N
- =~'i:il
,. ,. ! . : OG~Al .:):'.e.''! ~42
:1~ · : C7:3'.U::•'1:t:-·;~
\I ' ' ... :..>
• 'Ott.I J ' 3C
... .....
., ~~1!
_. -
~'O";:-:£ ·~--;ii;,._-L ~ C' ~

DC:"!\:! us.:.-.~ 1!£""...c::o.r~ - : 1-- 0r • .- !~E ~:JC ~ct.: .N.>: t ,)

SOLAS 2020 Damage Stability Regulations Hutcrinson and Scott

ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers
29
Lloyds Register of Shipping London United Kingdom 6th February 2018
 SOcAS :rio RO 00
=A SSl'•GE. It $M PS 0~, ..Y ll'ECF C.A.l.L'Y FCFI'
" J..:iE CASES WitCH •,• Q..: £A R O -~
~""' -llO~=.-.CE

Comparison of CZ"
) liC
,,.,.,.--f~'',
Damage ~f~
/
/ ~· ''
', : ...
I~
•I ' •'
.
....... \

Residual .... )

' I '
I ' \\
.. \

Stability _,
: •.t:

- :•l: I
I ' COJ.: ;j~ Ah:!::::-~ ' •\
0: I c;.t,e~ ..... c •.sso.~u ..•s ,
standards Ci : •::
If J, ,.~..
, """-..',. ' ' •\

'.
!., c ··~ .j

I '
:"'1"> '

for :s
el l •:):. / / --
~·::~rs ' '\ '\

SOLAS 2009 !
::l

~
)~

':'">)
) ~·:
., '
'

l• I I
~
I
,
I/ \
\
',
\
\
.
'
''
•,
"'
0 )!3 I 1 \ \
and O*l
scv.s
:~A\i"'Or:' I)
CARGO I
•, • I
I
\\
I
'\
'
f I \ \
6 :~CEGREES

SOLAS 2020 ~~4)

"" salAs
OJJ.l
J.:I C-E3-1"£ES

#
• I
( I
f I
\
\
\
\
\
\
t
l
i:mAAtHnJ f l \ '

(Final Stage J:m

Pi\SS: EhGE~
O·, tO ......, • DEGREES.
g, v f! OEGR:i S # \ \

of Flooding) 0
~NG .E C• l~Q.ll<ATIO~ 8 OEG•EES 1\
I,
,,.•
I

---------------- ~-ci~~es----------------~
I
I
I
d I
-------------------- -1~&-~u---- ------------------ ~
SOLA ~ ge Stab1I ty Regu !Jons Hutch on and S ott
rile Western Eu ope SE"c!lon of the Sooety of Nava Architects and M e Eng neers 30
Lloyd Reg ster of Shipp ng London u ted Kingdom F bruarv 2018
 .

- c='" - - - = = - - - ' --

Comparison of Damage Residual Stability standards

SOLAS 2009 and 2020 (Final Stage of Flooding)

SOLAS2009 SOLAS 2009 SOLAS 2020

CRITERION UNITS AND 2020 AND 2020 PASSENGER
CARGO PASSENGER RO-RO SPACES

POSITIVERESIDUAL RIGHTING LEVER (GZ)

CURVERANGE, 6,,. DEGREES 16 20

IJAXll.1Ul.I RESIDU.l\L RIGHTING LEVER GZl.'Al( l.1ETRES 0.120 0 200

ANGLEOFHEEL DUE TO UNSYt.11.1ETRICAL MINll.1Ul.1 7

FLOODING AFTER EQUALISATION DEGREES MINIMUM 25
(DAl.1AGE EOUIU SRI Ul.1 ANGLE). 11, M>.X IM lJA 3J MA.XIMUl.I 15

SOLAS 2020 Damage Stability Regulauons Hutchinson and Scott

ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 31
Lloyds Register of Shipping London United Kingdom 6th February 2018
Issues regarding SOLAS 2009 and
those relating to Passenger Ships

SOLAS 2020 Damage Stab ty Reg t ons H :ison and Scott

West n Europe SectJon of the Soaety of ava Atchlt cts and Mannf' fng neer 32
Lloyd Reg ster of Sh ilP ig London Ur t d Kmgdom ~ February 2018
SOLAS 2009 issues needing resolution at IMO

• Major issues with RO-RO damage stability-Rand s

• degree of applicability to smaller cargo ships
• the provision of information to the Master
• circumstances which would allow exemption from fitting a double
bottom
• whether ramps on the RO-RO deck should be made watertight
• treatment of ships with unusual waterline endings
• how to treat partial bulkheads I webs on the bulkhead deck now
that it may be flooded at equilibrium

SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scott

ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 33
Lloyds Register of Shipping London United Kingdom 6th February 2018
 .

- c='" - - - = = - - - ' --

Definition of s factor

• EN4 Reg. 7.1 "The s factor is dependent on the calculated survivability

of the ship after the considered damage for a specific initial condition"
I

• SOLAS Regulation 7-2.3 s .= K· ( GZmax. Range) :i

final,1 O.1 2 16

• Derived from Hs . = 4 _(GZmax. Range] and F

cnt 0.1 2 16 Hs callis10n

i.e. relationship between the parameters and the Hscrit(causing capsize

during the 30 minutes model tests with about 50% survival ratio) and
cumulative distribution function of the sea states recorded at the instant
of collision assuming H scrit is approximately equal to H scollision
SOLAS .!020 Damage Stability Regulauons Hutchinson and Scott
ftie Wes1em Europe Seeton of the Society of Naval Architects and Manne Engineers 34
Lloyds Register of Shipping London United Kingdom 6th February 2018
 - .=-- - - - =- -

Definition of s factor: CDF of Hs at instant of collision

0.9
1.0

, v
/ -- • The foregoing explains the s
factor in more detail and the
fact that the critical sea state
0.8

0.7
/ H scrit represents survival in
only 50°/o of the 30 minute
II
........
C/l
UJ
er: 0.6 model test results is not
[;j
~
.._, 0.5
obvious from SOLAS or
•
•
l"" 0.4 I Explanatory Notes
~ I • ENs for Regulation 7.1 or 7 .2
0.3
should be enhanced to give
0.2 explanation of the basis
0.1 behind the s formulae, as with
Explanatory Notes for IMO
0 Resolution A.265(VI11)
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
~ S0LAS ,a:: 1 Damage Slab1l1ty Regul.:l!ons Hutchinson and Scutt
.. ~
~u
ftie Western Europe Sectnn of the Society of Naval Architects and Manne Engineers 35
Lloyds Register of Shipping London United Kingdom 6th February 2018
Critical KG, or GMT for Probabilistic Criterion: Curves
CRITICAL KG, CU RVES

--- - UNSl ~BLE

·~

~ ...... '
,, ___
,_

l.-9 ~ I
~-

.,
v - ~
1.
~
··~'
STAE ~E

I~T .
' ::l l •

UBOMSION
OAOUW
DRAUGHT EXTREME (METR ES)

-1neac1c.-KG · OS-bythe~ lntad CrcQll KG. 1 5 MW.. by h Stem

lnlaa e r - KG· E"8'1 l<MI lntad Cttal KG· 20MW.. by1he St""'
1n1ac:t er-KG 0 5 MWes by.,,_ Stem lnlac:t Creal KG- 2 5 , _. . by IN Stem
1n1ac:t er-KG I 0 - by 11'9 Siem _ _ _,Cr&icalKG nacx:ocllia..-.
,,,_.sic o..n.ge~
S LA 2020 uamage Stab Regu ns H ch nson and Scott
Th8 Weste n Europe Se on of the Sooety of N Archrtects and Manne Eng ne rs
3
Lloyd Reg ster of Shipp ng London Un ted Kingdom Februaty 2o•e
 -

- c='" - - - = = - - - ' --

Critical KG, or GMTr for Probabilistic Criterion

• Is it technically legitimate critical KG, or GMT, for A >= R?

• Misleading for the Master to assume compliance with probabilistic
CCKG, ensures survival of almost unlimited damage as 1-A cases
permitted to result in capsize or sinking- possibly rapid in RO-ROs
• Master has little idea which are survivable damages and which are not
• A relatively minor damage on a ship observing the GMT,, Tand t
restrictions could still result in a rapid loss and hence high loss of life
• Helpful to the Master only to include the combined critical KG, or GMT,
criteria which are based on physical phenomena such as GZ, e, GMT,
• A >= R simply quoted in SIB as AIR for particular selected T, t, KG, to
show the Master those which are the most vulnerable sailing conditions
SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scutt
ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers
37
Lloyds Register of Shipping London United Kingdom 6th February 2018
 -

- c='" - - - = = - - - ' --

Variability of survivability over a ship's length

• SOLAS 90 was deterministic and included floodable length

• SOLAS 2009 different ships may have the same A but may have quite
different capabilities for withstanding hull damage over their length
• Distribution of survivability should be introduced so no part of the ship
is left unacceptably vulnerable to flooding hence capsize or sinking
• Could be facilitated by the application of 'local (partial) indices of
subdivision' with values maintained at I within some limits
• Suggested that two, or more, compartment local indices, Bj, are
considered as these are representative of a worst reality
• Suggested also one compartment local indices, Aj, or 'minor damage
indices' are also considered as such damages are probable
SOLAS 2020 Damage Stability Regul.:uons Hutchinson and Scott
ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers
38
Lloyds Register of Shipping London United Kingdom 6th February 2018
 -

- c='" - - - = = - - - ' --

Presentation of survivability: Information on-board

• SIB , Capacity Plan , Calibrations, Damage Control , Freeboard Plan

• Presentation of CCKG, or CCGMr data under consideration at IMO
- must be easy to use but reliable
- cover operational range of ship ( T, t etc .)
• 'Approachable' - clear, consistent, concise, usable, comprehensive
- 'simplified stability information' - Cross Curves, Wind data?
• Damage Control Plan I Book using the Probabilistic Approach
• Damage Consequence Diagrams 'Carpet Plots' (T, t, assumed KG,)

SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scott

ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 39
Lloyds Register of Shipping London United Kingdom 6th February 2018
- .=-- - - - =- -

Presentation of survivability: Consequence Diagram

G
......

•••~-~:::;-:-I ..
-'"" '._.... ,"""'--.........,
. ~.J1.; ;1; :,I.; : :t1:!r~·"'l"W',~ -
:;:;;;;;,;;;;;;,,wii;uqwco;;;;;u

SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scott

ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 4(1
Lloyds Register of Shipping London United Kingdom 6th February 2018
- .=-- - - - =- -

'Extant' Survivability: Consequence Diagrams

• Long advocated both as an aid to the Master in the event of damage

and for crew training purposes
• UK has permitted , if not actively encouraged , their provision on-board
• SLF 55 debated whether provided the legal means for fulfilling the
requirements of SOLAS Chapter 11-1 Part 8-4 Regulation 19.5
• Revisions not been agreed - remain as one possible means of
communicating information to the Master (MSC.1 /Circ.1245 para. 5)
• Regrettable as prepared in advance using assumed T, t and KG, hence
possibility of the information provided being misleading
• For Passenger Ships, on-board damage computation is the way ahead

SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scott

ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers 41
Lloyds Register of Shipping London United Kingdom 6th February 2018
On-board Damage Computation

• SOLAS Chapter 11-1 Regulation 8-1.3 provision of damage stability

information to the Master for 'Safe Return to Port' Passenger Ships
• Propulsion , steering , navigation, internal and external communication ,
fire main , watertight doors, flooding detection, systems deemed to be
vital to damage control efforts (SOLAS Chapter 11-2 Regulation 21.4)
• Ship assumed to survive flooding - in the past any Passenger Ship
would comply with at least a 1-compartment subdivision standard ,
however the harmonized SOLAS 2009 probabilistic damage stability
regulations no longer guarantee this!
• Moves to amend the regulations so ship itself must survive flooding of
any single watertight compartment with sufficient reserve to ensure the
systems functionality- candidate criterion s-factor of 1.0
SOLAS 2020 Damage Stab1l1ty Regulauons Hutchinson and Scott
ftie Western Europe Seeton of the Society of Naval Architects and Manne Engineers
42
Lloyds Register of Shipping London United Kingdom 6th February 2018
- -- - - -c=_---= - -

On-board Damage Computation

• Distinction between a computer on-board a Passenger Ship fulfilling

requirements of 'Safe Return to Port' (SOLAS Chapter 11-1 Regulation
8-1 , Chapter 11-2 Regulation 21.4) and 'system', including shore-based
support, sensors, tank gauges etc ., providing reliable and continuously
updated information to the Master on the actual damage situation
• Agreed no pre-determined 'Safe Return to Port' stability criterion , the
'system' is not expected to process data and advise the Master whether
or not to attempt to return the ship to port after damage
• Compute and communicate the physical status of the ship to the
Master and all interested parties given multiple-compartment, non-
continuous breaches of the bottom and side shell - decide the best
course of action given sea state, wind and tide conditions, weather
forecast, availability of SAR, distance from safe haven etc.
~ s: L\S 202' Damoge Stab1 r Regul •uons Hut.:h1nsor and Scott
Manlrne& ire Western Europe Sectnn of the Society of Naval Architects and Manne Engineers 43
~ Lloyds Register of Shipping London United Kingdom 6th February 2018
 SOLAS 2020
Damage Stability Regulations

Keith W Hutchinson, Babcock International Group, UK

Andrew L Scott, Marine and Coastguard Agency, UK

SOLAS 2020 Darriage Stability Regulations. Hutchinson and Scott

The Western Europe Section of the Sooety of Naval Architects and Marne Engineers 44
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
 Recent incidents

S LAS 2020 Damage Stab I ty R u t ons Hutchinson and Scott,

The West n E ope Section of the Soo ty ofN va Arch1t d5 and Mann fng ne s 4
L yd s Reg ster of Sh pp ng Lo:ldon U t d K;igdom February 2018
 SOLAS 2020 Damage Stability Regulations. Hutchinson and Scott
The Western Europe Section of the Society of Naval Architects and Marne Engineers 46
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
__ - - -_- --

MS Costa Concordia

• 114,147 GT 13 deck Cruise Ship for 3,780 passengers and 1,100 crew
• LoA 290.2 m, Lpp247.7 m, B 35.5m, D19.77 m, T8.2 m
• 6 x 12.6 MW DGs, 2 x 21 MW FPPs, 19.6 knots
• Keel laid 3th November 2004, Delivered 30th June 2006
• SOLAS 90 Chapter 11-1 Part 8-1 Regulations 4 to 8
• 21 :45 on 13th January 2012 struck rock on port side off eastern shore
of Isola del Giglio in the Tyrrhenian Sea off the western coast of Italy
with 3,229 passengers and 1,023 crew on board
• Drifted , grounded and partially capsizing to starboard with the
(probable) loss of 32 souls
SOLAS 2020 Damage Stability Regu!at1ons Hutchinson and Scott
The Western Europe Section of the Society of Naval Architects and Marne Engineers 47
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
- -- - - - -

S LAS 2020 Damage Stab I ty R u I Hutchinson and Scott,

The West E ope Section of the Soo ofN va Arch1t d5 and Ma n fng"""""" 48
yd Reg er of Sh pp ng Lo:ldon U t d K; dom February 2018
- -- - - - -

S LAS 2020 Damage Stab I ty R u I Hutchinson and Scott,

The West E ope Section of the Soo ofN va Arch1t d5 and Ma n fng"""""" 4
yd Reg er of Sh pp ng Lo:ldon U t d K; dom February 2018
MV Baltic Ace

• 23,498 GT 8 Deck Vehicle Carrier for 2,130 cars, crew 20 to 24

• LoA 147.93 m, Lpp 134.0, B 25.0m, D11 .8 / 25.2 m, T7.9 m
• 1x9.17 SSD , 1xFPP, 18.9 knots
• Keel laid 26th February 2007, Delivered 11th July 2007
• SOLAS 90 Chapter 11-1 Part 8-1 Regulations 25-1 to 25-10
• 18:15 on 5th December2012 collided with 6,370 GT, 698 TEU ,
Lpp 126.8, B 19.4 m, D 9.45 m container ship Corvus J 30 miles off
Rotterdam in the North Sea with -1 ,400 cars and 24 crew on board
• Car decks quickly flooded and she capsized and sank within 15
minutes with the loss of 11 souls
SOLAS 2020 Damage Stability Regulations. Hutchinson and Scott
The Western Europe Section of the Society of Naval Architects and Marne Engineers 50
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
 uU'•~2020 Damage Stab R u11::i111Mc:
The op Section of the Soo 1
Reg r f Sh pp ng Lo:ldon
 Issues with
RO-RO Passenger (RO-PAX) Ships

S LAS 2020 Damage Stab I ty R u t ons Hutchinson and Scott,

The West n E ope Section of the Soo ty ofN va Arch1t d5 and Mann fng ne s 2
L yd s Reg ster of Sh pp ng Lo:ldon U t d K;igdom February 2018
 I I I
.. ,, -...
:·::. .•.•••
\
~ 1••111 \ \
'
•..
•••• I •• \ \ \
I I

1I
I I

I
.' ,
I I I I

,
••

II \\\\\\\\\\\\\\\
\\\\\\\\\\\\ " '''',;. ·..---.....;.;___
\\\\\\

- •
I ,,,.
rI I 1111111111 • •
II • • •
STENA SFllTr:JNN/CQ

S LAS 2020 Damage Stab I ty R u I Hutchinson and Scott

The est E ope Section of the Soo ofN va Arch1t d5 and Ma n fng"""""" 3
yd Reg er of Sh pp ng Lo:ldon U t d K; dom February 2018
RO-RO Passenger Ship Damage Stability

• The original brief for SOLAS 2009 was to produce new harmonized
probabilistic regulations having an equivalent level of safety to
deterministic SOLAS 90 regulations
• Rands set at same level for 'conventional' Passenger Ships
• SOLAS 90 was deterministic and included floodable length

• Following Herald of Free Enterprise and Estonia losses the

Stockholm Agreement was added to SOLAS 90 for RO-PAX
operating in (NWE) Europe but not the rest of the world
• Research showed that SOLAS 2009 does not explicitly allow for
WOO and hence is SOLAS 90 without Stockholm Agreement

SOLAS 2020 Damage Stability Regulations. Hutchinson and Scott

The Western Europe Section of the Society of Naval Architects and Marne Engineers 54
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
RO-RO Passenger Ship Damage Stability

• Many non-EU IMO Member States have taken this to mean that
WOO is NOT to be allowed for in SOLAS 2009 as it is only an EU
regional matter necessitated by the particular sea conditions and
traffic density specific to EU waters
• Such Member States have never had anything equivalent to the
Stockholm Agreement in their own national regulations
• Many EU Member States have argued that the currents factor either
takes sufficient account of WOO or that allowing for WOO has little
or no effect on A anyway and cain therefore be ignored
• Damage statistics used for SOLAS 2009 showed collisions
predominantly occur in calm waters

SOLAS 2020 Damage Stability Regulations. Hutchinson and Scott

The Western Europe Section of the Society of Naval Architects and Marne Engineers 55
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
Research on SOLAS 2009 RO-PAX designs

• May be a problem withs factor RO-PAX ships

• SOLAS 2009 RO-PAX ships with LLH can sink, rather than capsize,
in calm conditions with only 8/ 10 side penetration into the LLH
• Problem appears to be lack reserve of buoyancy (floodable length)
• Possible revision to R, s, weighting factors , text may be necessary
• SOLAS 2009's lack margin line criterion more significant to RO-PAX
than 'conventional ' passenger ships - applicability same formulae
• Possible requirement for distinction between LLH and non-LLH type
RO-PAX if critical loss mechanisms different e.g. sinkage or capsize
• Research ongoing , EU RO-PAX still require Stockholm Agreement
almost certainly prevent LLH ships with 8/ 10 longitudinal bulkheads

• ~
Mantini&

~
S< L S 20~0 Damage Stab1ht Regu!at1nns Hui ·h1rsun drv:J Sc
The Western Europe Section of the Society of Naval Architects and Marne Engineers
Lloyds Register of Sh1pp1ng London United Kirgdorr 6 February 2018
56
 lime-line regarding development,
adoption and entering in to force of
Amendments to SOLAS 2009 RO-PAX
Damage Stability Regulations

SOLAS 2020 Damage Stability Regulations. Hutchinson and Scott

The Western Europe Section of the Sooety of Naval Architects and Marne Engineers 57
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
 - _- - -_-- ---

Time-line re R index and s rinal i factor development

'

- -
.•- or t
'
~--.
, .- .
-
~
• .. ,- nt H.:< • ,..,,, ~·
' (--- '~
.:.-.-c~.;;
--- -c--.
..... . -·-
"1-V• ... :.C>c
...--.--.,...
.,.._.
~

I '-•C.
,_

~~

- ·-
-····-·4!H. ........:
~
&.••o.
~.....-.:;.,..

...~--
-~
~-
-
~:.a.
:- · ·· -·~t.:t ·······
''
-
w-'N •

-·- ' .. ... ........ - ......, ...... ...

~,
/
' . . .·~· .;.:
•'
~
~- ~-~

.. -- --
c~~-- ;;J ;•
c...,.._ I .;.;...>.; .... I cc..;.,., -~- -~
,
,..,,.
~ cc:..;; ..
;
_
fj~
- i

..
2<2
-~-·
~
~ :=.....
,_.,,
C9~1

~-
-· __
- ·- -
~llCI...:.

, . ,,.. ·-.
·--Jltt1
fl.>,,.
-·
....- i.r

-·
•'
I
-
~ ~,

!§t ~·.;;w
,
~:;; .
p ,,_::;
;,j.f',o -.---1otu
-·-u j

-... ---·-··--· .
1 ~ mo.:. ••
... v~-
I· ~~ _,__.,i
1: §F
-·-· -
;

~
M

p
. §t
~
Fl ~·
.. i

-
= '
~

"•. ;;;...
w- -~ --· ··- I
I
~- I: - ·~ ,
=· --(
- '
-·
~c
•

-·
~
-·
•-<•
--·-.
•'

...
~~

. ·-- ·--
-i------r-T---,,.=~--'---- - --
.- - -----.tK.-· - . ...
....,r ..OO
.. .
i
>-:::
~·~-
... .,_
--- •'

_----··_.
~-

<-= ..... ~ lle< l "1i( ~ ··

~

-·--
..,
=::
-
··- _
:-.........•• · ~~

,. ..··-
c.-o:..;;-...
. :al:&. OU.
--·-?J-\~Cl(lllt.o(
.;;U'#!U .
~u-;.:...-.4
_ _ _ _ _ _ ....._ _ _ _ . . .
-
........
:.i:u.4 .......
..r...., .....
I u; - .... I
I
~oc JoCl;.1
,._\;---
-~-
-... .....
..~
..,_ -..
-~oc...<4- -·-·-
·-"-
'
• :.C: ..,,. n.;

•~
--
--~

---
~-=
...u-.
... au
-
~

SOLAS 20~0 Damage St.Jh1lit Regu!at1nns Hui ·h1rsun drv:J Sec

The Western Europe Section of the Sooety of Naval Architects and Marne Engineers 58
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
Discussion of R index and Srinal i factor
1
time-line (1)

• Time-line illustrates interactions over years between IMO, EC member

states, COSS (EC shipping safety legislative body), research institutes
involved in changes couple of SOLAS Chapter 11-1 formulae but which
have significant implications for the overall safety of RO-PAXs
• Time-line shows how long it takes for amendments to stability
regulations to be introduced I enter into force due to processes
• Illustrates research work required to provide evidence to change
existing legislation- typically undertaken by individual member states
or larger supra-national entities i.e. EU , can take years and be very
complex in nature, difficult to mainage and therefore very expensive
• Unless there is a particular need to introduce new legislation quickly,
say due to a major incident at sea, major amendments take time
SOLAS 2020 Damage Stability Regu!a!1ons. Hutchinson and Scott
The Western Europe Section of the Society of Naval Architects and Marne Engineers
59
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
Discussion of R index and Srinal i factor
J
time-line (2)

• The amendments approved at MSC 96 I MSC 97, will only be

applicable to RO-PAXs constructed on I after 1st January 2020, 13
years elapsed from initial identification of the issue in 2007
• Research shown almost conclusively that the SOLAS 2009 provide an
overall safety level equivalent to SOLAS 90 deterministic standards but
without provision for WOO which can cause rapid capsize of RO-PAXs
• In European waters WOO catered for by Stockholm Agreement, likely
debate if '20/20' amendments increase R index will allow to be revoked
• Few, if any, non-European countries implemented Stockholm
Agreement or their own equivalent to it
• Concern that during this 3+ year period a large number of RO-PAXs will
have been constructed to the arguably lower 'old' standard

• ~
Mantini&

~
S< L S 20~0 Damage St 1ht Regu!at1nns Hui ·h1rsun drv:J Sc tt
The Western Europe Section or the Sooety of Naval Architects and Marne Engineers
Lloyds Register of Sh1pp1ng London United Kirgdorr 6 February 2018
60
__ - - -_- - --

Main amendments to SOLAS 2009 Chapter 11-1

Parts A , 8 , 8-1 , 8-2 , 8-4 and C (MSC.421 (98))
Regulation Subjoot Amendment Reason
A 1 2_.:. Ca~o shfp-al~atil::wu Oela.ed AR retie ro1rele~·antfa l"IE(N SOLAS2009stt~

A2 Oefini11on-i Orai.g_rL trirr\ biJi:he:eo cem mooifled I nae gs~ clarity

8• Gen.gal a pp.lica bil it)' FootnoE e.du-ding >pscif.!Ed CitgO inip!> now., Regul.:i!K>n 1e:iG Fottrotes do not n9\fe i-3"11!: leg"91 autnofrty a;. Rt:gUla.Uai t=xt
8-1f..-1 lnfctmauon b tie f\1a.s.ter UmitlrQ KG GJ.J d3ta noo't to indud'e trm Exparston9odclarificatbl pstlCUl.!lrlycfR~. B-1 f--1 ~

S-172 Attalrted ln.ou A Treatrne.rL of t Irr in celoJls!Jo-n cf A lriae9SEo.Od3i!lty

8-17.3 Raldual Gl caJa.tlatiol" Coo.s. tam o 5 p4 goement r S!"l1Choed. ff eo: tri mring added Atlded w eight me:n.od can h e used d1.11ng IS floo:dirg
9 .1 7·2.!> S ul"heac deot "'tlf p:a.i-e'"Q9' ·;.nip;. and t-e fP-'·eboa!d cea Gf argo inic.s. Clarifcetkn · ap oH s in many Rl?9lJ la ti ens
8·17-2.5.~3 :r = O
I rrme"S ion giving lmrnersso pip;e oua in carnage e.nent. ifr is'f: of j:J"ogte.s.siv e rioadirig Excsrsfc:n-8110 cistificatioli
s -11.2e5 N !!!:O{lht n;tchmves Ref;rB'tzr: ES -;.m:in \l.'at&ftr;n llstai awe;. reoo~ 'Smalr tlldefine.d. w atstt:Qtl 11& eppnea ts gll harm m."'-"i
6 ·181 and2 e a<e e.nd damage pe.isengerships' Teet O'Ull"tQ$ 3flj ho.ot le 8CCO.H'4 fa trim Clsriflcatk1'lof ecten1off0te end damage
8-183 Extent cf d iWT'l&.ge .p:as:s: rt;i er l nips,) RsmO\oa r eferences ~ Nana L::. r'CN pcsrsorts ca!f1eo ~ns e quenf.AI t.Jpon cha~es to Rform;la ln Rieg 8-1 C2'3
8-29'3 Sl!'ldf..8 Daub.le ooOC:lcms Small w ell:. VB"tical came~ extelt, C8!p0"$hip'.i <Stl rrET5S ExP& rs ion a no dsrifaiion
S -2 12~ 1 roes Pe.sit bukl'-e!!d Fe.n5tr9tlcn pipe mirf b.e fihe.d -·ftT a butle'1ty v alve. caqio"!ihlp:. ConuJ-lidats a carnrrcn ~aabe Ro SOLAS
s-21e: T e..n:irld cf watertight cla1 ti' es: Title dianged, hatc:tl!:$ to.'* t!5te1l, ind I.Ide= imeU c:rgo>hips Testlng 0 f hatches. omi tle-:i P'Wi ous. ly
B·21C-1.2 V ena through the oulithe:31d deoi Ad Oi:-d "\ntenignt are.a":t., r smctreo allcYt'Shlf:. and ..IS Tof'urrnonise v.·iih !te r s t d pr.atabiU~tic SOLAS
8-217.? Ar oi;)es within a supers,trudure N.tw consiclered lr!f7cleaea if oowaleright me.3ftS d dc&u--e frrteo C4lns!5't1Uei1i911.1po.n switdi toprot:ebilisucSOLAS
s~

s...:20
6~201
1s2
a...: 1.s. i
Doe n • a tertig ht OQCl<n.
Carne~ oonttol c .rilti
in

Loe.oil"gjl a·f C:lli>el"Q9' snipi

Lce.dirg cfs.hiat
c.ss >ergs s h ~ O:Eleted

..,.,..
Wn tlfe rw:w a.ovis.ions. fa JH&>e.Jl;!!rthii::.s.
"Fe:s.ergs
Ship to be Lll7ight, IO->icfirq in !!o::o'CS!l06!'4'it1 SIS p.ermit!ld
Pttvls.ion n-"DW oond1uoral. >es f\.ts-C. 1.Circ. 1280
~rs e"QUenos
Now Steplj&J
d CoMii Conc:olda Les•
1> 911 inipi
lnitl311Gt net parrritte:l, c:onfrnution af anent p-a.:tlce
~2 11and ~ OoetatJon of •--.e.n.V1tdoo.n drill> u: placed ,..kh qpe-atlc:nal t:s& h.1o1-e 9ppro17iai~'k0fding
SA 211 Lea'ling con "le::wirg PO't" r epb'Ced with ...,oysge a:?ITWTV:!'ICEoS. Pon"& too>~ec:ific r eplaoe-o in »e~ gal Re.gulaticrs
e-~ '2.3 O penwate!'!ign.tdoof'S foo1ncn~ ~deo refEt'f irlJ 10 ne-w guiii:!nae., r..tsc tCirc Seo: 8 -4 192. Up;tl3t.ed Ci'c 12Stl no-t;'et aiv-ailable
C3e-114 Bilge Dl1lV• AD floodirg mnditioris. no;o. llmite'O oy r e-fere.ro: b ~ea- 8--1 S Cl3tiflc.atJon

SOLAS 2020 Damage Stability Regulations. Hutchinson and Scott

The Western Europe Section of the Sooety of Naval Architects and Marne Engineers 61
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
- _- - -_-- --

Main amendments to the Explanatory Notes (EN)

(SDC4/16 Annex 1, MSC.429(98))
Expll.n.atary Unr; SIE]Kt Am•ndmtH1l Reason
... ' l Ar~ICilO/) 1' --:i'~-:.;;!-1 £XP-'llL'lS Wl".3tli1i:*-"-l ,..,.~ i..:&l'atc" C»t.t:iE so~ exll'd M t..Jt R!'g A 1 3• '.!-O.e.!:eo
~A.Z 11 -~ "J %1V'te ,jr!J4" t R:b- 0!' tlll ~C.''Jti-."';"t '·r- J.-::a'.!-5" ex:J-...de-: ~J'i"CVa.~W a&r t ;i ~;i it U:o~J- D =.N:-t 5-t 4 ex:.<.. ru SV\IU 0').~~'1 !-C
S- 4 1 ~-.ara] ! t t ~li t =-"t'.': tat.-=a:l b:.:tr-t E C" C 2D W---Di<J-.::..! -l !-, J._ bct:.;;a: C:HI 11o..:;::i* - S:ll..~S~
~ ' !2
2-·!!•Vl:J!
Ir ti ::i s ti.biitly
Ara:~-:.:
" R!'i=.'"!'--::e t.c USCCit:...t 1!£.0f.::is ~!""'E m- ~-arr:o.e,j ,.~!'
Nn 9::i.:...C3-11--::e m~~
U~:.J~ ~ ... ~-2 G." w.hatmr..s~tlfE: J. ~ W't'
Da.r-4:.at~" GfY1f..e... r> v :$ ,..e a-: • t-i.el).r::a s1; 11~!'J.

3-l !-1 ~..P, 1r-~c.tf'l".ll)C.- t\.lt'l~ l,ti.il.:f Cc~.=t!- o/ 1= .l:~ l.:'l- ~l.:-:tC'J .... ,.,,..e\" ::tJ ;Jlll:l'.'S C..l! b:in- a a:i;~ w.l.l\ uw .. G1111d(G "'t~\>e-s er:
0-\f. Z:4 De-;~ en--t.:a ~ 0.:..> R ra d!-GE:r4et:)...Jy U1lC" N . ~JSOr..$0t' OU:~
2-t i' _.G...tt,;i 1'-Co:>- ll'~!-XA. N:>A' 2"N".s i' D t·lal<:!?d Catibatc-r ""' •~r~A ICIL::l.;J:tro:r: ~~ -~e-:i
:-r 1 z =~-uoftr.itn~"'"' lj,,~ : N S - 1 :-. 4-1! liJ.n"~i::t tf'T'I'
ra-:?...:.:eo r~eC'.. t l'i't:O E J D Yo.•e 1;3rc-11 t3.r~--:!~ trc:_... -~ ES _ ...J-*1 ~ =-• !--1
S-f ' ! ~ -!a.CJ-3 W .a:t *M'S NN' ~ ,.. se~=- rt!' :.ii;rJrn. cm : N-3-~ 2:N-4 G.. 01- ::e or tsl:t!:g- t:€"-Et!itar .,.. u: C.lf'oE':! wl rter -- e;
;..11 e
3-t ] - •
P--r i iuTfiO?i
~'t1e:
Nu EN
~fe....J,Je; .J
o.::..;.~: Vf·l!'1 uo=--4":E-.re~t:i :.N ft!rR J.r.:t:r ReJ --z.z H ~ oa: iJ• t '*S ..'1°f3!£S - = SGJ:CS G'8ood - 2
'-'1 '•ti*Mc:',..,.U:JO~ '?:r!yaO? * -J tai a:S ?l::se--as i:"a~:::· C-=:...bt~r OC!:)-f'S " s wraJt3:ia
S-i - ; ~ l.l·l'IOr «O}""a:S , ;,'! D:i:J '"__: Cltl -~; J.;N Cl...=·=-C'~r.-1. M:.3 ,.e11 :c~14'Cill!t 1 ~~ rre-t-:i 0-~'"E:llt -:-r-~l""e'U r-*'=-$ :1 h-.. <Jq~f s!':~:
0-t 7.. • >.< T'"-a::: .e:::a : utdiv: M T M> ~• 9ls w-;1r":i;a;;rvnsa-Jd~ (4 t ar"•.21 C;a:ba.K:,. G~b trai-re -to'm"'i:-'l!x w.11:-i-O!S i 't
:.-t 7-2.?2 s~~-~- tll l b:.>J r-g :x&n:: -,.ey ~lf'~.e-1 " r r-E« s.. o-;.va;; 1-ot":S i:r J ~r pa C_; ...fc:tot'r- -n'-ror-p.'il$.-C:.,.t1ton ~ .. 1"fSf" 1*~~
E-f 7·2-Z.4 ~= =-~"l-'f1 Exler.s k.'yivc;:v:Ja-::t ;"'°~ ll'.te:!'l'_;t·-'E orotie-Jw-e: Na" ~SC...?J!·l{~Z- ~woes MSC.2•!. S-Sr c.ir·fuatu
: -r -.21 R)..=\.-0..s,.... t.f!"tll' ~ "'~ :XC.:u__; 1i:S:01 T ;:-:--., a.:d T.. -.
2- t i'·24 fuW1'1'1.it!'PtlD J.:f' DP-*•-' l't:v• t?i ~vs' SOLAS WIS
E-1 ... ·2-4. I IJ!-'-,- t.ofl c't::r:J:Jl!"t 0:-!.Sl!'.I Ni>• ·11 ~v__:e-3 SOlAS wa.
_;..• 7-2: Dn=-r~1 ... <:ad:i.,.·c_; C-e..::&.l ~. .. f"!,'V--*j SOLAS W'l. fnf R!'.4 ;..1 --;_;_
2- f ?.3:2~ =-as.:·..: Oil tci t,, ~JIEd R l'»fO~ff::t :C!lt.::.- ;y _~N . ~~err oa.a:c
S-2~ Oet.t·>!- OC.!tOO!i ~.__s k tyEq::ir.Ja ::t'ar.a r-oa_,_~ • ~ .CD'13';t3!'f!e !-.tc li.afec"Ct: v..Jr;e - tr.* E;Ot.AS &xt u:i o;.u:: l
3--2 . 2 b • OOcrs i:t.U'l NN ·-a:~tn t:lll-SOtAS ~b l:f.e.? _:..;;-·~ - a;_,, T ~~
;.:_ ' 2.0 1 :!'l.tt-!'~\:.JJ -.E:S N~ i:N ~.Si!'ft.:-.J W1t' d.J,d"i.m5 '-if=~~ ~ :.r4e.s: -in- t"I-.: SOLAS l!X1
-;..2 12 •: 1· At =u, c.dra:-= Nu £N5: " S-a!~J W1tll a "1_;-t;;.rc:; A~~-,, e>.. idt$a J.ar~ st:',. w.~~-T'Tl - jai""""...a-e:
2-2 · 323 O: s~ ~,. P 'c::t N'-£4' :N ·- ~ C.ar t:.awr c 1 1w!t.:toc"--=tn:le-Jct£ CCSiO ·syste:n"!
3-2'?"'.'.e IEC s ti";ja"iJ OE~l!-J- =-.:ic r-t:
rore-d.g.1 --. SO\.AS tui
3-2 f!-t Qi:g-,, ¥~ ~:: r~ ; Nff~ · =n!--l Cu.-les•pe:.iieab· ry0-'R~s. 3-I •!... t - o ·!-t l
S-Z 17 T s.c_-.; ·1, _; igt'tl°l"l.:fOG.'-5. :.n:r.: ~~'y ~11'1::te:ja_r4 M>:fit~ ..- ,.. 11°9' ib:;nt"".!. :tc. -"o.s-JUSJrydarrkalll!!.. -ti:f ~ ~;y
3-Z 1- 3 A-;r c.-c'-i eCE- a-:i-: R-!'•.t::.!'d IU!e-~;: d::.r;:s or SOtAS ext
2-: 17-1 RO-RO =.~ !: NS' :.'t-~!-l C-=:;ie: a.l't" .c.ac 1yotR~:. 9-Z ~-.~ t 1 tr. 1--l >2
S-422 US.-!' 0 1 -=-o:.n N:!!W E:N -.Ji!.~~ :xriU3 Wd!Y :."!E Ot fl.:. eil""i. Woe'f

SC L S 2020 Damage Stability Regu!aunns Hut ·h1rsun dl"d Sc1..m

The western Europe Section of the Society of Naval Architects and Marne Engineers 62
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
Changes to R index and Stinat,i factor as
they effect RO-PAXs

S LAS 2020 Damage Stab I ty R u t ons Hutchinson and Scott,

The West n E ope Section of the Soo ty ofN va Arch1t d5 and Mann fng ne s 63
l...!oyd s Reg ster of Sh pp ng Lo:ldon U t d K;igdom February 2018
Comparison 0.95

of R index
for 0.90

passenger ll:
xw
ships: 0
~
0.85

z
0
(ij
>
i5 0.80
SOLAS 2009 ID
::::>
"'0w
er
5
@ 0.75
er
proposed
. •
rev1s1ons 0.70

0.65 ' - - - - ' - - - - ' - - - - ' - - - - - ' - - - - ' - - - - ' - - - - ' - - - - - '
SOLAS 2020 0 1 000 2,000 3.000 4,000 5,000
TOTAL i'U.1 BER OF PERSONS ONBOARD. N
6,000 7,000 8,000

Sul.AS 2020 Damage Stai 1ht Regu!at1nns Hui ·h1rsun and Seen
The Western Europe Section cf the Society of Naval Architects and Marne Engineers 54
Lloyds Register of Sh1pp1ng London United Kirgdorr 6 February 2018
 ;«e
"''i "'"'tj
1',
~,

=·=•~·-•.;;.i: ~a~ -;;:'i! ..

~J j;~~~,-;~~,
t I
l I
C. H

Comparison
of R index
for
passenger
and
cargo ships: 8
!
! (:~ 5 +---+--1--,,<--+----l--+---+--+----l
a ~ ~~-(l; : : ! ':.\n~ ::i.:
e: ~~.:;~-.~-.--:::\:.;;·c ·:.-- y,.~
~ o·--+- __,f- ~f-1--f- -fi1.!,..<0.1f ..,i -:v :n
SOLAS 2009 - -- -,,

_,
- 'f~ i&-1\-0 =s<.i:O.'\i
'jl;:
~~ - ~-cc....-~'f'I'
·~·-.t."l ::;· E O-ft : ,-.;.•.,3C=
•'C--::JJV
r-:S=,.: 1';::-::;.: ,,ll - -E,_:: •.op;
·• ~= !c:.=j..
£.!:". -t i: Vi.£
;-_::-.-;v<"...:... •.,...~ ..c a ·~· r -- i
"""-"-"'t.-"'-V' - :.~.:U: "f • Jit . ·, .,_,.
.: H +---l'--1---h =-r..:.-~ . !. ~ f4. v ! - · :! -L

~~=,,
~E~~~!~::!~::~
__. .,_ . ...... .....
- ~=-=

SOLAS 2020

Sul.MS 2r 2') Damage Sl: "11ht Regu!at1ons. Hutchinson and Scott

The Western Europe Section or the Sooety of Naval Architects and Marne Engineers 65
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
__ - - -_- --

Discussion of R index proposed by SOC 3 to MSC 96

• Substantial increase in the R index in comparison with SOLAS 2009

particularly in typical RO-PAX N 500 to 2,500 persons
• Agreed curve discontinuity N 1,000 as it was felt that to use the smooth
line proposed by EMSA3.2 would be too onerous for smaller ships
• Increase in R index could be achieved in a cost effective way, according
to the EMSA3 and confirmed by a Formal Safety Assessment exercise,
which was approved at IMO in November 2015
• Adoption of proposed revised R i ndex formulation would represent a
significant improvement in safety level for all passenger I RO-PAX ships
• MSC 3 proposal discussed at MSC 96 in May 2016 but not approved
• New submissions invited to MSC 97 in November 2016 - approved
SOLAS 2020 Damage Stability Regu!a!1ons. Hutchinson and Scan
The Western Europe Section of the Society of Naval Architects and Marne Engineers
66
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
Changes to S tinal; factor
I
agreed at SLF 55 to MSC 96
1

s tinaf.i =
K [ GZmax Range
. TGZmax . TRange
]4 K = Bma"< - Be
emax - Bruin

• K unaltered , Bm;n 7° for passenger ships, 25° for cargo ships

Bmax 15° for passenger ships, 30° for cargo ships
• GZmax is not to be taken as more than TGZmax
TGZmax = 0.200 metres for RO-RO passenger ships each damage
case that involves a RO-RO space
= 0 .120 metres otherwise
• Range is not to be taken as more than TRange
TRange = 20° for RO-RO passenger ships
each damage case that involves a RO-RO space
- 16° otherwise
- ~ ge Stab11ty Regulauons Hutch nson and S ott,
West n Europe Secllon of the Soaety of Naval Arch1t ct and ti. ne Eng neers 67
L vd s Reg ster of Shipping London United Kingdom February 2018
Likely effect of potential regulation
changes on RO-PAX design

S LAS 2020 Damage Stab I ty R u t ons Hutchinson and Scott,

The West n E ope Section of the Soo ty ofN va Arch1t d5 and Mann fng ne s 68
L yd s Reg ster of Sh pp ng Lo:ldon U t d K;igdom February 2018
--- - --

Likely effect of potential regulation changes on

RO-PAX Design (1)
• Presenters have no knowledge of designs optimized to SOLAS 2020
R index and s finat i factor hence comments only be speculative in nature
'
• tend to reduce A index as less damage cases +ve s finat,i
s finat,i factor wi 11
- some studies, A index reduced 1% to 3°/o for medium/large RO-PAXs
- recent Danish study, A index reduced10% to 12% for small RO-PAXs
• N 2,000 RO-PAX, R index increase significantly 0.78 to 0.86, according
to EMSA3 achievable whilst RO-PAX still operating economically
• New generation RO-PAXs, constructed to SOLAS amendments, will be
much safer but less economically competitive than current designs
• Potential risk of life-span of existing tonnage extended as far as
possible into the future and new-building orders delayed

SOLAS 2020 Damage Stability Regulations. Hutchinson and Scott

The Western Europe Section of the Society of Naval Architects and Marne Engineers 69
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
Likely effect of potential regulation changes on
RO-PAX Design (2)
• Designers will have to minimise economic dis-benefits of increased
safety level by careful selection of RCOs - increased freeboard I beam
• Economic climate important- if rates are good and prospects promising
then new ships will inevitably be ordered anyway
• 'level playing field ' for RO-PAX constructed after 2020 but, as
previously, no 'Grandfather Clause' so there will always be an
imbalance between the pre and post 2020 RO-PAX until older RO-PAX
are slowly withdrawn from service
• Anathema to RO-RO owners and operators is the RCO of utilising
watertight doors on the main car deck to achieve a high A index -
believed not now necessary, whereas much higher R index levels
proposed in earlier research work may have left this as only option
SOLAS 2020 Damage Stability Regulations. Hutchinson dnd Scu1l
The Western Europe Section of the Society of Naval Architects and Marne Engineers 70
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
Stockholm Agreement

S LAS 2020 Damage Stab I ty R u I Hutchinson and Scott,

The West E ope Section of the Soo ofN va Arch1t d5 and Ma n fng"""""" 11
yd Reg er of Sh pp ng Lo:ldon U t d K; dom February 2018
__ - - -_- --

Stockholm Agreement (1)

• Stockholm Agreement was concluded in Stockholm on 27th to 28th

February 1996 - original signatorries Denmark, Finland , Germany,
Ireland, Norway, Sweden and United Kingdom, into force 1st Apri I 1997
• Expanded to cover the Mediterranean countries
• Now applies to all EU countries and to domestic RO-PAX routes
• Few, if any, countries outside of the EU have adopted it, presumably
because they feel that the WOO effect was predominantly a
phenomenon peculiar to the sea conditions and high traffic densities
encountered in EU
• IMO, which usually prefers safety legislation of this type to be applied
worldwide, accommodated this difficulty by adopting a so-called
Regional Agreement based on local conditions

• ~
Mantini&

~
S< L S 20~0 Damage Stab1ht Regu!at1nns Hutchinson and Scott
The Western Europe Section of the Society of Naval Architects and Marne Engineers
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
72
__ - - -_- --

Stockholm Agreement (2)

• Recognises the inherent danger of rapid capsize following ingress of

water onto large open deck areas, as of those found on RO-PAXs, due
to the so-called WOO effect. Accounts for WOO in two ways:
1. SOLAS90 deterministic statical damage stability calculations
assuming 1/2 compartments breached below the vehicle deck (inc .)
but with the addition of a layer of flood water on the vehicle deck of
depth , when ship EK Upright, as a function of pre-damage residual
freeboard - depth 0.50 m if <= 0.30 m, depth 0.00 m if >=2.00 m
2. model tests worst damage cases in a Hs for operational area
• When probabilistic SOLAS 2009 replaced the SOLAS 90 hope that it
could be revoked , however extensive research clearly showed that
WOO was not accounted for in formulations COSS meeting in Brussels
late in 2008 I early 2009 EC decided that it was not safe to revoke it
SOLAS 2020 Damage Stability Regu!a!1ons. Hutchinson and Scott
The Western Europe Section of the Society of Naval Architects and Marne Engineers 73
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
__ - - -_- --

Stockholm Agreement (3)

• Designers cannot fully utilise the possibilities provided by SOLAS 2009

for optimising alternative subdivision arrangements (particularly those
involving combined transverse aind longitudinal bulkheads, such as
long lower holds) because of the continuing assumption of the
deterministic 8/ 5 damage penetration depth. Further complications
have emerged with the forthcoming new 2020 R index and RO-PAX
specific s finat,i factor
• Given these uncertainties over the possible combined effects of the
latest amendments on future RO-PAX designs, a major concern is
whether or not the regulatory authorities, in this case the EC , will
decide to revoke the Stockholm Agreement for new ships built to
comply with the possible 2020 SOLAS amendments

SOLAS 2020 Damage Stability Regulations. Hutchinson and Scott

The Western Europe Section of the Society of Naval Architects and Marne Engineers 74
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
__ - - -_- --

Stockholm Agreement (4)

• New research studies are under way in the EU to determine whether or

not the increased safety levels as represented by the new 2020 R
index and S rinat,i factor will adequately account for the WOO
• Experience gained thus far in optimising RO-PAX designs only to the
SOLAS 2009 amendments, without the support of the Stockholm
Agreement, led to some alarming designs, which , though compliant
with SOLAS 2009 had little or no survivability following specific damage
scenarios even in calm conditions - research is therefore essential , in
our view, before the 2020 R index and S rinat,i factor can be relied upon

SOLAS 2020 Damage Stability Regulations. Hutchinson and Scott

The Western Europe Section of the Society of Naval Architects and Marne Engineers 75
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
__ - - -_- --

Stockholm Agreement (5)

• Some advocate retaining Stockholm Agreement and requiring each

new RO-PAX be model tested and results analysed to assess how well
new s finat i factor predicts survivability of the damaged ship in a seaway
'
• We remain sceptical that a simplle s finat i formula has predictive capacity
I

and favour a formulation giving preference to increasing freeboard as

the main RCO as opposed to the proposed '20/20' formulation which
accepts WOO will occur but the RO-PAX has sufficient residual stability
• Given difficulty to predict WOO effect in a seaway, without advanced
numerical techniques, it is sensible to minimise water ingress by
increasing residual freeboard (say 2 metres+ for damages involving the
garage space) rather than basic damage stability characteristics
• UK will probably support retention until sufficient reliable evidence
SOLAS 2020 Damage Stability Regulations. Hutchinson and Scott
The Western Europe Section of the Society of Naval Architects and Marne Engineers 76
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
 Conclusions

S LAS 2020 Damage Stab I ty R u I Hutchinson and Scott,

The West E ope Section of the Soo ofN va Arch1t d5 and Ma n fng"""""" 17
yd Reg er of Sh pp ng Lo:ldon U t d K; dom February 2018
- -- - ---

Unresolved I Controversial Issues

• Revised s- should be based on residual freeboard not GZ

• Loss of 'margin line' concept
• Does SOLAS 2009 concentrate on loss by capsize omitting by sinking?
• Bulkhead or freeboard deck is no longer required to be weathertight
• Do air pipes with closing devices constitute down-flooding points?
• Why no adequate definition of s or R - both vital concepts. 2020 R is no
longer based on LSA provision. R covers overall ship survival only, not
individual survivors after an incident
• IMO still introduces amendments without being tested on new designs
• Many ask why new cruise ships are so ugly- like floating flat blocks.
Reduced transverse bulkhead weight allows more decks 'up top' (see
next slide)
SOLAS 2020 Damage Stability Regulations. Hutchinson and Scott
The Western Europe Section of the Society of Naval Architects and Marne Engineers 78
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
- - -- - - - - -

The Future and •• •

.........................•. .......

... the Past

- - ·..·~
... RMS Queen Elizabeth 2

- ·-··••-· "'I
;...z:,; .,~
....
.,...,
entered service in 1969

Hutchinson and Scott,

19
dom February 2018
Conclusions

• Update of regulatory situation at IMO and COSS re damage stability

• Agreed significant changes to R index and Srinat i factor
•
• Difficult to predict whether safety improvements will be sufficient to
allow the EU regulatory authorities to revoke the Stockholm Agreement
• Clear need for more research into whether the increases in R index
and s finat i factor will allow the Stockholm Agreement to be revoked and
•
hence total reliance placed on the increased safety level provided by
the latest proposed and future amendments to SOLAS
• Possible update to RINA at Damaged Ship IV in mid 2018 or The Naval
Architect magazine and other forums

SOLAS 2020 Damage Stability Regulations. Hutchinson and Scott

The Western Europe Section of the Society of Naval Architects and Marne Engineers 80
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
References: The Naval Architect, Journal, upcoming?

1. HUTCHINSON, K.W. . and SCOTT, A L. 'Damage stability: upcoming SOLAS 2020

regulations', The Naval Architect, The Royal Institution of Naval Architects, London, United
Kingdom, September 2017. ISSN: 0306-0209
2. HUTCHINSON, K.W. 'Damage Stability Update', The Naval Architect, The Royal Institution of
Naval Architects, London, United Kingdom, May 2017. ISSN: 0306-0209
3. WOODWARD, M.D. and HUTCHINSON, K.W. 'Uncertainty in ship stability', The Naval
Architect, The Royal Institution of Naval Architects, London, United Kingdom, September 2016.
ISSN: 0306-0209
4. WOODWARD, M.D., van RIJSBERGEN, M HUTCHINSON, K.W. . and SCOTT, A.L.
'Uncertainty Analysis Procedure for the Inclining Experiment', Ocean Engineering Elsevier,
Amsterdam, The Netherlands, Volume 114, 1st March 2016, ISSN: 0029-8018. (with.) - Open
Access at: http://www.sciencedirect.com/science/article/pii/S0029801816000287
TBC HUTCHINSON, K. W and SCOTT, A.L. 'SOLAS 2020 Passenger and Cargo Ship Damage
Stability Regulations", International Conference on The Damaged Ship IV, The Royal Institution
of Naval Architects, London, United Kingdom, 16th and 17th May 2018.

SOLAS 2020 Damage Stability Regulations. Hutchinson and Scott

The Western Europe Section of the Society of Naval Architects and Marne Engineers 81
Lloyds Register of Sh1pp1ng London United KirgdorT' 6 February 2018
References: RINA International Conferences

1. HUTCHINSON, K.W. and SCOTT, A.L. 'Passenger RO-RO Ferry Damage Stability: Status and
Development of International Regulations', International Conference on the Design and
Construction of Ferries and RO-Pax Vessels, The Royal Institution of Naval Architects, London,
United Kingdom, 251h and 26th May 2016. ISBN: 978-1-909024-54-0
2. HUTCHINSON, K.W. and SCOTT, A.L. 'Current and Possible Future Intact and Damage
Stability Regulations and Verification Tools for Passenger Ships', International Conference on
The Damaged Ship Ill, The Royal Institution of Naval Architects, London, United Kingdom, 25th
and 261h March 2015. ISBN: 978-1-909024-38-0
3. HUTCHINSON, K.W. and SCOTT, A.L. 'The Development and Impact of Current and Possible
Future National and International Damage Stability Regulations', International Conference on
The Damaged Ship II, The Royal Institution of Naval Architects, London, United Kingdom, 301h
and 31th January 2013. ISBN: 978-1-909024-11-3
4. HUTCHINSON, K.W., SCOTT, A.L., WRIGHT, P.N.H., WOODWARD, M.D., and DOWNES, J.
'Consideration of Damage to Ships from Conceptual Design to Operation: The Implications of
Recent and Potential Future Regulations regarding Application, Impact and Education',
International Conference on The Damaged Ship, The Royal Institution of Naval Architects,
London, United Kingdom, 26th and 27th January 2011 . ISBN: 978-1-905040-79-7
~ S< L S 20~0 Damage Stab1ht Regu!at1nns Hui ·h1rsun dl"d Sc
~ The Western Europe Section of the Society of Naval Architects and Marne Engineers 82
P(llll:cy Lloyds Register of Sh1pp1ng London United Kirgdorr 6 February 2018
- - -- - ~ - --

That's

all

folks!

S LAS 2020 Damage Stab I ty R u I Hutchinson and Scott,

The West E ope Section of the Soo ofN va Arch1t d5 and Ma n fng"""""" 8:3
yd Reg er of Sh pp ng Lo:ldon U t d K; dom February 2018
consider such uncertainty regarding KG1 This variation in G may
be substantially more influential than, for example, any possible
precision improvement gained from reforming the underlying
assumption in wall-sided theory regarding the stationarity of
the transverse metacentre (Mr) as traditionally applied in the
derivation of the as-inclined GMr-

Summary
What can uncertainty analysis actually do to aid our engineering
comprehension and hence our effectiveness as ship designers /
builders and practicing naval architects? Well, the important
thing to appreciate is that a wide uncertainty band does
not 'necessarily' mean a poor measurement; it is all about
information.
We might for example arrive at some 'estimated' value,
calculated as a function of multiple parameters that we have
measured directly. Taking into account the precision associated
with each of the values we measure, and accepting how these
uncertainties propagate as we calculate the value of interest, we
establish the related uncertainty.
If, however, instead of calculating the uncertainty, we choose
to repeat the measurement a great many times, we might, if we
are fortunate, get sufficiently close to the same value each time.
We could then take the uncertainty to be simply the standard
deviation of the mean value. We would therefore have a smaller
uncertainty and consequently more confidence in our result, but
with the associated additional cost in both time and resources,
and subsequently money.

Conclusion
So what's the bottom line? Uncertainty analysis can assist us, as
naval architects, to effectively decide how to best utilise our time
and resources in the most efficient manner in order to improve
confidence in the parameters that are of most significance and
importance to us, our clients and on the projects which we are
undertaking. The parameters that warrant the most attention
are those that present the greatest risk; that is risk to persons,
property and/or the environment.
Ship stability, as with other critical parameters relating
to ship operations, is of paramount importance, and should
therefore be dealt with accordingly with an appropriate
uncertainty analysis. NA

Disclaimer:
The views expressed in this article are those of the authors and
not necessarily represent those of the organisations with which
they are affiliated and the professional institutions of which they
are members.

References:
I.WOODWARD, M.D., van RIJSBERGEN, M., HUT CHINSON,
K.W. and SCOTT, A.L. 'Uncertainty Analysis Procedure for the
Inclining Experiment; Ocean Engineering, Elsevier, Amsterdam,
The Netherlands, Volume 114, I March 2016, Pages 79 to 86.
ISSN: 0029-8018 http://www.sciencedirect.com/science/article/
pii/S0029801816000287

The Naval Architect September 2016 39

 Ocean Engineering 114 (2016) 79–86

Contents lists available at ScienceDirect

Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng

Uncertainty analysis procedure for the ship inclining experiment

Michael D. Woodward a,n, Martijn van Rijsbergen b, Keith W. Hutchinson c, Andrew Scott d
a
Australian Maritime College, University of Tasmania, Launceston, Australia
b
MARIN, Wageningen, The Netherlands
c
Babcock International Group Centre for Advanced Industry, Newcastle-upon-Tyne, UK
d
Maritime & Coastguard Agency, Newcastle-upon-Tyne, UK

art ic l e i nf o a b s t r a c t

Article history: The inclining experiment is typically performed for all new-build ships and after any major refit. The
Received 31 July 2015 purpose of the inclining experiment is to establish the vertical distance of the centre-of-mass of the ship
Accepted 16 January 2016 above its keel in the lightship condition. This value is then taken as the point of reference when loading
the ship, for establishing the ‘in-service’ stability, throughout the life of the ship. Experimental uncer-
Keywords: tainty analysis is commonly utilised in hydrodynamic testing to establish the uncertainty in a result as a
Inclining experiment function of the input variables. This can in turn be utilised to establish an interval about the result that
Uncertainty analysis may be expected to encompass a large fraction of the distribution of values that could reasonably be
Ship stability attributed to the measurement. This paper provides a methodology for calculating a confidence interval
KG
for the location of the centre-of-mass of a ship from an inclining experiment; and ultimately, in any load
GM
condition.
The uncertainty compared to an assumed metacentric height of 0.15 m is provided for four classes of
ship: buoy tender 0.15 70.15 m ( 7100%); super yacht 0.1507 0.033 m (7 22.0%); supply ship
0.150 70.047 m ( 7 31.3%), container ship 0.150 70.029 m (7 19.3%), ropax 0.150 70.077 m ( 7100%).
& 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).

1. Aims and objectives This value is then taken as the point of reference when loading the
ship, for establishing the ‘in-service’ KG, throughout the life of the
The aim is to establish procedures for identifying the experi- ship. An accurate estimate of the limiting KG is absolutely neces-
mental uncertainty in the estimate of the centre-of-mass height sary for the safe operation of the ship, so as to ensure adequate
above the keel (referred to as KG) by method of an inclining stability. Clearly, this is dependent on an accurate estimate of the
experiment (IE). lightship KG obtained from the IE.
The first objective is to give procedures for performing a pre- While typically all attempts are made to conduct the IE in a
test analysis that can be employed to identify the best course of manner that minimises the introduction of error, many potential
action for reducing the experimental uncertainty. The second sources of error exist. For example, all attempts are made to
objective is to give procedures for performing a post-test analysis remove the influence of fluid free-surface effects, by emptying or
that can be employed to identify a confidence interval for the pressing-full all tanks. Any suspended loads are secured or
resulting estimate of KG.
removed and anything that may move is removed or made secure.
Similarly, all attempts are made to conduct the IE in calm condi-
tions, when the effect of wind, waves, current and the wash from
2. Background
passing ships is minimised.
Notwithstanding all attempts to minimise errors, sources of
The IE is a required procedure [unless exceptions apply; see
uncertainty will always be present – uncertainty being different
IMO, 2008] for all new-build ships and after any major refit.
from error. Due to the stochastic nature of the world, all input
The purpose of the IE is to establish KG, in the lightship condition.
variable measurements are only known with limited accuracy. The
uncertainty in the results (in this case the estimate of KG) is
n
Corresponding author. Tel.: þ 44 191 222 6750; fax: þ 44 191 222 5491. dependent on the magnitude of the uncertainties of each input
E-mail addresses: [email protected] (M.D. Woodward),
variable and on the particular sensitivity of the results to each
[email protected] (M.v. Rijsbergen),
[email protected] (K.W. Hutchinson), input, which is dependent on the form of the data reduction
[email protected] (A. Scott). equations.

http://dx.doi.org/10.1016/j.oceaneng.2016.01.017
0029-8018/& 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
80 M.D. Woodward et al. / Ocean Engineering 114 (2016) 79–86

2.1. Overview of the inclining experiment coefficient ci is the partial derivative of the results with respect to
any given input variable xi ; given by Eq. (5).
Explanations of the procedure for an IE exist in many texts, ∂y
with the fundamental description given by (IMO, 2008). In brief, ci ¼ ð5Þ
∂xi
an IE is conducted by forcibly inclining the ship by moving a
known weight a known transverse distance across the ship. The The standard uncertainty of any given variable is relatively easy
inclination is measured from the movement of a plumb-line to obtain. If a sufficiently large number of samples of measurement
relative to a mark-board, that is horizontal when the ship is data are available, the Type-A standard uncertainty for a single
upright. Typically, two or three plumb-lines are employed (for- sample is equal to the sample standard deviation. If there is no
ward-amidships-aft) to account for any torsional deformation of recent measurement data available, the limits of the uncertainty
the ship. Then, the metacentric height GM is obtained according need to be estimated or e.g. taken from a specification of a mea-
to, surement device. With these limits and an assumed probability
distribution, the Type-B standard uncertainty can be derived (for
wd
GM ¼ ð1Þ application guidance see (ISO/IEC, 1995) Section 4.3).
ρ∇ tan θ
where w is the mass of the weight moved, d is the distance the
weight is moved, ρ is the water density, ∇ the displaced volume of 3. Derivation of sensitivity coefficients
the ship and θ is the induced heel-angle. Eq. (2) calculates the
height of the metacentre above the centre-of-buoyancy as a By assuming linearity, for small changes in draught T, for the
function-of-form for the given draught. variables KB, I and ∇, the sensitivity coefficients can be obtained
I directly. Going to the hydrostatic tables for the ship, the tangent to
BM ¼ ð2Þ the curves at the lightship ‘as inclined’ draught are utilised to
∇
obtain the coefficient αn and constant terms βn shown in Eq. (6).
In Eq. (2), I is the transverse second moment of area of the
water-plane at that draught. The height of the centre-of-buoyancy KB ¼ α1 T þ β 1
above the keel KB, (the centroid of volume at that draught) being a I ¼ α2 T þ β 2
geometric property, is readily calculated from the hydrostatic ∇ ¼ α3 T þ β 3 ð6Þ
particulars. The height of the mass-centroid (centre of gravity)
above the keel KG, is then given by Eq. (3). Eq. (7) is obtained by substituting Eqs. (1), (2) and (6) back into
Eq. (3).
KG ¼ KB þ BM  GM ð3Þ
  " #
α2 T þ β 2 
wd

ðα 1 T þ β 1 Þ þ 
α3 T þ β 3 ρ α3 T þ β3 tan θ
2.2. Overview of experimental uncertainty analysis KG ð7Þ
Simplifying as much as possible, the relevant sensitivity coef-
The expression of experimental uncertainty is generally dealt
ficients are then given by Eqs. (8)–(12), for the ith heel-angle
with by National Metrology Institutions. However, for the appli-
measurement induced by weight shift. In Eq. (12) the gradient
cation of specific procedures, scientific committees or societies
terms αn are replaced with the specific differential terms, as they
more often take responsibility. Considering hydrodynamic testing,
are perhaps more meaningful.
the International Towing Tank Conference (ITTC) provides Proce-
dures and Guidelines for many aspects of ship related testing. ∂KG wd
c1i ¼ ¼ ð8Þ
Though the IE is not within its scope; one procedure (ITTC, 2008) ∂θ i ρ∇ sin 2 θi
does have relevant information, as it describes the application of
uncertainty to hydrodynamic testing. Also, the development of all ∂KG wd
new procedures and guidelines should be expressed in line with c2i ¼ ¼ 2 ð9Þ
∂ρ ρ ∇ tan θi
the International Organisation for Standards (ISO), Guide to the
Expression of Uncertainty in Measurement (ISO/IEC, 1995). ∂KG d
In accordance with ISO uncertainties can be categorised into c3i ¼ ¼ ð10Þ
∂w ρ∇ tan θi
Type-A and Type-B. Type-A uncertainties are components
obtained utilising a method based on statistical analysis of a series ∂KG w
of observations. Type-B uncertainties are components obtained by c4i ¼ ¼ ð11Þ
∂d ρ∇ tan θi
means other than repeated observations. For the IE most mea-
surements are Type-B; or at least must be treated as such due to  
∂KG ∂KB 1 ∂I ∂∇ ∂∇ wd
the nature of the measurement methods applied. In many respects c5i ¼ ¼ þ  BM þ ð12Þ
∂T ∂T ∇ ∂T ∂T ∂T ρ∇ tan θi
however, the distinction is arbitrary as, for onward calculations,
Type-A and Type-B uncertainties are treated in the same way. In The uncertainty in the ship geometry is an important con-
its most simple form, the combined uncertainty in a result uc ðyÞ is sideration in comparison to the drawings. This takes into account
the root-sum-square of the standard uncertainties uðxi Þ for each the uncertainty in the position of the centre-of-buoyance and the
ith input variable multiplied by a corresponding sensitivity coef- metacentre, from which all other calculations are taken. Taking the
ficient ci for each variable, given by Eq. (4). partial derivatives of Eq. (3) (with Eqs. (1) and (2) substituted
accordingly) the sensitivity coefficients given by Eqs. (13)–(15) are
X
N
obtained.
u2c ðyÞ ¼ c2i u2 ðxi Þ ð4Þ
i¼1  
∂KG 1 wd
c6 ¼ ¼ 2 I ð13Þ
Of course, this is a somewhat simplified form, neglecting the ∂∇ ∇ ρ tan θi
possibility of correlation between various variables. Such correla-
tion will be dealt with later in the paper, but for the immediate ∂KG 1
c7 ¼ ¼ ð14Þ
discussion this simplified form is sufficient. The sensitivity ∂I ∇
 M.D. Woodward et al. / Ocean Engineering 114 (2016) 79–86 81

∂KG giving in Eq. (20), is obtained.

c8 ¼ ¼1 ð15Þ  max 
∂KB   s smin
u η ¼ pffiffiffi ð20Þ
2
If the induced heel-angle is given by Eq. (16), then the sensi-
4. Identification of the variable uncertainties tivity is the partial derivative of θ with respect to η, given by Eq.
(21).
With various types of calculation involved in an analysis, a ∂h ηi l
description of uncertainty in ‘levels’ is more practical. That is to tan  1 ¼ ð21Þ
∂η l η2 þl2
say, use the sensitivity coefficient and standard uncertainty at one
level to output the combined uncertainty. Then use this as the In a similar way, the sensitivity with respect to the plumb-line
input standard uncertainty at the next level. An example of such length is given by Eq. (22).
an approach is implemented within this methodology, utilising the ∂h ηi η
tan  1 ¼ ð22Þ
output combined uncertainty for the heel angle measurement as ∂l l η2 þ l2
input standard uncertainty for the next calculations. The next
section will look at the necessary variables and provides practical It is important to remember that although several plumb-line
methods for obtaining the required values. measurements are taken at various locations, these are not inde-
pendent measurements of the same thing. In actual fact, these are
  discrete measurements each contributing to a part of a data
4.1. Uncertainty in the heel-angle by plumb-line measurement, u θ reduction equation. In this case the data reduction equation is
rather simplistic, being simply the mean value for N plumb-lines.
Taking the length of the plumb-line to be l, and the horizontal From this, the sensitivity coefficient for each measurement can be
measured plumb-line displacement to be η, then the heel angle θ, shown to be equal to N1 . Bringing together Eqs. (20)–(22) into the
is given by Eq. (16). form given in Eq. (17), the uncertainty in the heel-angle induced
η by the ith moment (induced by weight shift) is obtained as given
θ ¼ tan  1 ð16Þ in Eq. (23). Here, the standard uncertainty of the jth plumb-line
l  
length u lj is the combination of two uncertainties. The first is the
The combined uncertainty for the measured heel angle is best measurement capability of the measuring equipment utilised
dependent both on the standard uncertainty in l and in η; as given to measure it, including components such as calibration uncer-
by Eq. (17). tainty and resolution. The second is the uncertainty in the mea-
 2  2 suring process with contributions such as alignment, repeatability.
  ∂θ ∂θ  
u2 θ ¼ u2 ðlÞ þ u2 η ð17Þ 82 32 2 32 2 32 9
∂l ∂η N  2 > >
  X 1 <4 lj smax
ji  smin
ji  ηji  =
u 2
θi ¼  5 4 pffiffiffi 5 þ 4 5 u2 lj
Typically, the plumb-line will be swinging back-and-forth in an >
N : η2 þ l 2 2 η2ji þ lj
2 >
;
j¼1 ji j
approximately sinusoidal oscillation. The value for η is typically
ð23Þ
obtained by trying to estimate the middle of the plumb-line swing.
Ideally the estimate of the uncertainty would be obtained as the  
sample standard deviation of the signal, over a sufficiently large 4.2. Uncertainties related to the water density, u ρ
number of cycles. In the case of the IE however, the time history of the
plumb-line displacement is typically not recorded. Taking the Typically, the water density around the ship will be sampled at
extremes of the swing would somewhat overestimate the uncertainty. several locations and at more than one depth. The average water
A reasonable estimate for uncertainty in the plumb-line displacement density is then taken as the basis for subsequent calculations.
measurement can nevertheless be obtained in terms of the approx- Utilising this method there are two main areas to be considered.
imate maximum and minimum observed values. The standard Firstly, there is uncertainty related to the best measurement cap-
deviation of a sinusoidal signal σ S , of amplitude ζ can be shown to be ability of the device employed to measure the water density.
as given in Eq. (18); with proof provided in Appendix A. Secondly, there is the uncertainty due to the measuring process.
If for example, the water density is determined by measuring
ζ the specific gravity, then the best measurement capability is the
σ S ¼ pffiffiffi ð18Þ
2 combined uncertainty of the calibration uncertainty as provided
by the calibration certificate
 and the resolution (smallest scale
Assuming that the swinging plumb-line motion is a pure
division on the gauge), u ρbmc . The second source of uncertainty
sinusoid, then the signal height is the maximum observed value to be considered is the uncertainty in the measuring process. The
minus the minimum observed value. The amplitude is by defini- main contribution to this uncertainty is the process of sampling.
tion half the signal height; given by Eq. (19), Since the samples can be assumed to be independent, the standard
 max  uncertainty of the mean value can be calculated by dividing the
s  smin
ζ¼ ð19Þ sample standard
2   deviation by the square root of the number of
samples, u ρσ .
where smax is the maximum observed swing of the plumb-line and The uncertainty for any necessary temperature correction
smin the minimum. Considering that the plumb-line will be oscil- associated with the hydrometer reading can also be taken into
lating about both the reference position and then later about the consideration by applying ITTC (2011). However, based on the
measurement position, the uncertainty related to both situations findings of the case studies (in Section 8), such finesse may be
needs to be taken into account. If the magnitude of the oscillations superfluous. The total uncertainty associated with the water
 
is not far different in either case, the uncertainties in the ampli- density u ρ , is then given by the root-sum-square of the com-
tudes are correlated. Then the standard uncertainty in η is equal to ponent uncertainties; given by Eq. (24).
2σ S . Substituting Eq. (19) back into Eq. (18), and multiplying by      
u2 ρ ¼ u2 ρbmc þ u2 ρσ ð24Þ
two, the uncertainty in the estimated plumb-line displacement, as
82 M.D. Woodward et al. / Ocean Engineering 114 (2016) 79–86

4.3. Uncertainty in the weight of objects moved, uðwi Þ 4.5. Uncertainties related to the draught marks, uðT Þ

In an ideal situation, a quayside crane will be employed to The estimate of the draught marks has two sources of uncer-
move the inclining weights. However, more typically, a forklift tainty. The uncertainty related to the position of the draught
truck will be employed to move the inclining weights and then marks and the uncertainty of the water-level with respect to those
return itself to a known position. Similarly, the staff involved in marks. For the first of these, the draught mark represents a dis-
conducting the IE must also return to known positions before the tance above the keel. The flat bottom of the ship however has itself
necessary measurement readings are made. The uncertainty rela- some variation. Realistically, adjudging the ‘flatness’ of the keel to
ted to items such as the forklift, the personnel and any other be, say plus-or-minus 10 mm, then the uncertainty of the draught
equipment are covered in Section 4.7. marks must be at least this. Depending on the construction
The uncertainty of the mass of each inclining weight is methods and the quality of build, the task is to make a realistic
assumed to be equal to the calibration uncertainty of the mea- judgment on the likely building tolerance; here represented by
suring device utilised to weigh it. If a given weight is made up of uðϵM Þ.
multiple smaller weights, each having been weighed separately on In addition to this, the effect of surface tension causes an
the same device, then their uncertainties in mass are correlated. uncertainty in the exact position of the water level due to the
 
This results in a simple addition of the individual uncertainties curved meniscus; here represented by u γ . The magnitude of this
instead of a root-sum-square calculation. Eq. (25) gives the depends on the roughness of the surface that the fluid is in contact
uncertainty for each ith inclining weight, where N is the number of with. A typical value would be in the order of 3 mm and should be
component weights making up each inclining weight. added (as a root-sum-square) to the other draught related sources
of uncertainty.
X
N  
uðwi Þ ¼ u wj ð25Þ As the water surface is invariable moving and, to some extent,
j¼1 the ships itself, then the measurement is problematic. This can be
improved upon by the use of a glass tube to damp out the wave
action; but some oscillation will always be present. For compar-
4.4. Uncertainty in the distance objects are moved, uðdÞ ison with the above, typical amplitudes could be in the order of
50 mm. For simplicity, a reasonable estimate of the uncertainty
When considering the placement of inclining weights, two may be obtained by multiplying the oscillation amplitude by the
sources of uncertainty must be taken into account. Specifically, the standard deviation of a sinusoidal signal; described in Section 4.1
uncertainty in the location of the marks made for positioning the and Appendix A. Letting the maximum local observed jth draught
weights and the uncertainty of the placement of the weights with mark be τmax j and the minimum be τmin j , then Eq. (27) gives the
respect to those marks. combined uncertainty for the draught measurement as,
If for example a measurement mark were made on a piece of 2 !2 3
white paper with a fine pencil and a steel rule calibrated in mil- X3
24
τmax –τmin  
þ u γ þu ðϵM Þ5
j j
2
uc ðT Þ ¼ c5j pffiffiffi 2 2
ð27Þ
limetres, then it would be fair to say that the uncertainty was plus- 2 2
j¼1
or-minus a millimetre. Conversely, just because a tape measure
calibrated in millimetres is utilised to mark the placement of the where j ¼ 1 corresponds to the forward draught measurement,
inclining weights, to assume such accuracy would be spurious. 2 the measurement amidships and 3 the aft measurement. Taking
Stretching a tape-measure across a, perhaps uneven, deck and into consideration the hog/sag correction and the layer correction,
marking with chalk or sticky-tape, or some such similar crude the draught at the longitudinal centre of flotation is given in
marking, could be more realistically considered as plus-or-minus a Eq. (28) (which is typically the reference point in tables describing
centimetre. Of course, a more sophisticated method might be the ship hydrostatic characteristics),
employed such as a laser measurement, to improve accuracy.
Notwithstanding, the task at hand is to make a realistic judgment 1 ðT 3 T 1 Þ
T LCF ¼ ðT 1 þ 4T 2 þ T 3 Þ þ LCF ð28Þ
of the accuracy that can be assumed with the tools utilised. When 6 Lbm
taking multiple measurements to calculate the total distance the
where LCF is the position of the longitudinal centre of flotation
total measurement uncertainty is taken as the root-sum-square of
with respect to amidships and Lbm is the length between draught
the contributing measurement uncertainties (or simply the sum if
marks. The corresponding sensitivity coefficients c5j are given by
the individual measurements are correlated e.g. taken with the
Eqs. (29)–(31).
same device). Then, the measurement of the mark dMi relating to
the ith inclining weight has an uncertainty uðdMi Þ. ∂T LCF 1 LCF
c51 ¼ ¼  ð29Þ
As with the above, when trying to line up an inclining weight ∂T 1 6 Lbm
(itself on a forklift truck pallet) with a mark made with sticky-tape,
then to assume millimetre accuracy would be spurious. As above, ∂T LCF 4
c52 ¼ ¼ ð30Þ
the task at hand is to make a realistic judgment of the accuracy ∂T 2 6
that can be assumed with the tools utilised. Then, alignment with
respect to the mark dAi for the ith inclining weight has an uncer- ∂T LCF 1 LCF
c53 ¼ ¼ þ ð31Þ
tainty uðdAi Þ. ∂T 3 6 Lbm
For each ith inclining weight moved, the total uncertainty is the
By taking an average from N draught measurements and
root-sum-square of the uncertainty related to the marks and the
assuming that their uncertainties are independent, the uncertainty
uncertainty related to the position with respect to the marks.
of the average draught is given by Eq. (32).
Then, Eq. (26) gives the uncertainty of the distance the ith
inclining weight is moved.   X N  2
1
u2 T ¼ u2 ðT i Þ ð32Þ
u2 ðdi Þ ¼ u2 ðdMi Þ þ u2 ðdAi Þ ð26Þ i¼1
N
 M.D. Woodward et al. / Ocean Engineering 114 (2016) 79–86 83

4.6.
 Uncertainties
 related to hull-form tolerances, uð∇Þ, uðI Þ and displacement uðΔÞ can be obtained from the density and volume
u KB uncertainties (given in Eqs. (24) and (35) respectively) by Eq. (42).
 

uð∇Þ u ρ
Taking the usual definition of volume to be ∇ ¼ LBTC B and uðΔÞ ¼ Δ þ ð42Þ
∇ ρ
taking logarithms, Eq. (33) is obtained.
log ∇ ¼ log L þ log B þ log T þ log C B ð33Þ
4.8. Uncertainties related to free-surface corrections uðFSC Þ
dy
Recognising that if y ¼ log x then dx ¼ 1 so dy ¼ dx
x, Eq. (34) is
x
obtained. After the IE is conducted a correction to KG may be required-if
there are any free-surfaces aboard the ship during the test.
∂∇ ∂L ∂B ∂T ∂C B
¼ þ þ þ ð34Þ Assuming tanks to be approximately rectangular, the free-surface
∇ L B T CB
correction is given by Eq. (43). In the equation ϱi is the density of
Considering the change in any given parameter to be the the fluid in the ith tank and ai and bi are the length and breadth of
manufacturing tolerance in that given dimension (denoted ϵ), then that tank respectively.
Eq. (34) can be rewritten. To assign a tolerance to the block coef-
ficient an assumption is made that any horizontal transverse ϱi ai b3i
FSC ¼ ð43Þ
measurement from the centre-line has the same tolerance as that ρ 12∇
of the breadth. This leads to a simplification (factor of 2 on breadth The sensitivity coefficients for the free-surface correction are
tolerance) where Eq. (35) gives the uncertainty in displaced given in Eqs. (44)–(48).
volume. 3
ϵ ∂FSC 1 ai bi
L ϵB ϵT  c12i ¼ ¼ ð44Þ
uð∇Þ ¼ ∇ þ2 þ ð35Þ ∂ϱi ρ 12∇
L B T
In a similar way, assuming that the water-plane area can be ∂FSC ϱ ab 3

approximated by a rectangle, the second moment of area is given c13i ¼ ¼  2i i i ð45Þ

3
∂ρ ρ 12∇
by I ¼ LB 12 . Again taking logarithms and with the same process as
above, Eq. (36) gives the uncertainty in the transverse second ∂FSC ϱi bi
3
c14i ¼ ¼ ð46Þ
moment of water-plane area. ∂a ρ 12∇
ϵ ϵB 
L
uðI Þ ¼ I þ3 ð36Þ
∂FSC ϱi ai bi
2
L B
c15i ¼ ¼ ð47Þ
From a similar analogy, Eq. (37) gives the uncertainty in the ∂b ρ 4∇
height of the centre of buoyancy.
  ϵ  ∂FSC ϱ ab 2
T c16i ¼ ¼  i i i2 ð48Þ
u KB ¼ KB
T
ð37Þ ∂∇ ρ 12∇
 standard uncertainty for the density of fluid in the ith tank
 The
4.7. u ϱi is obtained
 in a similar way as the uncertainty for the sea-
 Uncertainties
 related to the removal or addition of weights
water density u ρ ; see Section 4.2. The standard uncertainties for
u δG
the length ai and breadth bi of each tank are taken as the cali-
It is necessary to remove the inclining weights and other bration uncertainty of the device utilised to measure them, and
equipment from the ship after the IE is finished. The estimate of the uncertainty in ship displaced volume uð∇Þ as given in Eq. (34).
KG must then be amended accordingly. In addition, though not
ideal, the ship may well have weights on-board that will be 4.9. Other sources of uncertainty
removed or still to be added. Eq. (38) gives a change in the position
of the ships centre-of-gravity due to the addition on removal of an 4.9.1. Uncertainty of the position of the inclining weight centroid
ith weight of vertical distance hi from the original centre-of- While methods do exist for finding the centroid of a mass by
gravity (wi will be a negative value for the removal of a weight). direct measurement, they are unlikely to be undertaken. Provided
the inclining weights are not rotated when moved, the position of
hi wi
δ Gi ¼ ð38Þ the centroid is not important. That is to say, the distance moved by
Δ þ wi the centroid will be the same as the distance moved by any point
The sensitivity coefficients for a shift in the centre-of-gravity, of reference. Therefore, careful attention to the procedure can
due to the addition or removal of an ith weight are given in remove this source of uncertainty.
Eqs. (39)–(41).
  4.9.2. Uncertainty of the marks made on deck for longitudinal
∂ δ Gi hi Δ
c9i ¼ ¼ ð39Þ placement
∂wi ðΔ þ wi Þ2 The difficulty with the longitudinal marks is more one of
  finding a suitable point of reference. If a hatch combing or
∂ δ Gi wi
c10i ¼ ¼ ð40Þ accommodation block bullhead is utilised for reference, then the
∂hi Δ þwi
uncertainty in their placement must be considered. Sighting
  transversely across the deck, at right angles to the parallel-mid-
∂ δ Gi  wi hi
c11i ¼ ¼ ð41Þ body, is again not without difficulties. Considering this, an
∂Δ ðΔ þ wi Þ2
uncertainty of approximately 10 cm is reasonable. While this may
The standard uncertainty of the mass of the ith weight uðwi Þ sound alarmingly large, remember this value will be multiplied by
and the height of the ith weight uðhi Þ should be taken as the a sensitivity coefficient. This then considers the sensitivity coeffi-
combined uncertainty of the calibration uncertainty of the devices cient for the change in I and ∇ with respect to a small change in
utilised to measure them (or a realistic estimate) and the uncer- trim. These terms will be negligibly small provided the trim is
tainty in the measurement. The standard uncertainty for the minimal.
84 M.D. Woodward et al. / Ocean Engineering 114 (2016) 79–86

4.9.3. Uncertainties when utilising ballast tanks as inclining weights 6. Expanded uncertainty (U)
In some cases the general arrangements of the ship prohibit the
use of mobile inclining weights. In such cases, the ballast tanks are The combined uncertainty is defined as equivalent to one
employed as an alternative. For example, a port side tank may be standard deviation. This corresponds to a confidence interval of
filled. Then, when ready, the tank will be emptied and an approximately 68% if the uncertainty can be assumed to be nor-
equivalent tank on the opposite side filled. In such cases, the mally distributed. In engineering applications a higher confidence
uncertainty is related to the relative positions of the centroid of interval when expressing the uncertainty is more practical. This
can simply be achieved by multiplying the combined uncertainty
each tank, the volume of each tank and the density of the fluid
uc by a coverage factor k, which gives the expanded uncertainty U.
used to fill them. Taking the root-sum-square for these items then
For example, assuming a normal distribution, k¼ 2 gives a U 95
the sensitivity can be taken with respect to the induced moment.
with a 95% confidence interval.
Also, the uncertainty in any free-surface correction must be taken
into account.
7. Method

An experimental uncertainty analysis may be performed prior

5. Combined uncertainty
to the IE, as a process of experimental design, or post analysis to
establish a confidence interval in the result. The main difference is
It is not uncommon in an IE to take multiple measurements by
that, prior to the test being conducted, the limits of some para-
additional or repeated weight movements. Estimates of the ran-
meters must be estimated. In either case the calculations are
dom uncertainty from the standard deviation of the mean are
relatively straightforward and can be performed easily with a
possible, if multiple truly independent measurements are made. typical spread-sheet application. Also in either case, the process is
This however provides only the uncertainty in the estimate of GM predominantly the same and can be structured into seven key
and not KG. The estimates of KB and BM, both necessary for the steps as described in Fig. 1.
estimate of KG, are dependent on parameters also measured as Utilising the methods outlined for Step 6 (Sections 4.7 and 4.8),
part of the IE; and must be dealt with appropriately. Notwith- the uncertainty in GM for any load condition can readily be
standing, more likely the individual measurements are not truly obtained.
independent. For example, the second induced angle may include
the moment from both the first and second inclining weights.
Similarly, a third weight move may be achieved by returning the 8. Case studies
first weight to its original position. The uncertainties should thus
be assumed to be fully correlated and combined accordingly. As, in To establish the fitness-for-purposes of the procedure and to
this case, the data reduction equation is a simple average then the meet with the objectives of the paper, the procedure is applied to
uncertainty for fully correlated variables is also a simple average, five case-study ship inclining experiments. In line with the
objectives of the paper, the results are utilised to find the uncer-
given by Eq. (49).
tainty in the estimated KG and, explore the origins of contributing
  1X N   uncertainties to help target improvements in the experimental
uc KG ¼ u KG i ð49Þ procedures.
Ni¼1
As the data is historic, not all of the necessary parameters
specified by this procedure are available. Nevertheless, the data
For items that are to be removed (as described in Sections serves perfectly well to perform a typical pre-test analysis. This
4.7 and 4.8), the corresponding
  uncertainties should be included has in fact some advantages in that environmental inputs are
after the samples of u KG i are combined utilising Eq. (49). made the same for all five ships, making the results more directly

Fig. 1. Step-by-step flow chart for the implementation of the uncertainty analysis procedure for an inclining experiment, either pre- or post-analysis.
 M.D. Woodward et al. / Ocean Engineering 114 (2016) 79–86 85

Table 1
Results from case study ships.
   
Parameter Lbp (m) ΔDesign KG (m) uc KG U 95 (m) U
95 GM
(units) (tonne)
(m) (%)a

Buoy 37 453 3.580 0.075 0.15 100

tender
Super 50 698 4.340 0.016 0.033 22.0
yacht
Supply ship 51 904 4.173 0.024 0.047 31.3
Container 124 15,718 10.245 0.014 0.029 19.3
Ropax 204 23,370 16.620 0.077 0.15 100

a
The expanded uncertainty is given as a % of an assumed metacentric height of
0.15 m.

comparable. For example, in all cases wave amplitudes of 5 cm are

assumed for the water surface when taking draught readings.
Similarly, plumb-line readings are all assumed to be oscillating
with amplitude of 1mm and the water density is assumed to have
a 5 kg/m3 standard uncertainty in all cases. In all cases the man-
ufacturing tolerances are assumed to be 7 10 mm in the length
and 73 mm in all other dimensions.
For commercial considerations, the full details of the particular
ships are not published. Table 1 contains however all of the per-
tinent values necessary to form a judgment. For reference, the
table gives the length between perpendiculars and the design
displacement for each ship, together with a descriptive ship-type
Fig. 2. Component uncertainty contribution in the vertical location of the centre of
title. In each case, the estimated KG is given together with the
mass for various inclining experiment parameters for the five case study ships.
combined uncertainty and the expanded uncertainty for a 95%
confidence interval. This contains only the uncertainty associated
various inputs for each ship. On examination the importance of
with the light ship estimate and not the uncertainty of all other
heel angle is clear; and notably so for both the Buoy tender and the
items (cargo, fuel, water, ballast etc.) on-board the ship in its
Ropax. The Buoy tender does have the smallest average induced
loaded condition. The uncertainty of the light-ship KG is the
heel angle ( E0.6°). The Ropax however has an average induced
minimum possible values and the implication of this for the
operation of the ship is certainly worth considering. Comparing heel angle in line with and sometimes greater that the other ships
the uncertainty in KG to the value of KG is not particularly examined. This is a function of the sensitivity of the results to the
meaningful in this case as the magnitude of KG is somewhat heel angle that depends on the relationship between various
arbitrary, and will change as the ship is loaded. Strictly speaking, parameters (heel angle; plumb-line length; plumb-line oscillation;
the uncertainty in the position of the centre of mass ðGÞ is GM). This perhaps exemplifies well the value of performing a pre-
important and not its magnitude with respect to an arbitrary test uncertainty analysis, to avoid such situations. Notwithstand-
reference point such as the keel ðK Þ. Consideration of the magni- ing, the two parameters here that may be readily controlled are
tude of the expanded uncertainty for a typical GM limitation is the induce heel angle (which should be appropriately large) and
perhaps more meaningful. Considering the basic IMO requirement the plumb-line length (which should be as long as possible).
for GM to be greater than 0.15 m, the given values of expanded The second most influential parameter appears to be the
uncertainty can simply be added on to find the necessary GM that draught measurement. In actual fact, the 5 cm wave amplitude is
would have a 95% confidence of achieving the given criterion. For most likely very optimistic, and could be much larger. Never-
comparison, the percentage of expanded uncertainty with respect theless, repeating the draught measurement more than once
to an assumed GM of 0.15 m is given in the last column of Table 1. quickly reduces the uncertainty in this parameter. Establishing the
For the three smaller ships, if the confidence interval encom- minimum number of necessary draught measurements needed for
passes a potentially negative GM, this does not necessarily present any particular wave condition is a relatively easy process using this
a problem, as they would not normally be loaded to this limit (or procedure.
be required to do so). In the case of the container ship however The next most important parameters appear to be the estimate
GM would typically be close to this limit; to prevent high roll
of displaced volume, followed by the estimate of the second
accelerations that might otherwise cause damage to the container
moment of water plane area. These are dependent on the manu-
stacks. In this case the ship would have to be loaded to a GM value
facturing tolerances, and the estimate thereof. Of course, this can
of nearly 0.18 m to ensure a 95% confidence of compliance. Simi-
vary depending on the shipyard. More sophisticated ways of
larly, the Ropax would typically load close to the limiting GM to
measuring the ‘as-built’ form/dimensions may be considered if
reduce the risk of high acceleration causing a shift of cargo. In this
case the ship would have to be loaded to a GM value of nearly this parameter is identified as significant.
0.33 m; more than double the criterion limit. Note, this estimate It is worth also considering the inferred relationships from the
does not accounting for uncertainties in the loading of the ship; sensitivity coefficients. Assuming that the ship is simply a box with
that could be much larger. the same length, breadth and draught but with a block coefficient
It is clear from the results that the magnitude of estimated tending to unity, then Eq. (12) can be substantially simplified. The
uncertainty varies widely for the ships considered; with at least centre of buoyancy of a box is always at half the draught, so
∂KB
one, the Buoy tender, showing a markedly high value. To better ∂T ¼ 0:5. Also, the second moment of water plane area does not
∂I
explore the origins of the uncertainties, the contributions from change with draught, so ∂T ¼ 0. Substituting also Eq. (1) and
various inputs are examined. Fig. 2 gives the uncertainties for recognising that GM  BM ¼ BG, Eq. (12) can be reduced to the
86 M.D. Woodward et al. / Ocean Engineering 114 (2016) 79–86

simplified form given in Eq. (50). Appendix A. Standard deviation of a sinusoidal signal
∂KG ∂∇ BG
¼ 0:5  ð50Þ Taking the definition of standard deviation to be given by
∂T ∂T ∇
Eq. (A1.1), where xi is the ith sample amplitude, μ is the mean
This indicates that, to reduce sensitivity, BG must be as high as value of all samples and N is the number of samples.
possible. As the height of the centre-of-buoyancy at a particular vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u N
draught is fixed by the geometry of the ship, a more generally u1 X  2
σ¼t x μ ðA1:1Þ
inference can be made in that the centre of gravity must be as high Ni¼1 i
as possible. Also Eq. (50) indicates that ∇ must be as small as
possible. Inspection of Eq. (1) shows that both situation result in The mean value μ, of a sinusoidal signal, between the limits of
increased induced heel angles. Some caution should be exercised zero and 2ωπ , will be by definition zero. Then, replacing xi with ζ
however as, while large heel angles may reduce uncertainty, they sin ωt [where ζ is the amplitude, ω is the frequency and t is time]
will at the same time increase error due to changes in the position we get Eq. (A1.2).
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
of the metacentre. Nevertheless, heel angles in excess of 7° would u
u Z2ωπ
be needed before metacentric theory is seriously compromised; uω
far in excess of those needed for a successful IE. σS ¼ u
t ζ 2 sin 2 ωtdt ðA1:2Þ
2π
0

Integrating between the given limits, gives:

9. Conclusions and recommendations
2π
Zω
2ωπ
t sin ð2ωt Þ
The aim of this study was to establish procedures for identi- ζ2 sin 2 ωtdt ¼ ζ
2

fying the experimental uncertainty in the estimate of KG, obtained 2 4ω 0
0
by IE. The objective were to give procedures for performing a pre-
test analysis to help reducing the experimental uncertainty and which, by substituting in the values for the limits, can be seen to
equal ζωπ . Substituting this back into Eq. (A1.2), we obtain:
2

post-test analysis to identify a confidence interval for the resulting

estimate of KG. sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ω ζ2 π
A procedure is provided together with case studies, demon- σS ¼
strating how the uncertainty in an IE can be utilised to improve 2π ω
the design of the experiment. No one parameter can be identified Cancelling out, the standard deviation of a sinusoidal signal for
in all cases as problematic from the case studies. There is however any number of whole cycles, is by definition thus given by
a strong indication that the uncertainty in the heel-angle mea- Eq. (A1.3).
surement is important but this may be a function of other factors
ζ
such as GM. Nevertheless, the longest possible plumb-line (or σ S ¼ pffiffiffi ðA1:3Þ
perhaps an electronic alternative) with sufficiently large induce 2
heel angles should help to reduce uncertainty. The draught mea-
surement uncertainty was also seen to be important, but can be
substantially improved with increased sample size. Also, the References
knowledge of the ‘as-built’ condition in terms of manufacturing
tolerances was identified as important. If this were identified as IMO, International Code on Intact Stability, 2008 (2009 Edition), IMO Publishing,
critical for any particular ship, alternative methods could be ISBN: 978-92-801-1501-7, MSC 85/26/Add.1 ANNEX 2, pp. 71–89.
ISO/IEC Guide 98-3:2008, Uncertainty of Measurement – Part 3: Guide to the
employed to establish the as build dimensions more accurately. Expression of Uncertainty in Measurement (GUM:1995).
A procedure is provided for estimating a confidence interval for ITTC, Fresh Water and Seawater Properties, ITTC Procedure 7.5-02-01-03, Effective
KG and argued to be more usefully considered as a confidence Date 2011, Revision 02.
ITTC, Recommended Procedures and Guidelines: Guide to the Expression of
interval for GM. The case studies show that, for some ships, a
Uncertainty in Experimental Hydrodynamics, 7.5-02-01-01, Effective Date
substantial increase in the minimum GM may be necessary to 2008, Revision 01, pp. 27.
ensure safe operation.
In addition to the original objectives, the methods outlined for
the addition or removal of weights and for free-surface correction,
provide a full and complete procedure for establishing the
uncertainty in GM for any load condition.

View publication stats