CONTAINER SHIP UPDATE - DNV GL SE · PDF fileFor 18,000 TEU ships, the material scantlings are between 80 and 95 mm, where the larger thicknesses are for high tensile steel with

No 01 2014

Container ship

UPDATE No 01 2014

Performance monitoring

Development of large container ships

Hull optimisation

DNV GL

2 Container Ship Update

CONTENTS

Front cover photo: © DNV GL – Per Sverre Wold-Hansen

1204 10

Development of large container ships

– what are the limits? ..................................................................... 4

Will the Suez Canal limit the size of ULCS? .............................. 8

Hull optimisation – making good ships better ....................... 10

Performance monitoring

unlocks dormant energy savings ...............................................12

LC and RSCS – boosting container ships’ cargo intake

and stowage lexibility and improving safety ..........................16

Parametric roll – risk reduction

through real-time detection ..................................................... 20

LNG as fuel is spreading into container shipping ................. 23

ReVolt – the unmanned, zero emission,

short sea ship of the future ....................................................... 24

Recent newbuildings ................................................................. 30

© H

ase

np

usch

No. 01 2014

Container Ship Update 3

NEEDS AND EXPECTATIONS

Your needs and expectations matter to us. “Cost control and eficient operation”, “safety

and compliance”, and “ship availability” came out on top when we asked for your top three

concerns in the present market situation. “Your front-line people, the surveyors and the

approval engineers, make the difference,” is what we heard.

Daniel Abt is one of them. He is in charge of our successful RSCS and LC class notations.

He talks about the logic and advantages of the new lashing philosophy – to adapt the

requirements to the area where the ship operates. After all, “North Atlantic winter” condi-

tions do not exist everywhere in the world. This makes good sense. At the same time,

parametric rolling has challenged the industry for decades. Roberto Galeazzi has com-

pleted his PhD thesis at the Technical University of Denmark. This thesis is about giving

reliable warning prior to the onset of parametric rolling. His algorithm has been tested

out on data collected by DNV GL from full-scale measurements at sea and is part

of our focus on safe cargo operations.

Cost control and eficient operation are all about optimising ships for their future service

and trading area as well as monitoring and improving performance during operation. We

bring you some interesting articles by Volker Bertram, a well-known specialist in the indus-

try. He guides you through the many alternatives to be considered, with sound advice

based on his life-long experience.

Size matters, when we are talking about container ships. But what governs the development

of bigger ships? What are the limits? Will we see 24,000 TEU ships on the seven seas in

a few years’ time? We try to provide you with some of the answers on the next pages.

The Suez Canal limits the size of tankers, but container ships carry volume cargo, so they

are different.

We hope you ind this issue interesting. We will be happy to discuss any queries you

may have.Published by DNV GL Editorial committee: Jan-Olaf Probst, Jost Bergmann, Knut A. Døhlie Magne A. Røe, Editor Lisbeth Aamodt, Production

Design and layout: coormedia.com 1408-001

DNV GL NO-1322 Høvik, Norway Tel: +47 67 57 99 00

© DNV GL www.dnvgl.com

Jan-Olaf ProbstDirector Business DevelopmentExecutive Vice [email protected]

Jost BergmannBusiness Director Container [email protected]

CONTAINER

SHIP

UPDATE

© D

NV

GL

DNV GL


Text: Jan-Olaf Probst, Jost [email protected], [email protected]

Beside the economies of scale, the development of several types

of container ships was inluenced by the restrictions imposed by

the present Panama Canal locks, which allow free operation of ves-

sels with a beam of up to 32.25 m. The new Panama Canal locks

under construction can handle vessels with a beam of up to 49.0 m,

i.e. container ships of up to 12,800 TEU.

In addition to the Panama Canal, the limitations on the draft and

length in South America and on the air draft in the Port of New York

have a signiicant inluence on the Paciic and Trans-Atlantic trades.

These boundary conditions are a driving factor in that different

large container ships have been developed for the Asia–Europe

and South America–Europe/Asia trades. Especially on the Asia–Eu-

rope trade, ever-larger container ships can be used because up to

now there have been no harbour or operational limitations.

Will the trend for ever-larger container vessels continue or is there

any limitation from a technical point of view? This question should

be discussed in this paper by verifying the limitations of the

principal dimensions.

Parameter ship length

In regard to the ship length, the maximum value quite often men-

tioned is 400 m, which results from a former and present require-

ment that, for vessels above this length, the actual wave behaviour

in sea conditions must be veriied by a direct wave analysis, a

requirement which is nowadays no longer an issue.

In the case of container ships, the standardized container box allows

clear enlargement steps which result from the container itself, the

clearance in the cargo hold, and a typical transverse bulkhead

width which all add up to 14.6 m.

Enlarging length inluences the wave bending moment and the

still water moment by the power of two. This factor has a signii-

cant inluence on the steel plates needed in the upper hull girder

of a container ship due to the cross-section as an open u-shape.

For 18,000 TEU ships, the material scantlings are between 80 and

95 mm, where the larger thicknesses are for high tensile steel with

a yield strength of 400 N/mm2 and the smaller value is for a yield

strength of about 470 N/mm2, both of which are recognized mate-

rials. However, there are other criteria, such as the production of

thick high-tensile steel plates and Rule requirements for the global

strength of the bottom area, which limit the use of the 470 N/mm2

yield strength steel.

Considering a single main engine design of an 18,000 TEU

container ship with four 40´ bays behind the funnel, and eleven

40´ bays between the funnel and deck house, then a practical

elongation would be to add one further 40´ bay forward of the

engine room, combining the small cargo hold above the engine

room to form one cargo hold with two 40’ bays. This arrangement

will lead to twenty-ive 40´ bays in the vessel, and a length overall

of approximately 414 m, which could still be designed in normal

practice. Lengths above this value are possible, but have to follow

other cross-section arrangements which must be developed and

investigated.

The harbour limitation with regard to the berth length or diameter

needed for turning inside the harbour is a factor which should not

be underestimated.

Parameter ship breadth

Enlarging the breadth of a vessel had a large impact on the neces-

sary main engine power for a fast vessel where speed was a main

design parameter in the past. This requirement no longer exists

DEVELOPMENT OF LARGE

CONTAINER SHIPS

– WHAT ARE

THE LIMITS?

Economies of scale are the most important factor for container ships in service and the

development of new designs. The transportation rate per TEU has driven the trend towards

larger vessels and has particularly contributed to the great success of post-Panamax vessels.

Today, post-Panamax container ships above 11,000 TEU have the major share of the total

transport volume and slot capacity. Vessels approaching the capacity of 18,000 TEU are already

in operation and several will be delivered during the next 12 months, and the step to 20,000 TEU

is already under development.

No. 01 2014


because the maximum design speed has been reduced from 25

knots to 23 or even 22 knots, and this has a large inluence on the

opportunity to adjust the length-to-breadth ratio. The possibility

of reducing the length of the vessel and enlarging the beam by

keeping the same nominal capacity has been used by many yards

in present designs. Enlarging the breadth and reducing the length

have a positive inluence on the total building costs, a factor which

counts in these times of lower newbuilding prices.

Compared to the length, the beam would be enlarged by 2.5 m

steps due to the standardized container size including cell guide

and clearances. From a strength point of view, the enlargement

in width would have only a linear inluence on the wave as well as

on the still water bending moment. That means widening a design

has a much smaller impact on the global longitudinal strength

than elongating the length. Beside this effect, the widening of

a longer vessel has a better inluence on the enlarged nominal

capacity than an elongation.

However, widening a vessel will have a negative inluence on the

so-called warping deformation, which is the deformation of both

sides of the vessel in different longitudinal directions. This defor-

mation has a large inluence on the hatch cover movements and

therefore on the wear and tear of the bearing pads, which leads to

maintenance costs or an increase in the number of hatch panels.

In addition, the deformation has an inluence on the containers in

the cargo hold and the interaction of the containers on the hatch

cover with the lashing bridge. To reduce the warping deformations

for larger container vessels the deck house has been shifted to the

forward area to achieve a closed strength member below the deck

house leading to a twin island design. Compared to the elonga-

tion of the present designs, widening is the more practical solution

from a strength point of view.

The rolling and transverse acceleration forces of container ships

depend on the width of the vessel and inluence the permissible

loads on the lashing equipment as well as the crew on the bridge

deck. This effect has led to increasingly high lashing bridges with

only one target – to allow the stowage of heavier containers in

higher tiers or reduce the racking loads from the transverse forces

during rolling on the lower container. Increasing the vessel’s beam

will have an inluence on the loadable containers so that the nomi-

nal container intake igure is not comparable to the real container

intake igure.

The current limiting factor of some container terminals in regard

to the 60-metre gantry crane outreach should be considered when

widening present designs – especially on the Asia to Europe trade

© H

ase

np

usch

DNV GL


where large vessels will be deployed – but is not a limiting factor

for the vessel itself.

Parameter ship depth

The vessel’s depth depends on the number of containers in the

cargo hold, the height of the hatch cover and the space between

the top of the container and the lower side of the hatch cover. In

addition, a passageway of at least two metres has to be ensured

above the upper deck and top of the hatch cover. The height of

the container itself, which is today 9’ 6” for a standard 40’ contain-

er as well as the usual height for reefer containers, must be taken

into consideration. However, the majority of containers in use to-

day are still of the 8’ 6” size and the fact that the nominal container

capacity of a container ship will be measured in 8’ 6” tall TEUs

(twenty-foot equivalent units) is more important. Considering this

aspect, there are two possible container arrangements, ten tiers of

high cube 9’ 6” containers with a height of 28.96 m or eleven tiers

of standard 8´ 6” containers with a height of 28.50 m. This results

in a difference of only 459 mm at the side depth, which means

that most designs could accommodate ten tiers of high cube in

the cargo hold.

According to ISO 1496/1, the lowest container in the hold may be

over-stowed by 192 t (based on a maximum vertical acceleration

of 1.8 g), which would result in an average container weight of

around 31 tons for a 40´ container and 28 tons in the case of eleven

tiers, which is equal to the typical 14-ton homogenous loading

conditions. Exactly this loading condition is a main contractual item

between the shipowner and building yard, so yards and designers

are hesitant to enlarge the number of tiers in the cargo hold or the

ship depth because then the maximum container load would differ

from the maximum 14-ton homogenous loading condition in the

cargo hold and would thus lead to a different interpretation.

Although the test load of the container itself has been changed

from 86t to 96t since 2005, this does not result in a higher side

depth for the vessel unless there is an intermediate stopper for

the container stowage in the cargo hold, which is not practical

during the loading and unloading of the vessel.

In the case of container ships, three different drafts could be con-

sidered – the design draft, the scantling draft and the operational

draft. Traditionally, the design draft is used for contractual items

such as speed and cargo capacity, while the scantling draft is the

basis of all International Regulations as well as Class Rules. Today,

the different operational drafts are used to calculate the speciic

fuel oil consumption at different loading conditions. Starting from

about 13,000 TEU ship size, the design draft is between 14.0 and

14.5 m, while the scantling draft is in general 15.5 to 16.5 m.

However, by keeping the vessel’s depth and draft and widening

the vessel, the additional freeboard will be reduced which could

lead to weathertight hatch covers having to be installed – leading

to more maintenance work and costs. The ship owner should

therefore carefully consider if the large scantling draft is needed

or whether maybe the same deadweight can be achieved by

reducing the draft and having a larger block coeficient – an

indication of a fatter vessel.

Another aspect of the draft-to-beam ratio is the Suez Canal. The

Canal limitations are set by a combination of draft and beam. A

draft of 16.8 m would allow for a maximum beam of 60 m, which

does not impose any restriction on current designs. However,

should the beam increase to 65 m, then the corresponding draft

limitation would be 15 m and could thus inluence the lexibility of

new designs.

Economy of scale – where is the end?

Vessel size

14,000 TEU 16,000 TEU 18,000 TEU 21,000 TEU

Uti

lisa

tio

n

100% 100% 97% 91% 89%

95% 105% 101% 96% 94%

90% 110% 106% 101% 98%

85% 117% 112% 106% 103%

80% 123% 119% 112% 109%

75% 131% 126% 119% 116%

Development of container ships. The evolution of the container ship 1960–2015

20,000

18,000

16,000

14,000

12,000

10,000

8,000

6,000

4,000

2,000

–

1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015

TE

U

Year built

No. 01 2014


Bay 13 14 15 16 17 18 19 20 21 22 23 24 25 26Row13 3,500 TEU 3,500 TEU 3,650 TEU 4,300 TEU 4,900 TEU 5,060 TEU

(32.25 m) (212 m) (225.5 m) (254 m) (262 m) (275 m) (283 m)14 4,250 TEU 4,500 TEU(35.0 m) (253.4 m) (268.5 m)15 3,600 TEU 4,500 TEU 4,600 TEU 4,900 TEU(37.5 m) (219 m) (249 m) (254.7 m) (269.2 m)16 5,500 TEU 5,900 TEU 6,800 TEU(40.0 m) (257.4 m) (273.45 m) (300 m)17 7,090 TEU 8,063 TEU 8,600 TEU(42.8 m) (300 m) (323 m) (334 m)18 8000 TEU 9,000 TEU 9,200 TEU 10,000 TEU(45.6 m) (300 m) (320 m) (336.7 m) (349.7 m)19 8,800 TEU 11,500 TEU 12,600 TEU(48.2 m) (300 m) (349.7 m) (366 m)20 13,300 TEU 14,000 TEU(51.2 m) (366 m) (383 m)21 14,800 TEU 16,000 TEU(54.0 m) (383 m) (399 m)22 Emma M CMA CGM(56.2 m) (397m) (399 m)23 19,000 TEU 20,270 TEU 20,800 TEU(58.6 m) (400 m) (414 m) (429 m)24 20,750 TEU 21,250 TEU 21,750 TEU(61.2 m) (400 m) (414 m) (429 m)25 21,700 TEU 22,200 TEU 22,750 TEU(63,8 m) (400 m) (414 m) (429 m)

Development of container ships

Parameter air draft

The maximum number of containers above the hatch cover is

restricted to the container strength, analogue to the depth of the

vessel. The major difference is the racking load due to the rolling of

the vessel – which does not exist within the cargo hold because of

the cell guides. This effect could be compensated by higher lash-

ing bridges, which are today up to three tiers, and different stow-

age devices. In this respect, the number of tiers on the hatch cover

have been step-wise increased to 8, 9 and 10 by the installation of

1-, 2- and 3-tier lashing bridges. It is important that the total stack

load can only be slightly enlarged due to the container strength of

the lowest container, but the weight distribution in the stack can

be inluenced. There are already designs and ships waiting to be

delivered that have eleven tiers on the hatch cover. Another factor

is that the number of tiers on the hatch cover will enlarge the deck

house and therefore the height of the bridge deck. This will have

a larger inluence on the transverse acceleration and therefore on

the safety of the crew. The installation of protection for the conning

position as used on VLCCs may be necessary.

Port restrictions like the bridges in Osaka, Hong Kong and Hamburg,

as well as the air draft limitation below the gantry cranes of some

harbours on the Asia to Europe trade, may also be limiting factors

for the vessel and must be taken into consideration by the operator.

Conclusion

The request of container liners to operate vessels above 20,000

TEU exists. However, at present we have to solve technical restric-

tions and the main issue is the maximum plate thickness in the

upper hull girder. A simpliied elongation or widening of present

designs is possible to a certain extent. Further enlargement is

only possible with a different design layout similar to the step

from semi-aft deck house design to the twin-island design. Beside

this technical restriction the port facilities or canal limitations as

mentioned should not be underestimated and be considered for

the next steps. ❚

24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1

Illustra

tion

: HH

I

DNV GL


Text: Knut A. Døhlie [email protected]

WILL THE SUEZ

CANAL LIMIT THE

SIZE OF ULCS?

The Suez Canal is undergoing continuous development with a

substantial growth in capacity, allowing ever-bigger ships to pass

through it. The current ship size limitations were published in 2010

– the Suez Canal Authority Beam and Draft tables (Circular 2-2010).

Combinations of beam and draft produce a “bounding box”, limit-

ing the ship’s cross-sectional area. There is no restriction on the

ship length at present. What happens if the ship length increases

signiicantly beyond 400 m is yet to be seen.

A Suezmax tanker has a draft of 21.1 m and a beam of 50 m, produc-

ing a bounding box of 1,005 m2. Liquids are heavy cargo and draft

is therefore often the limiting parameter for tankers. The beam is

taken from the tables as a function of the draft. Containers are vol-

ume cargo and other combinations of beam and draft are possible.

Design Loa [m] B [m] Tdesign [m]

Tscant [m] Airdraft [m]

DSME 405 59.0 14.6 16.0 73 from keel

HHI 400 58.6 14.5 16.0 71.2 from keel

STX 401 59.0 14.5 16.0 68.2 from keel

OMT/SWS 399 58.5 14.0 16.0

Suez bounding box

Unlimited 60.0 16.8 68 above sea

Suez bounding box

Unlimited 65.0 15.0 68 above sea

Cross sections of 18+kTEU designs compared to Suez limitations

Current 18,000 TEU newbuilds are designed with a cross section close to the limits imposed by

the Suez Canal. With a beam of less than 60 m and a scantling draft of 16 m, they are within the

limits set by the Suez Canal Authorities (SCA). The Canal limitations may, however, restrict the

choices for designers should the industry go for even bigger ships.

No. 01 2014


Suez Canal

Centre Line

Container ship

B=65m, T=15m

Container ship

B=60m, T=16.8m

Suezmax tanker

B=50m, T=21.1m

24 m

Half cross section of canal with half cross

section of ship type boundary boxes

156.5m (full width=313m)


24m

313m

24m 121m

5200m2

CONTAINER SHIP

B=60m T=16.8m

CONTAINER SHIP

B=65m T=15m

SUEZMAX TANKER

B=50m T=20.1m



Suez Canal cross section

Suez Canal half-cross section

Bounding boxes giving maximum dimensions

© F

oto

-do

ck

All current 18k designs have a beam of just under 60 m, primarily

due to the outreach capacity of port cranes. The maximum draft at

this beam is 16.8 m, which is fairly close to the scantling draft of all

current designs (16 m), so the cross section of the current 18k de-

sign is close to the limits. To increase capacity, the obvious answer

would therefore be to increase the ship length. However, length is

the most expensive dimension to increase, so wider and deeper

ships may be a preferable option to increase capacity.

A ULCS in the range of 20k–24k TEU would in such case beneit

from a wider beam allowing for better hull optimisation. A 65 m

beam would give a permissible draft of 15 m, which may be an

acceptable design draft for normal operations even though the

scantling draft would be higher. This could in fact be a not insignii-

cant limitation.

The Canal has undergone a continuous widening and deepening

over the decades, increasing its capacity in step with the general

increase in ship size. Is this trend likely to continue? We believe

that will depend on demand and on the earning potential for the

Canal. So far, the SCA has not published any plans for a further

general increase in depth and cross-sectional area. ❚

Suez Canal cross section with ship type “bounding boxes”

“Bounding box” of typical ship designs

DNV GL


Text: Volker Bertram [email protected]

Largest potential for fuel savings lies in design

High fuel prices and IMO regulations to curb CO2 (carbon-dioxide)

emissions put pressure on ship owners to obtain more fuel-eficient

ships. The largest fuel saving potential lies in the design stage.

Traditional ship designs relect much lower fuel prices than the in-

dustry is facing today and is likely to face tomorrow. Consequently,

these designs are no longer it for the current market or for the

future. Hull optimisation is one of the “low-hanging fruits” for

signiicant fuel savings. In fact, the hull optimisation of ultra-large

container ships may have payback times that can be measured in

weeks.

The term “hull optimisation” is widely used and in many cases

abused. Often the term “improvement” would be more appropri-

ate than “optimisation”. Simulation-based hull improvement is

commonly employed. In this design approach, a few (typically

fewer than 20) hull variants are generated and assessed using

computational luid dynamics (CFD) and/or model tests. This is

standard design procedure and employed by many model basins,

but we frequently see that signiicant improvements beyond this

approach are possible. (Formal) optimisation looks at thousands

or even tens of thousands of designs and uses an optimisation

algorithm to ind the best design, achieving typically 4–6% more

than the standard CFD guided approach (Hochkirch and Bertram,

2012).

Geometrical constraints are usually speciied by the shipyard.

These concern bow lare, clearance for the propeller and rudder

and hard points for the engine in the aftbody. In addition, there

are indirect geometry constraints on the displacement volume

and initial stability. A key input from ship owners is the considera-

tion of the ship’s actual operational proile instead of just one

design (or contract) point (Hochkirch et al., 2013). A “point” in this

respect is the combination of load condition and speed. Ships are

frequently operated at lower speeds or partial drafts (with vary-

ing trim angles). The design should in such case strive for the best

compromise between these conditions, yielding e.g. the lowest

yearly fuel costs.

Numerical towing tests and optimisation

The key to unlocking the theoretical potential often lies with the

ship owner and building speciications. A case study may illus-

trate the application in practice. United Arab Shipping Company

(UASC) wanted to expand its leet to include new super-eficient

container vessels. Setting its sights on competitive slot cost, UASC

planned to order an initial ive 14,000 TEU and ive 18,000 TEU

vessels.

Pre-contract phase

Recognising fuel consumption as the single most important factor

in determining the proitability of new vessels, UASC partnered

HULL OPTIMISATION

– MAKING GOOD

SHIPS BETTER

Formal hull optimisation typically leads to 4–6% savings for large

container ships, depending on the owner-speciied operational

proile and yard-speciied design constraints. Such hull optimisation

should already be included in the build speciications, ensuring

smooth cooperation between the owner, yard and service provider.

Numerical towing tank tests based on CFD can help ensure an

objective third-party assessment of the design as early as in the pre-

contract phase.

No. 01 2014


with FutureShip (now DNV GL Maritime Advisory) in order to ind

the most eficient design possible. A number of designs were

assessed based on the total cost of transport per container-mile

for UASC’s speciic trading pattern, taking into account both

capital investment and operational costs.

Four sets of designs were shortlisted for intensive evaluation

using numerical tank towing tests based on Computational Fluid

Dynamics (CFD) simulations for the ship with propeller. In only

four weeks, FutureShip ran thousands of tests to determine the

speed-power relationship for the two ship classes at two drafts for

each of the four competing designs. The use of massively parallel

computation allowed multiple tests to be run at the same time,

over speed and draft ranges which relected realistic operational

proiles. The four designs were all well matched, with Hyundai

Heavy Industries (HHI) emerging as the winner.

Post-contract phase

Taking HHI’s designs as the starting point, the ship owner, shipyard

and FutureShip joined forces to ine-tune the vessel performance

in a formal parametric optimisation. The optimisation employed

more than 60 free parameters with the objective of reducing fuel

consumption as much as possible, taking into account hydro-

dynamic power requirements, the speciic fuel oil consumption

of the respective engines and UASC’s speciic operational proile

for speed-draft combinations. HHI and FutureShip revisited the

powering concept and more than 35,000 hull shape variants were

investigated for each hull design. As a result, better performance

could be achieved for both the 14,000 TEU and 18,000 TEU vessels

compared to the initial proposal. For inal validation, model tests at

the Hamburg Ship Model Basin conirmed the CFD predictions.

Cooperation between owners, yards and optimisation experts

is vital

The inal signiicantly higher fuel eficiency of the new designs was

the result of fruitful cooperation between the ship owner, shipyard

and DNV GL Maritime Advisory as a ship-optimisation service pro-

vider. Thus, fuel savings start with the owner speciications. ❚

ReferencesHochkirch, H.; Bertram, V. (2012), Hull optimisation for fuel eiciency – Past, present and future, 11th Conf. Computer and IT Applications in the Maritime Industries (COMPIT), Liege (2012), pp.39–49Hochkirch, K.; Heimann, J.; Bertram, V. (2013), Hull optimisation for operational proile – The next game level, 5th Int. Conf. Computational Methods in Marine Engineering (MARINE), Hamburg

Screen shots of CFD analyses for 14,000 and 18,000 TEU container designs©

DN

V G

L

© D

NV

GL

DNV GL


Text: Volker Bertram [email protected]

PERFORMANCE MONITORING

UNLOCKS DORMANT

ENERGY SAVINGS

© G

etty Im

ag

es

No. 01 2014


Monitoring – why and why now?

Performance monitoring of ships is by no means new to the indus-

try. Jensen and Söding (1985) present one of the irst attempts at

an onboard hydrodynamic monitoring system. Grabellus (1990)

of Hamburg-based ship operators Hamburg-Süd expanded on

their work to create a voyage-optimisation advisory system based

on data collected regularly on board ships in operation. Grabellus

determined the ship performance based solely on onboard data,

using statistical regression techniques. Voyage legs in shallow

waters in port approaches were not evaluated. A lot of data were

acquired automatically, but some had to be added manually. The

system was abandoned after some years due to frequent sensor

failure and overall disappointing accuracy.

The subsequent decades have seen assorted variations on this

theme. These have varied in detail, but overall shared a similar

approach. For almost three decades, performance monitoring has

remained an elusive issue outside the limelight of mainstream ship

operations. Several game-changing developments have recently

altered this:

■■ In 2013, IMO made the SEEMP (Ship Energy Eficiency Manage-

ment Plan) mandatory for all ships. Monitoring the effective-

ness of measures to increase energy eficiency is one of its four

elements. The irst year after its introduction brought a sobering

realisation for many ship operators: our traditional monitoring

systems fall short in terms of accuracy and insight.

■■ Increased fuel prices combined with low freight rates brought an

end to previously more lenient business relationships between

charterers and owners. A surge in claims questioning contrac-

tual performance led to both owners and charterers measuring

and assessing the performance of ships more closely, avoid-

ing disputes through more accurate monitoring on both sides.

Similarly, coating manufacturers and suppliers of energy-saving

devices have an interest in demonstrating measurable savings to

their customers.

■■ The proposal of a new ISO standard for the hydrodynamic per-

formance of ships in service, ISO (2014), stimulated additional

discussion and illuminated the shortcomings of older perfor-

mance monitoring systems. The open discussion with interna-

tional panels of experts brings more transparency and allows

performance monitoring methods to be compared. DNV GL has

from the beginning taken an active role in furthering these goals.

Navigator PerformanceSmart solution for onboard voyage reporting and onshoreoperations and leet performance analysis

Modules:Voyage administrationShip general dataEvent reportingFleet voyage/statusDashboards

DNV GL


You cannot manage what you do not understand

All performance monitoring systems have a similar basic ap-

proach: raw data acquired on board are iltered and corrected

for external inluences. The corrections are based on engineering

models which differ in sophistication, accuracy and required effort.

Errors in the resulting hull performance rating thus stem from two

main sources:

Input data error

Onboard sensors have inherent limited precision. The crew is also

a “sensor” in this sense, adding errors due to faulty estimates,

deliberate lies or reporting errors. Speed through water is a noto-

rious example of sensor-based errors: “Although the speed log is

one of the oldest sensors on board ships, it is still one of the most

inaccurate instruments and it does not give the speed through

water with an acceptable accuracy” (Boom et al., 2013).

Model errors

Models are by deinition approximations of reality. Some inlu-

ence factors are omitted for convenience, making the model

faster and easier to handle. Other factors may be approximated,

for example using “representative” or “average” semi-empirical

estimates that are derived from many ships and are thus by

nature not exact for an individual ship. Ignoring the trim and

using cruise model-test-based interpolation to show the effect of

draft on the required hull power are typical model errors in most

performance monitoring systems. In our experience, container

vessels are particularly sensitive to such modelling errors. Model

test results typically cover only two load conditions (design

and sea trial), and only a small speed range near to the design

speed. Power requirements are extrapolated for lower speeds

and interpolated for intermediate draft and trim values between

the speed-power curves for design and sea trial conditions. Even

for tankers, average errors of 7% are added by this popular, but

dangerously crude, model approach (Bertram and

Lampropoulos, 2014).

The more you know, the faster you learn

While each tool must have an engineering or hydrodynamic

knowledge base, the approach chosen to generate this knowl-

edge base determines the costs and accuracy of the software.

There are two fundamentally different approaches to deriving

performance monitoring systems.

The irst approach is based on system identiication of the actual

ship. Typically some machine-learning techniques are employed

which perform in essence the task of putting smooth “curves”

through data. The more parameters involved and the more ran-

dom scatter the data have, the slower the computer learns. For

example, ambient conditions introduce unavoidable scatter in

data. Typically, machine-learning systems must be exposed to an

initial dedicated training period with varying key parameters (such

as draft, trim and speed for hull performance) during days when

the ambient conditions do not contaminate the data sets too

much (“sea trial conditions”). After that, illing in missing patches in

the knowledge base and updating the existing knowledge takes

lifelong learning. This continuous learning is called “dynamic” in

marketing jargon.

The second approach uses insight into physics and advanced

simulation techniques to feed the knowledge base. The advan-

tage of this approach is that there is no random ambient con-

tamination. Simulations allow systematic and complete investiga-

tions of possible operational conditions, including for rare and

extreme occurrences. The disadvantage is that any model is only

an approximation of reality. If models are validated, i.e. prove to

be accurate against precisely measured data, this approach is

superior to system identiication in terms of speed and cost, and

sometimes also in terms of accuracy.

In practice, the best approach uses insight into physics and

advanced simulations as far as possible to build the knowledge

base and reduce the remaining system tuning to a minimum. The

more knowledge furnished in the beginning, the faster and better

machine-learning approaches can ill the gaps. In short, the more

you know already, the faster you learn the rest.

DNV GL’s solutions – committed to user-friendliness, transparency

and accuracy

Building on our long experience with onboard measurements and

advanced engineering, DNV GL has developed software to sup-

port the shipping industry focussing on three key pillars:

■■ User-friendliness – minimise the input for users, especially

for onboard systems; use intuitive displays; display only key

information for quick access, using a pull-down menu for added

options and information on demand.

■■ Transparency – document the underlying engineering models

and system identiication approaches in publications, often with

peer review; we strive to use the best possible approaches and

believe that transparency is crucial to build trust.

■■ Accuracy – using data checking at the source, adapted iltering

and calibration combined with advanced simulation techniques,

performance monitoring systems can be made signiicantly

more accurate.

Our Navigator Performance software is the backbone for subse-

quent business intelligence analyses using ECO Insight. Navigator

Performance collects data using a stringent “enter data once only”

philosophy. The system then re-uses existing data for assorted re-

porting tasks to different stakeholders (owners, operators, authori-

ties, cargo owners, etc.), signiicantly reducing the crew’s adminis-

trative workload. ECO Insight uses these data and possibly other

monitoring systems to track the performance of the hull, propeller,

machinery and main consumers.

Through transparency to savings

The accuracy of performance monitoring depends on the quality

of the data input, the sampling frequency and the quality of the

engineering model for correcting changing ambient and opera-

No. 01 2014


tional conditions. Trend monitoring supports management deci-

sions on daily operation and maintenance strategies.

The key to unlocking the dormant savings potential lies in shared

transparency between all stakeholders, including onboard crews.

A new generation of DNV GL software tools now supports these

performance monitoring tasks, incorporating stringent data re-use

for user-friendliness and cutting-edge simulation technology for

accuracy. ❚

ReferencesBertram, V.; Couser, P. (2014), Computational methods for seakeeping and added resistance in waves, 13th Int. Conf. Computer and IT Applications in the Maritime Industries (COMPIT), Redworth, pp.8–18

Bertram, V.; Lampropoulos, V. (2014), A hull performance monitoring model for energy eicient operation, 6th European Conf. Production Technologies in Shipbuilding (ECPTS), Hamburg

Boom, H.v.d.; Huisman, H.; Mennen, F. (2013), New guidelines for speed/power trials level playing ield established for IMO EEDI, Hansa Journal 150/4ISO (2014), ISO 19030 (Measurement of changes in hull and propeller performance), 2nd revised draft, International Organization for Standardization, Geneva

© D

NV

GL

Saving fuel with ECO Insight A container vessel operator utilises four sister vessels in the same trade pattern. Speed-performance evaluations revealed large diferences in speed proiles on comparable voyages. By closely monitoring the average speed and waiting times through ECO Insight, the overall average speed could be reduced without changing schedules, with an associated fuel saving of three percent just from this single measure.

DNV GL


LC AND RSCS –BOOSTING CONTAINER SHIPS’ CARGO

INTAKE AND STOWAGE FLEXIBILITY

AND IMPROVING SAFETY

Text: Daniel Abt [email protected]

The maximum stack weight and the weight distribution within the

stack are key design criteria for container ships – for boxes carried

both on deck and in the hold. The limits for both parameters are to

a large degree determined by the loads acting on the containers

during the voyage, mainly represented by the transverse accelera-

tions that are experienced.

Until recently, the transverse acceleration calculations were usually

based on empirical rule formulas which had to be adapted con-

stantly in order to cope with the growing size of container ships.

As an alternative, we offered for many years the option to apply a

design-speciic “Individual Acceleration” analysis, which resulted

in tailored and more precise loads.

Based on experience from classifying more than 2,500 container

ships and continued research and development work, including

onboard measurements, we started about two years ago to de-

velop a new concept for calculating the loads acting on containers

and lashing equipment. The aim was to maximise a vessel’s cargo

capacity while at the same time improving safety and reliability.

No. 01 2014


What is the core idea behind the LC/RSCS approach?

The core idea was to update the container lashing rules for a

modern approach to cover all ship sizes and their motion charac-

teristics within one set of approximation formulas. This would offer

a uniied way to calculate consistent load assumptions without

the need for high-effort direct calculations for each individual ship

class.

The route-speciic extension was a side product born from the

idea to check the inluence of route-dependent trading patterns.

As the irst estimations showed signiicant beneits, the idea has

been developed further without complicating the application of

the container lashing system and with ample attention to maintain-

ing safety margins.

What does LC mean and what benefit does it bring?

We introduced the new class notations LC and RSCS to be able

to differentiate between vessels using old rules and vessels which

fulil the latest requirements. As RSCS is just optional, two class

notations had to be created, LC for unrestricted worldwide service,

which is mandatory for all newly contracted container ships, and

RSCS for sailing in restricted sea areas.

The new approximation formulas are based on an evaluation of

the direct calculations of maximum transverse accelerations car-

ried out by us for hundreds of different container vessels during

the past 20 years, as well as onboard veriication measurements

Compliance with the new concept is documented by two class

notations LC (Lashing Computer) and RSCS (Route Speciic

Container Stowage), which will be assigned to either new-

buildings or existing vessels that can be adapted to the new

standard.

The innovative concept is explained by DNV GL’s Senior

Engineer and responsible technical expert Mr Daniel Abt.

Daniel Abt is our expert advisor for the LC and RSCS notations.

© D

NV

GL

© H

ase

np

usch

DNV GL


done on a number of different sized container ships. As a con-

sequence, with the new unrestricted service load assumptions,

a beneit of 3–10% – depending on the vessel’s size – can be

achieved. In general, one can say – the larger the vessel, the big-

ger the acceleration beneit given an equal loating condition.

All relevant container vessels must now be equipped with an ap-

proved lashing computer, something which was not mandatory

before and was only applied by a few owners. This is in order to

ensure that the container securing equipment is not overloaded.

What does RSCS stand for and what is in it for owners?

RSCS means Route Speciic Container Stowage and is an option

for newbuild vessels as well as for existing ships. Direct calcula-

tions based on route-speciic wave scatter data showed that a

uniied route-speciic reduction factor for transverse accelerations

could be determined irrespective of the vessel size. That is why,

for a speciic route, just one reduction factor can be applied to

unrestricted service accelerations as determined according to LC

requirements. This keeps the whole calculation system very simple

and thus easy to apply.

Further, to the beneit of the new unrestricted accelerations, the

route reduction factors offer reductions of between 3% (North

Atlantic Service) and 18% (North Sea – Baltic Service). This can

result in a signiicant increase in the intake of laden containers and

higher stowage lexibility.

When was the concept officially released and on how many ves-

sels has the concept now been implemented?

Formal inclusion in the rules was on 1 May 2013. Until now, about

80 vessels have fulilled all approvals and the inal survey on board

– for both legacy GL and legacy DNV.

Which owners are typically applying for the concept/these class

notations?

Up to now, the new class notations have been granted to vessels

from Panamax-size (about 5,000 TEU) up to 14,000 TEU post-

Panamax-ships. Most of the owners charter their vessels to the

big liners, which are also starting to apply the new rules to their

own vessels now. The main customers of this service as of now

are Reederei Claus-Peter Offen, Rickmers Reederei, E.R. Schiffahrt,

Reederei NSB, Danaos, China Shipping, Hapag-Lloyd, Seaspan,

APL, Hansa Shipping and many more of our good customers.

How many vessels are in the pipeline and how are we dealing

with the huge demand?

About 300 vessels are in the approval process right now, while

we expect even more to apply for the concept in the near future.

These class notations are also available for all DNV GL classed

vessels and a irst newbuild series applying the RSCS concept

has been inalised, with more to follow.

} Increased stackloads for the Europe to South America trade due to RSCS class notation.

}} IACS scatter diagram with the Europe to South America plotted.

No. 01 2014


Due to this outlook, we carried out cross-organisation training

of colleagues to increase the number of our employees that can

serve our customers faster in a consistent way.

Have we experienced any damage on ships where LC or RSCS has

been implemented?

We are of course glad that we have not been notiied of any inci-

dents on vessels utilizing our LC and RSCS concepts. By introduc-

ing LC/RSCS, the safety margin related to the load assumptions

has been lowered a little, but this has been compensated by

increasing the safety level at another point – we introduced the

mandatory requirement that the ships had to be itted with an ap-

proved lashing computer system, a requirement which was already

overdue in our opinion.

We are convinced this will create more sensitivity to the limits of

the lashing system in the container shipping business. It’s the irst

time that loading conditions have to be checked by an approved

lashing computer system, which will deliver consistent results

independent of the manufacturer.

We are supported in our decision by the publication of similar

ideas by other IACS class societies, which seem to be working on

similar concepts. ❚

A 19% increase of the stackload for the Asia to Europe trade with RSCS notation.

DNV GL


Text: Roberto Galeazzi, Associate Professor, Technical University of DenmarkContact: [email protected]

PAROLL is an innovative condition-monitoring system for the timely detection of parametric

roll on merchant vessels. It has been invented and developed by the Technical University

of Denmark. DNV GL and Wallenius Marine have supported the development and full-scale

validation of this monitoring system.

PARAMETRIC ROLL– RISK REDUCTION THROUGH

REAL-TIME DETECTION

Parametric roll in head seas is nowadays a well-known resonance

phenomenon that threatens a ship’s stability by inducing rapidly

growing extreme roll motion, and hence may cause considerable

damage in terms of cargo losses, hull integrity and crew safety.

Signiicant research followed the multimillion-dollar incidents

suffered by the APL China in 1998 and Maersk Carolina in 2003.

Since then, DNV GL has actively contributed to the understanding

of the phenomenon’s root causes in order to improve the opera-

tional safety of merchant vessels.

In 2006, the article “Parametric Rolling – a problem solved?” was

published in the DNV Container Ship Update. In this, DNV GL

Vice President Knut Døhlie made a clear analysis based on the

know-how accumulated by DNV through scrutiny of ships’ motion

data and research regarding container-ship seakeeping. Analysis

of Paciic and Atlantic passages revealed that the availability of

weather routing systems and implementation of navigation strate-

gies could mitigate the risk of parametric roll occurring by sailing

across routes that minimize the vessel’s exposure to head-sea

conditions. This approach completely disregarded the physics of

the phenomenon based on frequency and phase synchronization

between roll and pitch motions and aimed at preventing para-

metric roll by avoiding head-sea conditions. Sometimes, however,

head sea is simply where you have to go.

19:00 20:00 21:00 22:00−20−10

0

10

20

φ[deg]

0

0.5

1

A

19:00 20:00 21:00 22:000

0.5

1

Sφ

2θ

φ2θ

19:00 20:00 21:00 22:000

1

2

3

4

5

G

Time[hh:mm]

00.51

Roberto Galeazzi, Associate Professor, Technical University of Denmark

No. 01 2014


Døhlie pointed out that, although ensuring a safe passage is

certainly important, it is also signiicant to assess the effects that

this strategy has on schedule reliability and fuel consumption.

Hence, he identiied the need for second-generation warning

systems to not only provide operational guidance in the form of

polar diagrams to indicate the risk of parametric roll based on

the combination of the ship’s forward speed and heading, but

also monitor the presence of the physical conditions that trigger

parametric roll.

The PhD project Autonomous Supervision and Control of Para-

metric Roll Resonance (2006–2009), run at the Technical University

of Denmark by Ass. Professor Roberto Galeazzi under the super-

vision of Professor Mogens Blanke and in collaboration with Ass.

Professor Niels K. Poulsen, responded to this call for innovation by

investigating signal-based detection methods which could extrapo-

late from ship-motion measurements the existence and persistence

of the conditions for parametric roll to unfold. The research project

resulted in PAROLL, a patented condition-monitoring system. Initial

testing of data from model tests performed by Dr Gaute Storhaug

– DNV GL Principal Specialist – showed the potential of the detec-

tion algorithms. However, to obtain robust routines, extensive

testing on real full-scale motion data with and without parametric

roll was paramount.

Parametric roll – triggering conditionsParametric roll is a ship dynamic stability problem that afects large merchant vessels such as container ships and car carriers. Empirical conditions have been identiied that may trigger parametric roll:•Theperiodoftheencounterwaveisapproximatelyequaltohalfofthe

natural roll period•Thewavelengthisbetween1to2timestheshiplength•Thewaveheightisgreaterthanaship-dependentthreshold•Theship’srolldampingislow

When these conditions are met and the ship sails in moderate to heavy longitudinalorobliqueseas,thenthewavepassagealongthehullandthe wave-excited vertical motions result in variations of the underwater hull geometry, which in turn change the roll-restoring characteristics.

} Fig. 2 Detection of parametric roll events on board a 2,800 TEU container ship. The risk index informs the crew that conditions for the triggering of parametric roll are present from about two hours prior to the large event occurring a few minutes before 21:00.

| Fig. 1 A 2,800 TEU container ship experiences parametric roll while crossing the Atlantic. The wave radar and spectral analysis show that the wave peak period is close to half the natural roll period. The time analysis confirms that before and during the first parametric roll event, roll and pitch are synchronized in phase.

19:00 20:00 21:00 22:00−20−10

0

10

20

φ[deg]

19:00 20:00 21:00 22:00−5

−2.50

2.55

θ[deg]

Time [hh:mm]

0.1 0.2

45

225

90270

135

315

180

0

Wave radar - Ti me: 20:00

0.1 0.2

45

225

90270

135

315

180

0

Wave radar - Ti me: 21:30

0 0.1 0.20

0.2

0.4

0.6

0.8

1

f [Hz]

Power spectra

Φ/ Φmax

Θ/ Θmax

20:54 20:56 20:58 21:00−20

−10

0

10

20

Time[hh:mm]

φ [deg]θ [deg]

DNV GL


The development has continued through a Proof-of-Concept

project in collaboration with the Norwegian University of Science

and Technology (NTNU) and DNV GL as an industrial advisor;

later on Wallenius Marine entered the project with their know-how

regarding ship operators. The research collaboration with DNV GL

and Wallenius Marine allowed the unique opportunity to inally

test and validate the PAROLL monitoring system on large full-scale

motion data sets.

PAROLL has also attracted the interest of providers of decision-

support systems such as Amarcon, a member of the ABB group,

which is integrating it as a part of its ship-motion monitoring and

advisory system.

PAROLL: real-time detection of parametric roll

PAROLL is a novel condition-monitoring system that timely detects

the development of parametric roll on board merchant vessels re-

lying on low-cost motion sensor information. PAROLL implements

signal-based automated detection algorithms, which extrapolate

information about the levels of frequency and phase synchroniza-

tion between the roll and pitch motion.

Two statistical change detectors are at the core of PAROLL: the

spectral correlation and the phase synchronization detectors.

The irst assesses if the natural roll period is approximately twice

the period of the pitch oscillation that in turn relects the wave-

encounter period. The second monitors if the roll and pitch

motions are synchronized in phase or, quoting Knut Døhlie,

if “the roll and pitch peaks are lining up”. If both detectors’

outputs are above their thresholds, the monitoring system

issues an audible alarm.

To provide the crew with an intelligible system that gives ad-

equate information from an operational perspective, the output

of the monitoring system has been enhanced with a colour-coded

risk coeficient which provides a real-time measure of closeness

to a parametric-roll event by combining the current outputs of the

two detectors. The colour-coded risk coeficient can help gener-

ate a state of alert for the navigator and will allow the crew to

start taking pre-emptive actions to counteract parametric roll and

mitigate its effects before the phenomenon unfolds to its possibly

devastating magnitudes.

From an operational viewpoint, it is also important not to increase

the level of nuisance on the bridge, where several other decision-

support systems are integrated. Due to this, PAROLL has been

further enhanced with a check-of-the-roll amplitude, which lastly is

used to determine whether to issue an alarm. The selection of the

roll level that serves as a inal alarm threshold should be based on

irm investigations into the oficers’ attitudes and it can therefore

be tuned on board to achieve the best balance between early

alerts and avoiding unnecessary alarms on the bridge. It is noted

that the visual warning level is exclusively based on the risk coef-

icient calculated from the outputs of the two detectors, while it is

solely the audible alarm that can be adjusted to users’ needs via

the roll amplitude level parameter.

It is important to emphasize that both detectors only use the

measurements of the pitch and roll angles provided by the on-

board IMU. No knowledge about the speciic vessel being moni-

tored is needed, making the PAROLL system portable and robust

against ship-model uncertainties.

Full-scale validation on container ships and LCTCs

A full-scale validation of the PAROLL monitoring system was inal-

ized during the irst quarter of 2014. This was uniquely possible

thanks to the valuable collaboration with DNV GL and Wallenius

Marine, which have provided long-term motion data for two dif-

ferent vessels: selected parts of data from a 2,800 TEU container

vessel for which data were available over a two-year period, and

motion data from all the voyages of a large car carrier during a

one-year period. Further, motion data from two of the parametric

roll events reported in the literature have been made available by

Wallenius Marine.

The validation has been a true success, showing that PAROLL

is effectively a robust and reliable solution for the timely detec-

tion of parametric roll; conirming the already promising results

obtained earlier on model tests data. In approximately 70% of

the cases where parametric roll determined motions larger than

ten degrees, PAROLL provides an alarm as soon as ive to 40 roll

cycles before the maximum roll amplitude are achieved, meaning

that the crew has between 1.5 and 12 minutes to take pre-emptive

actions.

From the investigations of motion data from the large car carrier

and container vessel, it is possible to conclude that parametric roll

is far more common than is reported. With PAROLL, timely detec-

tion is inally available so that remedial actions can be taken well

before parametric roll develops to severe magnitudes. ❚

No. 01 2014


Text: Gerd-Michael Wü[email protected]

LNG as a ship fuel has been discussed in shipping circles for ap-

proximately ive years now. However, the irst non-gas-carrier ves-

sel running on LNG came into operation in Norway in 2000. This

was the only such vessel for 13 years, until the irst LNG-fuelled

vessels not operating in Norwegian waters came into service in

2013. Now, LNG as a ship fuel is spreading into worldwide com-

mercial shipping. Today, 50 LNG-fuelled ships are in service and

six of them operate outside Norway. This year was the irst time

that the order book for LNG-fuelled ships contained more ships

than the number of LNG-fuelled ships in operation. Sixty-six ships

are on order and only 19 of them will operate in Norway, while 20

will operate in the US and 23 in Europe outside Norwegian waters.

The irst container ship on order will come into service in 2015

but 14 of the 66 vessels on order are container ships, ten of them

classed by DNV GL. All container ships on order will operate in

European and North American waters. They are the answer to the

SECA/ECA fuel-cost challenges.

The low gas prices in the US make LNG a fuel alternative for all

vessels sailing between the US and Europe or US and Asia, even

in competition with today’s HFO prices. With the IMO worldwide

sulphur limit of 0.5% becoming applicable in 2020, global con-

tainer shipping will also beneit from lower fuel costs. First movers

are preparing for this by using DNV GL’s LNG Ready Approval In

Principle. For example, UASC is currently building 17 large con-

tainer vessels (14,000 and 18,000 TEU capacity) that will also be

prepared for running on LNG. DNV GL certiied these vessels with

an Approval In Principle for the LNG system which is intended to

be installed if the vessels are converted to LNG.

With an increasing number of vessels running on LNG, the avail-

ability of this fuel will improve very quickly. DNV GL predicts 1,000

vessels operating on LNG in 2020 and these vessels will need

four million tonnes of LNG as fuel. In 2020, we expect approxi-

mately 250 vessels running on LNG to come into service. This will

be approx. 9% of all newbuildings. On the other hand, 91% of

all newbuildings will not be built for using LNG. It is obvious that

these 91% should at least be LNG ready to ensure that they will be

able to operate successfully during their lifetime. DNV GL can help

them be really LNG ready with its LNG Ready service, which is a

modularized service that includes modules for reviewing docu-

ments in an early project state and Approval In Principle certiica-

tion during the design phase. ❚

LNG AS FUEL IS

SPREADING INTO

CONTAINER SHIPPING

0

40

80

120

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

Number o

f ships

Ships on order

Ships in operation

LNG ready ships

Development of LNG fuelled fleet as per August 2014

© U

AC

S

DNV GL


Text: Hans Anton Tvete, Bjørn-Johan Vartdal [email protected], [email protected]

The road network in Norway and the EU is heavily congested. At

the same time, the use of road transport for the freight of goods

is steadily increasing. The consequences are increased road wear,

more accidents and higher emissions. Towards 2040, a large

increase in population growth in the urban regions will lead to

an even greater demand for the transport of goods. This trend

will make demands on the roads which are beyond much of what

today’s road network can handle. Governmental bodies in Norway

and the EU acknowledge this fact and want to move more freight

from roads to sea and rail. However, primarily due to the high

energy cost, high operational costs and high taxation level, the

proit margins in the short sea shipping segment are small.

DNV GL Strategic R&I has designed a concept ship, aimed at the short sea shipping segment,

with the objective of moving freight from road to sea. Taking the design and technologies to

the extreme is intended to provoke discussions while addressing DNV GL’s ambitions for a

safe and sustainable future for shipping.

ReVolt– THE UNMANNED, ZERO EMISSION,

SHORT SEA SHIP OF THE FUTURE

No. 01 2014


Can present-day technology improve the cost and effectiveness

whilst improving the safety and environmental performance of

short sea shipping? To answer this question, a multi-disciplinary

team at DNV GL has come up with the concept ship “ReVolt – the

unmanned, zero emission, short sea ship of the future”.

ReVolt – main features

■■ Fully battery-powered

■■ Ballast-free design

■■ Unmanned operation

■■ 2-bladed propellers

■■ Fast cargo handling

■■ No direct emissions

Battery-powered

■■ A power demand of only 50kW in calm sea

■■ Short transport legs

■■ Re-charging facilities in every port

■■ High eficiency

■■ No direct emissions

Ballast-free design

■■ Inclined keel

■■ Straight bow

■■ Less wetted surface

■■ Less need for rotating machinery and ballast water

treatment plant

Autonomous operation

■■ Increased safety

■■ Increased situational awareness

■■ Lower operational cost

Propulsion

■■ High eficiency

■■ No cavitation

■■ Podded propulsion for increased manoeuvrability

■■ Retractable bow thruster for port manoeuvring

Cargo handling

■■ Extended cell guides

■■ No manual lashing

Operational profile

ReVolt’s operational proile has been established by analysing

Automatic Identiication System (AIS) data from vessels that oper-

ated in the Norwegian economic zone (NEZ) in 2012. The ship

type and cargo capacity are obtained by pairing the AIS data with

the DNV GL ship register data. This has in turn been translated

into ReVolt’s needs and requirements.

By counting unique IMO numbers from the AIS database, it has

been identiied that general cargo ships are the most represented

ship type in the NEZ. As much as 23.4% of all vessels traficking

in or around Norway are in fact general cargo vessels. Further-

more, research carried out by the Norwegian Institute of Transport

Economics with respect to freight eligible for transfer from road

to sea and rail has singled out general cargo as a cargo type that

is particularly suitable for this transfer. This cargo can be contain-

erised, which will promote multimodality and enable fast cargo

handling. As a result; a container-type vessel has been chosen as

the basis for further design.

Following analyses of the AIS data, the following requirements

were chosen as the basis for ReVolt’s operational proile:

■■ Speed: six knots

Table 1 – ReVolt main particulars (60.3, 5.4, 2-bladed)

Main particulars Capacity Machinery

LOA 60.3 m Cargo capacity 100 teu Battery 3000 kWh

LPP 58 m Deadweight 1,250 mtPropulsion:2 x azimuth pods @ 120 kW each 2-bladed propellers, diameter 3 m

1 x retractable bow thruster

Beam 12 m Cruising range 100 nm

Depth 12 m

Draught (full) 5.4 m

Draught (ballast) 4 m

Service speed 6 kn

DNV GL


■■ Range: 100 nautical miles

■■ Cargo capacity: 100 TEU

■■ Route: coastal trafic from Oslo to Trondheim

■■ Port stay: four hours on average

The low speed of six knots will force the logistics chain to be built

up in a completely new manner. The transit times will inevitably be

longer but, by applying a conveyor belt approach to the logis-

tics chain – with frequent departures and reduced downtime – it

should be possible to transfer applicable cargo types to this mode

of transport.

Hull

The ship’s hull has been designed to optimise ship eficiency

whilst ensuring that the safety and operational requirements are

fulilled.

One additional requirement for ReVolt was the need to be able

to operate without the use of ballast water, thus doing away with

the requirement of equipment for ballast water management. This

reduces the vessel’s energy demand and need for maintenance.

Due to ReVolt’s low speed, resistance will mainly be due to hull

friction and any external forces acting on the vessel. Wave-making

resistance will be modest. For this reason, a straight vertical bow

design was chosen in order to minimise the resistance over the

whole operational proile. Choosing a bulbous bow design will

only prove a good solution for a few load conditions where wave-

making resistance makes a signiicant contribution. In addition,

a sharp waterline will have a piercing effect and be favourable in

adverse wave conditions. For the design chosen, CFD calculations

show a calm water resistance of only 50kW. However, due to the

low speed, added wave and wind resistance was shown to con-

tribute quite a lot to the overall ship resistance. By incorporating

metocean data sets from along the intended route into the resist-

ance calculations, an average ship resistance of 120kW was found.

An investigation into the possible use of composite materials was

carried out. Although it was found that the lightweight material

would lead to a greater decrease in the wetted surface than nor-

mal steel would, the ship’s propulsion eficiency would decrease

as a result of the composite material yielding a lower draught,

which would in turn require a propeller with a smaller diameter

and therefore lower eficiency. Hence, considering the overall ef-

iciency, steel proved to be the best solution. Steel is also a proven

and less costly hull material.

Propulsion

For ReVolt most of the normal constraints concerning propel-

ler design are absent due to the low vessel speed. Maximum

diameter, blade strength, RPM, noise and vibration tolerances

predetermine much of a propeller’s geometry but, apart from the

diameter and blade strength, these criteria are pretty much non-

existent for ReVolt. Hence the designers could focus on optimis-

ing the eficiency. Cavitation is also more or less absent. This has

made it possible to reduce the number of propeller blades to

only two. In turn, this reduces the viscous losses in the propeller

considerably and, for the inal design, a propulsion eficiency of

76% was achieved.

ReVolt is to operate independently of tugs and needs good

manoeuvrability. Due to the shape of the hull and in order to

reduce the amount of rotating parts on the ship, podded propul-

sion was chosen over conventional shafted propulsion. Two stern

pods for main propulsion and one retractable bow thruster for

manoeuvring were selected.

Machinery

A fully battery-powered solution was selected for ReVolt in order

to maximise eficiency, eliminate local emissions and reduce the

number of rotational components requiring maintenance. Pro-

vided the power used to charge the electricity comes from hydro-

power, batteries will provide a highly eficient propulsion system.

The energy loss from the water reservoir to the propeller is esti-

Figure 3 – Podded propulsionFigure 2 – CFD analysis of ship resistanceFigure 1 – AIS-plotted trade route

No. 01 2014


mated to be only 40%. In contrast, similar diesel-powered ships

have well-to-propeller losses of up to 85%. Assuming hydropower

is used for charging, batteries can also be considered a zero emis-

sion solution. Finally, to minimise the maintenance requirements

in order to enable an autonomous solution, batteries represent a

nearly maintenance-free alternative.

Autonomy

The shipping industry’s safety record is currently relatively poor,

with an average of 900 fatalities yearly. This number is 90% higher

than for comparable land-based industries. Studies show that as

much as 85% of accidents in shipping are the result of human

error. DNV GL has an ambition to reduce the number of fatalities

to that of land-based industries. If this ambition is to be met, the

accidents caused by human errors need addressing. One way to

do so is by introducing automation to support or replace the hu-

man element.

One concern with respect to the reduction and even elimination

of the crew is the need for continuous maintenance. To address

this, ReVolt has been designed to minimise the maintenance re-

quirement. The main cause of technical breakdowns in shipping is

related to rotating machinery. A ballast-free, fully battery-powered,

unmanned vessel like ReVolt will have a minimal need for rotating

equipment. In fact, the only rotating machinery on board ReVolt is

related to the propulsion pods and bow thruster, which in turn are

located outside the ship hull.

When it comes to autonomous navigation, sensor fusion from sen-

sors like ECDIS, GPS, radar, cameras and LIDAR has the potential

to create complete situational awareness around the vessel. These

are sensors which are all available on the market today.

ReVolt has been made fully autonomous with the purpose of

taking the different applications and technologies to the ex-

treme, and DNV GL believes that many intermediate steps like

condition- and sensor-based monitoring, enhanced navigational

assistance and remote operation will have to be taken before fully

unmanned ships can become a reality.

Cargo handling

Due to ReVolt’s low speed, it is important not to waste unneces-

sary time in port. Hence, a quick turnaround time in port is of the

essence. By using state-of-the-art technology in automatic moor-

ing systems, like grip arm and vacuum-based mooring, ReVolt

will be moored fast and without the need for ropes and winches

which are highly dependent on manual assistance and regular

maintenance.

With the aim of saving time, several automatic options for loading

and discharging the containers were explored. The beneit of an

automatic cargo handling system, apart from a faster handling

time per container, is that the cargo handling process is less

dependent on manual assistance. However, the disadvantages

are that such systems occupy potential cargo space and are quite

maintenance–intensive, with a high risk of breaking down. Given

this fact, it was decided that all cargo handling should be car-

ried out from the shore side using standard container cranes. By

extending the hull sides as high as the containers are stacked, cell

guides can be installed in the full height of cargo holds, which in

turn will provide faster cargo handling and eliminate the need for

stevedores and additional lashing.

In order to ensure the fast transfer of cargo from the ship to other

modes of transportation, the ports’ shore facilities also need to be

highly effective. Having dedicated terminals with easy access for

trucks will achieve this.

Energy efficiency measures

By analysing historical metocean data sets, statistical distributions

of average and extreme sun, wind, wave and current conditions

were established at evenly spaced points along the route. The

data was used to analyse the potential for utilising energy eficien-

cy measures based on renewables, such as solar panels, foils, kites

Figure 5 – Wireframe arrangementFigure 4 – Battery pack

DNV GL


Figure 6 – Cell guides in the cargo hold Figure 7 – ReVolt lifetime cost

Figure 8 – ReVolt model demonstrator Figure 9 – ReVolt

02 4 6 8 10 12 14 16 18 20 22 24 26 28 30

10

20

30MUSD

Years

40

50

60

70

Replacement of battery pack

Conventional diesel powered shipReVolt

and lettner rotors. Given favourable weather conditions, these

measures have the potential to improve the energy eficiency sub-

stantially. However, in order to evaluate the real added beneits of

the different measures, the average beneit of each measure must

be calculated. As a result, the average beneits are moderate and

cost-effective implementation is questionable.

There is also a paradox associated with installing energy eficiency

measures on a fully battery-powered vessel. For safety purposes,

the battery pack will need to be dimensioned and be of such a

size that it is not dependent on additional assistance from the

energy eficiency measures in any operational mode. The cost

beneit of the measures will therefore be limited to the smaller

requirement for battery charging in port. With a relatively low price

of electricity, the return on investment period will be too long to

justify the added cost of such measures.

Cost

ReVolt is unique in terms of safety and environmental perfor-

mance. However, one question still remains; can this solution be

cost effective?

ReVolt’s autonomous capabilities eliminate the need for facilities

related to crew. The superstructure, auxiliary machinery and outit-

ting can be signiicantly reduced, if not eliminated. The absence of

crew will also give increased space for payload. The battery pack

on board ReVolt is, however, extremely capital intensive, with an

estimated cost of 1,000 USD/kWh. Due to the battery’s perfor-

mance degradation, this is also a cost which will be incurred twice

over the vessel’s estimated lifespan of 30 years. However, this is

a cost that is likely to be reduced as battery technology matures,

and predictions indicate a signiicant reduction in battery price.

Taking into account local incentive programmes like the Norwe-

gian NOX-fund, it is estimated that the CAPEX of ReVolt is equal

to that of a conventional ship of equivalent cargo capacity.

No. 01 2014


I� �� , � � � ��a��at �x�� a� ��e�t� ��tt o� ��

� ��f ��g� a� �a��a�� tt o� �educed signii-

cantly compared to those of a diesel-powered ship. The crewing

costs will also be reduced. How much will depend on the shore

infrastructure needed to maintain the autonomous operations.

As a result, ReVolt is proitable from day one. Over its lifetime,

ReVolt has an estimated reduction in operational costs of 34 MUSD

compared to a conventional diesel-powered ship. In addition,

future governmental emission reduction incentives are likely to

increase these margins further.

Model demonstrator

In order to demonstrate ReVolt’s autonomous capabilities and

test other features of the design, a model demonstrator has been

produced. The demonstrator is a 1:20 scale model in which all the

concept’s design features are replicated. The model is powered

by batteries and propelled by two azimuthing stern pods and

a retractable bow thruster. The demonstrator will be itted with

cameras, GPS, compass and motion sensors to ensure situational

awareness around it.

By using the demonstrator to learn about the challenges and pos-

sibilities related to autonomous navigation, DNV GL is preparing

for a future where autonomy is part of the solution for increasing

the level of safety at sea.

A vision for the future

ReVolt is a vision for the future and will not be built until several

of the technologies introduced become more mature. However,

building and operating this ship would be feasible with today’s

technology. ReVolt is intended to serve as inspiration for equip-

ment makers, ship yards and ship owners in their efforts to de-

velop new solutions on the path to a safe and sustainable future. ❚

DNV GL


RECENT

NEWBUILDINGS

Delivery of the irst 10,000 TEU container ship HANJIN BUDDHA at

the Jiangsu New Yangzi Shipbuilding Co., Ltd. earlier this year.

From left to right; Mr Lee, Young Hwan, Project Manager, DNV GL, Mr Chen, Keng, Area Manager, DNV GL, Mr Du, Cheng Zhong, Vice General Manager of YZJ Shipbuilding Group, Mr Torgeir Sterri, Regional Manager, DNV GL, Mr Wang, Jian Sheng, General Manager of YZJ Shipbuilding Group, Mr Huang, James, Regional Market and Business Development Manager DNV GL, Ms Hou, Juzhen, Regional Business Development Director, DNV GL and Mr Wang, Dong, Vice General Manager of YZJ Shipbuilding Group

© D

NV

GL

© D

SIC

Da

lian

Sh

ipb

uild

ing

Ind

ustry C

o. L

td

No. 01 2014


Name TEU Owner/manager Shipbuilder Class notation

APL ��OSA 13,892 NOL Liner (Pte.) Ltd/Neptune Shipmanagement

Hyundai Samho Heavy Industries

DNV +1A1 Container Carrier E0 CLEAN BWM-T COAT-PSPC(B) BIS TMON NAUTICUS(Newbuilding)

THALASSA MANA 13,806 Enesel S.A. Hyundai Heavy Industries Co GL +100 A5 Container Ship RSD(gFE) IW ERS BWM(D2) DG +MC AUT CM-PS EP-D

ULSAN EXPRESS 13,167 Hapag-Lloyd AG Hyundai Heavy Industries Co GL +100 A5 Container Ship IW NAV-O RSD ERS BWM(D2) DG +MC AUT EP-D

HANJIN BUDDHA 10,000 Seaspan YZJ 983 Ltd/Seaspan Ship Management

Jiangsu New Yangzi Shipbuilding Co., Ltd.

DNV +1A1 Container Carrier SafeLash DG-P COMF-V(3)C(3) E0 NAUT-OC CLEAN BWM-T COAT-PSPC(B) BIS TMON NAUTICUS(Newbuilding)

CSCL SUMMER 10,000 CSCL Summer Shipping/China Shipping Container Lines

DSIC Dalian Shipbuilding Industry Co., Ltd.

GL +100 A5 Container Ship IW NAV-O RSD ERS BWM(D2) DG + MC AUT EP-D

MSC AZOV 9,403 Ample Pointer Ltd/Costamare Shipping

Shanghai Jiangnan Changxing HI

GL +100 A5 Container Ship IW NAV-O RSD ERS BWM(D2) DG +MC AUT EP-D

CHARLOTTE SCHULTE

5,400 Bendemeer Park Shipping Co. Pte. Ltd/B. Schulte Shipmanagement

HHIC Phil Inc. (Hanjin Heavy Industries Corporation Philippines)

DNV +1A1 Container Carrier DG-P E0 CLEAN BWM-T COAT-PSPC(B) BIS TMON NAUTICUS(Newbuilding)

HAMMONIA ISTRIA 4,957 MS HAMMONIA ISTRIA Schiffahrts GmbH & Co. KG/ Hammonia Reederei

Jiangsu New Yangzijiang Shbldg

GL +100 A5 Container Ship IW NAV-O RSD BWM DG +MC AUT CM-PS EP-D

BARTOLOMEU DIAS 4,800 Hamburg Süd/Alianca Navegacao

Shanghai Shipyard Co Ltd GL +100 A5 Container Ship IW NAV-O RSD BWM DG +MC AUT EP-D

BANAK 2,550 Klaveness Container AS/Klaveness Ship Management

Jiangsu Yangzijiang Shipbuilding Co

GL +100 A5 E Container Ship IW NAV-O RSD BWM(D2) SOLAS-II-2,Reg.19 +MC E AUT

NORDLION 1,700 Schifffahrtsgesellschaft MS NORDLION mbH & Co. KG/Reederei Nord

Zhejiang Ouhua Shipbuilding Co

GL +100 A5 Container Ship IW NAV-O ERS BWM(D1) DG +MC AUT CM-PS EP-D

During the past months a number of newly designed container ships

have been delivered to the DNV GL class, please see the list below.

© H

ase

np

usch

© H

ase

np

usch

SAFER, SMARTER, GREENER

www.dnvgl.com

Documents

CONTAINER SHIP UPDATE - DNV GL SE · PDF fileFor 18,000 TEU ships, the material scantlings are between 80 and 95 mm, where the larger thicknesses are for high tensile steel with