Harmonizing SEEMP with effective vessel operation

Gregory J Grigoropoulos1), Dimitris Konstantinos N Theodossiou2)

1)
National Technical Univ. of Athens, Greece, [email protected]
2)
DANAOS Management Consultants SA, Greece, [email protected]

MEPC are the establishment of the Ship Energy Effi-

Abstract ciency Management Plan (SEEMP) and the introduction
of two indices, the Energy Efficiency Design Index
Current high prices of fuel necessitate the management (EEDI) and the Energy Efficiency Operation Index
of ship operation for energy efficiency. This is further (EEOI). The former index, EEDI, which correlates the
supported by the recent resolution issued by emissions with the transport capacity and speed, will be
IMO/MEPC which obligates shipping companies to effective on the beginning of 2013 for new buildings.
implement a Ship Energy Efficiency Management Plan The latter index, EEOI, with quite similar concept as the
(SEEMP) and to associate an Energy Efficiency Opera- former, will apply to existing ships as well, taking into
tional Indicator (EEOI) to each ship. This resolution account the operational profile of the vessel. Both indi-
provides a good basis to keep a reliable record of the ces, however, assume a direct relation between the fuel
ship performance with respect to energy efficiency, to consumption and the bulk of emissions.
set goals in terms of improving it and finally to evalu- Shipping companies, in order to conform to the above
ate the effectiveness of their policies. A suitable perfor- IMO Resolutions, have to prepare a plan which keeps
mance indicator is the necessary tool for these tasks. In track of the evolution of energy consumption onboard in
this paper: (a) the available performance indicators are conjunction with its associated operational parameters.
evaluated and a new one is proposed for internal use in On the basis of these data the respective performance of
the company, and (b) a methodology is proposed, based the vessel will be monitored and evaluated. In case
on the new indicator to keep track of the energy effi- discrepancies with similar archived data are encoun-
ciency of the ship to identify the operational parameters tered, the company should be able to identify whether
leading to deviations and, to support the decisions for they are due to a modification of the condition of the
medication policies in line with the company specific vessel or to inconsistency of the reported data. In the
goals for energy efficiency. The root cause analysis for former case, the medication actions to resolve the dis-
poor performance is identified and the impact of each crepancies should focus on the identification of all the
policy available is evaluated in quantitative way. The parameters that may affect the deterioration of energy
impact of trim, ballast, hull fouling, propulsion system efficiency of the ship (hull or propeller fouling, main
performance etc. are analyzed. The applicability of the engine rating etc.). The improvement of the perfor-
proposed method has been already tested during mance and the subsequent reduction of the emissions,
onboard a containership within EU Project ESPRIT are carried out by taking all the necessary medication
(Loukakis et al, 1992). actions to eliminate any stimulants of excessive fuel
consumption. In the latter case, the reported data are
Keywords checked for consistency. The paper will focus on cargo
ships and on the measures that should be taken with
Energy efficiency; environment; index; optimization; respect to the propulsion of the ship, considering that
hydrodynamics; system identification. the maintenance policy affecting the consumption of the
auxiliary engines, which are also taken into account in
EEDI and EEOI indices, is quite different.
1. Introduction The major difference between the EEDI and the EEOI is
that in the latter the operational profile and the actual
Although waterborne is a low polluting (3% of the total cargo carried by the vessel are taken into account. Fur-
pollution) transportation mode, serving about 90% of thermore, the EEDI is based on the results of the sea
the world trade, its sustainability and environmental trials, where the hull is new and both the hull and the
performance is concentrating the interest of both inter- propeller are clear. The authors claim that the consump-
national maritime organizations and national authorities. tion along an operational profile of a ship can mainly be
The International Maritime Organization (IMO) has reduced using a state-of-the-art routing system
formed a special committee for the environmental poli- (Theodossiou and Grigoropoulos, 2005). This compares
cy MEPC to propose measures for reducing emissions the route proposed by the master with optimal route
in the coming decades. The major tools proposed by based on vessel characteristics, weather forecasts, navi-

1
gation restrictions and charter party obligations. Thus,  Target-orientated: The choice of indicators should be
the operational profile will not be treated in this paper. directed by company goals.
A new index is necessary that will trace the effect of the  Balanced: Environmental performance must be re-
quality of the hull and the propeller surface on the con- flected with a balance of both problems and benefits.
sumption, taking into account the loading.
 Continuity: The indicators must have the time series
With the aforementioned comments in mind the authors and time units.
evaluated the available efficiency indicators for cargo
 Frequency: The frequency of measurements must be
ships to discover their capability not only to monitor the
sufficient for action to be taken in time.
current situation of the vessel and its rank with respect
to energy performance, among similar vessels, but also  Comprehensibility: The indicators must be in corre-
to guide the operator in his attempt to improve it and spondence with the needs of information, as well as
operate it more efficiently. Since none of them is found understandable for the user.
suitable for all these tasks, a new index is proposed. Finally, EMSA (2007) identified the following charac-
The new indicator will form the basis for a new meth- teristics to be fulfilled by a proposed European “Clean
odology which keeps track of the energy efficiency of Ship" Label:
the ship, and identifies the operational parameters that 1. Comprehensive: It should not be restricted to certain
cause any apparent deviation. Thus, the proposed meth- vessel types or pollution types.
odology supports the decisions via which the available 2. Complete: The requirements of the label should be
policies to meet the company specific goals in terms of met by both the management of shipping companies
energy efficiency are implemented. Furthermore, the and ships by ship design and technical equipment. It
reliability and the consistency of the daily reports of the also should include on-board social and safety as-
vessel are evaluated and in case of systematic deviations pects.
a root cause analysis is recommended.
3. High public visibility: High public visibility could
enable use of the label in environmental marketing
strategies.
2. Selection of Ship Performance Index (SPI) 4. Flexible: The system should be updated regularly for
adaptation of technological developments and “ideas
2.1 Review of existing Indices & their Characteristics about environmental protection and pollution preven-
tion”.
A variety of tools for communicating environmental and
sustainability performance of products, services, activi- 5. Result-oriented: EMSA highlighted that the possibili-
ties and organizations are available nowadays. Svensson ties for improvement of environmental performance
and Andersson (2011) carried out an exhaustive review are numerous. The proposed labels should thus be
and evaluation of environmental performance indices based on achievable results and not on specific tech-
for shipping. Their evaluation was based on the follow- nologies or measures.
ing characteristics selected among those proposed by 6. Easily verifiable: Compliance of the proposed system
Meadows (1998) for ideal indices: should be verifiable bon y either Port State Control or
classification societies acknowledged by IACS.
• Clear in content: easily understandable
7. Simple: Complicated ranking systems should not be
• Feasible: measurable at a reasonable cost
used. EMSA highlighted that too many and too ambi-
• Sufficient: not too much or too little information tious criteria may lead to high bureaucracy. It was
• Appropriate in scale: “not over- or under-aggregated” proposed that the system should be divided in a set of
mandatory requirements and a set of optional re-
• Democratic: People should be able affect the choice
quirements (with a certain percentage that needs to be
of indicators and get access to results.
fulfilled).
• Hierarchical: The user should quickly get the general
Svensson and Andersson (2011) collected 38 pertinent
message, but also be able to view detailed infor-
initiatives falling into one of the following categories:
mation.
1. IMO Instruments
• Physical: Physical units are preferred before monetary
units. 2. National Instruments and Initiatives
• Leading: Information should be given in time for 3. Classification Societies
action. 4. Ports and Port Associations
• Tentative: Should be up for discussion and have the 5. Cargo Owners, NGOs and Shipping Associations
ability to change. 6. Proposals and On-going Initiatives
On the other hand, the standard ISO 14031 contains the 7. On-board tools, Software and Calculators.
following principles for the derivation of environmental Furthermore, the aforementioned authors proceeded to
indicators (Jasch, 2000): the creation of a short list of ten of them and then se-
 Comparability: The used indicators must be compara- lected the following three for further elaboration:
ble. 1. Energy Efficiency Operational Indicator (EEOI)

2
2. Clean Cargo Working Group (CCWG) Performance • 100% deadweight, for bulk carriers, tankers, gas tank-
Metrics Tool ers, Ro-Ro cargo and general cargo ships
3. Clean Shipping Index (CSI) • 70% deadweight, for containerships
The above indices possessed better qualities in response • 100% gross tonnage for passenger and Ro-Ro passen-
to the following four questions: ger ships
A. What is the intended use and who are the intended Of course the reliability of the aforementioned quanti-
users? ties, when used for ranking a ship among similar ones,
B. What is the general basis and scope of the initiative? is quite questionable.
C. Which of the voluntary initiatives are based on an Finally, a table for CO2 Conversion Factors CF is pro-
indexing system? vided. RightShip categorizes ships in 12 types and de-
D. By evaluation of a selection of indices, how do they fines their GHG Emission Rating in an A-G scale by
correspond with principal aspects and criteria found comparing the index of an existing ship with the dis-
in literature? crepancy from the mean value of similar type and/or
size. The discrepancy in expressed in terms of standard
The final elaboration on the three indicators led to the deviation of the distribution of the ships of the same
conclusion that the CCWG tools is convened by the type and/or size over the logarithm of EVDI (ln EVDI).
Business for Social Responsibility (BSR), a global net-
work aimed at developing sustainable business strate- 2.3 Proposal of a New Energy Efficiency Index
gies and apply to ocean container ships only, and only
the member carrier fleet of the network. In addition, the In the previous section we investigated various available
Clean Shipping Index (CSI) is a regional initiative on operation and design indices and their characteristics.
the Swedish west coast, applies to ten types of existing Furthermore, in the introduction of the paper we
ships, it uses a quite complex scoring system with max- claimed that an operational index, without the opera-
imum 150 points, which takes into account the perfor- tional profile of the vessel, would be more suitable to be
mance at five environmental areas. Geographically, the included in SEEMP, since the operational performance
Performance Metrics Tool covers a total of 25 global may be treated by routing software, as it is described by
trade lanes. However access to the database is limited to Theodossiou and Grigoropoulos (2005).
membership. In this section we will discuss the introduction of an
Thus, according to Svensson and Andersson (2011), Energy Efficiency Indicator, which will support a ship-
EEOI is the most promising tool for monitoring opera- ping company in carrying out efficient management of a
tional efficiency. It is applicable to all ships engaged in ship or a fleet of similar vessel. The proposed index can
transport work, i.e. carrying cargo or transporting pas- be used by the company in improving the overall opera-
sengers. The guidelines of the system emphasize the tion of the vessel, independently from the operational
importance of a uniform data collection and that suffi- profile of the vessel. The new index is based on the
cient information is collected to enable a realistic as- EEDI, properly modified to encompass quantities that
sessment. The EEOI enables both assessment of current can be available at the technical department of a ship-
performance and performance trends. A shipping com- ping company. These quantities can provide useful
pany could use the EEOI data for establishing a baseline guidance to the technical staff of the company to take
and to set performance targets thereafter. This index can effective medication actions to improve the energy
be used as communication of the performance in the efficiency of a ship or a fleet.
form of a certificate, a label, an award or a score in a In the aforementioned context, both fi and fw coeffi-
rating system. Furthermore, a variety of systems has cients, responsible for ice operation and service in ad-
evolved that uses the EEOI as a basis. EEOI could be verse weather and sea conditions (wind, waves), which
used as self-assessment at ship owner level without a are included in EEDI, should be eliminated from the
coordinating organization involved with the index sys- proposed index, since fi is a characteristic of specific
tem. type of ships, while fw can only be affected by the de-
On the other hand, RightShip (2011), an Australian signer during the preliminary design phase by modify-
company extended EEDI, which is applicable only to ing main particulars of the vessel or hull form parame-
new buildings laying keels from beginning of 2013, to a ters (Grigoropoulos, 2004) in order to improve the
new index denoted as Existing Vessel Design Index seakeeping qualities of the ship.
EVDI. They based their index on data retrieved from Furthermore, the new index should be referred to both
their own Ship Vetting Information System (SVIS), IHS existing ships and new-buildings. Thus, in case the
Fairplay database, Classification societies, Owner’s data company has no data for the SFC of either the main
and Ship-sourced data. The information used in EEDI engine or the auxiliary engines, tables similar to those
and missing for existing ships, i.e. the reference speed used by RightShip (2011) should be used. However, the
Vref at 75% of MCRME, the specific fuel consumptions effect of speed on the power is quite strong. Following
for the main engine SFCME and the auxiliary engines the Admiralty coefficient as well as the experience on
SFCAE, are replaced by the design speed VS, SFCME and ship hydrodynamics of the first of the authors (see also
SFCAE provided by tables according to the manufacture Newman, 1977) it would be reasonable to consider that
date for the former. The capacity is replaced by: the required SHP of the main engine SHPME, which
forms the dominant part on the numerator of the ratio

3
defining EEDI, is analogous to VS3. Thus, we could for each speed, as the proposed CWEEI, since usually
replace MCRME by the product of CME * VS3, where CME half of the service time is spent in full load and the rest
may be defined by the following formula: in ballast condition. Any other normalized weighted
CME = MCRME / Vref 3
(1) sum may be accepted for special ships, like the contain-
erships. Since the proposed index is mainly intended for
Of course, in case there are data from sea trials, the
use within the company, the technical staff will decide
pertinent information for the speeds during the sea trials
for the most suitable combination to be used. Thus, it is
could replace the cubic law. Both the conversion factor
proposed that:
of fuel mass to CO2 mass coefficient CFME and the spe-
CWEEI FL (VS ) + CWEEI BL (VS )
cific fuel consumption SFCME should be treated in the CWEEI (VS ) = (4)
2
same way RightShip (2011) does. On the other hand the
presence of VS in the denominator of the same ratio is In relation (4):
reasonable to take into account that faster transportation  nME 
rates are awarded lower energy efficiency (EE) indices.  ∑ PMEX (i ) (VS )CFME (i ) SFCME (i ) (VS ) 
Also the product of the capacity and the= speed is a well CWEEI X (VS )  1 +
established performance reference. In both of the above Capacity X VS

cases the proposed EE index is replaced by a speed ( PAE CFAE SFC AE )

+ (5)
dependent EE curve. This curve could be also approxi- Capacity X VS
mated by a polynomial, using best fitting methods, for
In relation (5) the index X refers to either the full load
later use.
(FL) or the ballast (BL) condition.
Another component not accounted for in a design index,
The SFC and the CF for the main engines (index ME)
like EEDI, but useful for monitoring the evolution of the
and the auxiliary engines (index AE) could be drawn
performance of a vessel, is the displacement. Usually,
from tables (e.g. RightShip, 2011) in case no specific
merchant ships sail at two loading condition, the fully
data are available. The same reference is recommended
laden and the ballast one. An additional condition may
for the definition of the “Capacity” in the denominator.
be considered for containerships at the 70% deadweight.
When we are seeking a unique index, VS may well be
Full load and ballast conditions can also be defined for
the Vref as defined in the EDDI by IMO/MEPC. The
passenger ships and RO-RO passenger ships. The for-
formula for CWEEIX(VS) may be appropriately modified
mer will correspond to the maximum number of passen-
in case the auxiliary power is provided by the main
gers and heaviest combination of cars and trucks. As-
engine via a Power-Take-Off (PTO) shaft alternator or a
suming that data for two limiting loading conditions, at
PTO/PTI (Power-Take-In) system.
a heavy and a light displacement, are available, results
for the power at any intermediate loading condition PPL The proposed index can only be used externally in order
may be extrapolated to the full load condition for the to compare the energy efficiency of its ships with other
same speed PFL using the Admiralty relation i.e.: ships of the same or other company, if data derived
2/3 using the aforementioned methodology is available.
PFL = PPL (DFL/DPL) (2)
Furthermore, the new index provides full flexibility, in
In case, information for more than one loading condi- case the operator decides or the charterer request opera-
tions are available they should be used instead of the tion at a lower than the design speed for more economic
Admiralty relation in priority. In that case the interpola- operation. The authors decided to maintain in the pro-
tion for any intermediate displacement should be carried posed index the penalty for high speed imposed by
out by applying the following equation which is a gen- IMO/MEPC on both EEDI and EEOI indices. However,
eralization of Admiralty equation: if a more flexible index is adopted by IMO/MEPC, the
PFL = PPL (DFL/DPL)n (3) installation of highly de-rated main engines onboard
The exponent n in (3) is defined on the basis of the will be avoided, and the fleet will have the capacity to
respective values of the two curves at the same speed. In operate at higher speeds, for example in case of emer-
the aforementioned correction the trim has been includ- gency.
ed implicitly in the correction. Usually there are not Although in this section of the paper a single energy
available results for the same loading condition and efficiency index is proposed, it is recommended that
different displacements. So we have to handle these two both energy efficiency curves vs. speed should be filed
parameters together, assuming that intermediate dis- up in the SEEMP Manual activated in the company.
placements are bound with intermediate trims. This is This information provides valuable reference to the
sufficient for the level of detailed analysis applicable to methodology to be described in the next section, aiming
an index. However, attention should be kept to the fact at identifying the operational parameters causing devia-
that only conditions where any bulb fitted on the bow of tions from the reference operating conditions and at
the vessel is submerged during the operation of the supporting the decisions for medication policies in line
vessel. with the company specific goals for energy efficiency.
Now we have concluded with two Calm Water Energy
Efficiency Indicator (CWEEI) curves, the CWEEIFL(VS)
and the CWEEIBL(VS). They can provide an index for 3. Updating the Calm Water Performance
each speed under consideration. The authors propose to
take into account the mean value of these two indices The analysis up to this point, concluded to two speed-

4
dependent energy efficiency curves for the ballast and amount) is balanced by the thrust T, i.e. the force pro-
the full load condition of the ship in calm water, as well duced by the rotating propeller(s). However, the thrust
as to a rational single indicator for the average energy is a larger force than the resistance to account for the
efficiency of a ship in calm water. losses due to the interaction of the propulsion system
However, these curves refer to a vessel with clean hull with the hull, expressed via the propulsive performance
and propeller upon delivery or immediately after a dry- coefficients of the hull. The relation between resistance
docking. If the data in (5) are based on the sea trials of and thrust is:
the vessel during the delivery, they refer to a new build- R
T= (10)
ing. 1− t
With the time passing (ageing of the ship) the quality of where:
the hull surface is deteriorated, while fouling further
increases the hull and propeller roughness and, there- t = thrust deduction coefficient
fore, the power requirements. Furthermore, the trim- On the other hand, this thrust depends on the open water
ming angle plays sometimes a beneficial, some other characteristics of the propeller(s), which should be
times a tortuous role on the power requirements. specified by model tests in the laboratory. To be more
In the following paragraphs a methodology will be de- specific the following quantities are defined:
scribed to update the information for some critical pa- The thrust:
rameters involved in the estimation of the power re- R
quired for a ship to attain a specified speed VS. On the =
T = KT ( ρ N 2 D 4 ) (11)
1− t
basis of the modification of these parameters a fault tree
analysis will be carried out to identify any problem where:
apparent that caused increased power requirements and N = propeller revolutions
to remedy it, as it will be described in the next section.
D = propeller diameter
Using basic ship hydrodynamic theory and avoiding an
in depth analysis, in order to tow a vessel at a forward KT = thrust coefficient
force should be exerted on it. This force in case the ship The advance coefficient J:
sails in smooth sea is denoted as calm water resistance.
The calm water resistance consists of the resistance due =
J VS (1 − w) / (ND) (12)
to the waves generated by the sailing ship RW, plus the where:
viscous resistance (RF+RVP). The latter is mainly due to
the friction of the wetted surface of the ship with the w = wake fraction
water RF which is slightly (about 20%-30%) increased In order to sail at a specified speed VS, the power re-
to take into account the hull form RVP. To be more spe- quired by the main engine(s) is derived using the torque
cific, calm water resistance is decomposed as follows: coefficient:
RT = RW + RF + RVP (6) K Q ( J ) = SHP ηR ηS / (2πρ N 3 D 5 ) (13)
where
where:
RW = 1 ρ WS VS2 CW (7) SHP = shaft horsepower
2
= RF + RVP 1 ρ WS VS2 CF (1 + k ) (8) ηR = relative rotative efficiency coefficient
2
ηS = shaft efficiency coefficient
In the above relations (7) and (8) the following quanti-
KQ = torque coefficient
ties are used:
In order to simplify the calculations in the case of fixed
ρ = water density
pitch propellers (FPP), we assume that the non-
WS = wetted surface of the ship dimensional coefficients KT and KQ may be expressed as
CW = a coefficient depended on the hull form a third-order polynomial of J, where the coefficients KTi
and on the speed and KQi are specified by curve fitting on the respective
CF = a coefficient depended on the hull roughness experimental curves:
and on the speed KT ( J ) =KT 0 + KT 1 J + KT 2 J 2 + KT 3 J 3 (14)
k = a coefficient depended on the hull form
K Q ( J ) =K Q 0 + K Q1 J + K Q 2 J 2 + K Q 3 J 3 (15)
The wetted surface of a ship may be expressed in the
form of (si are coefficients to be defined): More complicated formula fitting curves can be devised
for Controllable Pitch Propellers (CPP).
WS =s0 + s1TM + s2TM2 + s3 tr (9)
In a similar manner the thrust deduction factor t is ex-
where: pressed via a second-order polynomial of TM and VS. (ti
are coefficients to be specified):
TM = midship draught
tr = trim, i.e. the difference between the draughts t=t0 + t1TM + t2TM2 + t3VS + t4VS2 + t5TM VS (16)
at the bow (FP) and the stern (AP) In order to update the parameters of the propulsive per-
In merchant ships the resistance, which is a virtual formance of a ship, and hence its energy efficiency

5
footprint, some full scale experiments in calm water 3.1 Updating the Propulsive Performance at an Avail-
have to be carried out. During each of these tests the able Loading Condition
captain operates the ship at a constant speed for about
We further distinguish two cases, depending on the
20 minutes, keeping track (recording) of a set of indica-
availability of not of a thrust meter on board:
tions that are currently available onboard of most mod-
ern ships. Using this information, System Identification If the ship is instrumented with a thrust meter, two full
(SI) techniques are implemented to derive the current scale trials are adequate. In this case shaft horsepower
values of the critical parameters that describe the pro- of the engine, propeller revolutions, ship speed, propel-
pulsive performance of the ship. ler thrust as well as wind speed and direction are meas-
ured during the full scale experiments. The values of RT,
Thus, the updating and the variation of the following
w, t, ηR are accessible in the SEEMP, derived either
parameters: KT, KQ, w, t and, CF is the prime outcome of
from model tests or the sea trials. These values are to be
the proposed methodology. Afterwards, taking into
updated together with the propeller characteristics ΚT,
account the relations that constitute the hydrodynamic
ΚQ. The values of relative rotative efficiency ηR and
problem of ship propulsion, e.g. the co-operation of the
shafting efficiency ηS are assumed to remain constant.
propeller with the main engine to propel the ship at a
Then, the following system is to be solved:
given speed, the energy efficiency of the vessel at the
specified speed is derived. The number of runs that are K Q ( J1 ) + δ K Q =
SHP1 ηR ηS / (2πρ N13 D 5 )
necessary for each case depends on the number of un-
known parameters. The more unknown parameters we KQ ( J 2 ) + δ KQ =
SHP2 ηR ηS / (2πρ N 23 D 5 )
have to estimate the more runs have to be carried to K T ( J1 ) + δ K T =
T1 / (ρ N12 D 4 )
form a system of equations that could be solved leading
to a unique solution. KT ( J 2 ) + δ KT =
T2 / (ρ N 22 D 4 )
Furthermore, since it is generally easy to find calm J1= VS 1 (1 − w1 − δ w) / (N1 D) (22)
water conditions, but usually some wind is apparent, we
J 2= VS 2 (1 − w2 − δ w) / (N 2 D)
have to compensate the results for this wind. To do so,
the ship operator has to record both the speed and the RT 1 + X W 1 + 1 ρ WS VS21 δ CF = T1 (1 − t1 − δ t )
direction of the apparent wind, which however, is quite 2
trivial. Then, the force due to the wind XW, which should RT 2 + X W 2 + ρ WS VS22 δ CF = T2 (1 − t2 − δ t )
1
2
be taken into account in the estimation of RT, is given in
the following expression: Subscripts "1'' and ''2'' refer to the first and the second
experiment, respectively. Solving the system, the values
X W = 1 ρ C X L2OA VR2 (17)
2 of J1, J2, δw, δt, δCF, δΚT, δKQ are obtained. For each
where the wind coefficient CX depends on the angle of ship speed VSj the new values of RTj, wj, tj are:
wind incidence γ, the ship frontal area (transverse area wj = wj + δw (23)
of the ship above the waterline, being a function of TM) tj =tj + δt (24)
AT and the length overall LOA, and VR is the apparent Due to the small variation of advance coefficient J, the
speed on the sailing ship, i.e. additive quantities δKT and δKQ which adjust the propel-
C X = C X (γ , AT / L2OA ) (18) ler polynomials, are assumed to be constant for the
entire range of ship speeds. Thus the new expressions
AT = a0+a1TM (19)
for the ΚT, ΚQ polynomials are:
This coefficient has been derived experimentally for the
ΚT(J)= ΚT(J)+δΚT (25)
most common ship types by various authors. A rela-
tively recent reference is provided by Blendermann ΚQ(J)=ΚQ(J)+δΚQ (26)
(1994). In case the ship is not equipped with a thrust meter
On the basis of the equations (10), (11), (12), (13) and (more common situation), the shaft horsepower of the
the following two equations expressing the increase of engine, propeller revolutions, ship speed, as well as
calm water resistance due to fouling and the effect of wind speed and direction are measured during the full
wind: scale experiments. The initial values of RT, w, t, ηR are
available in the SEEMP Manual. The values of RT, w, t
δ RT = 1 2 ρ WS VS2 δ CF (20) are to be updated as well as the propeller characteristics
KT, ΚQ. The values of relative rotative efficiency ηR and
RT + X W + 1 ρ WS VS2 δ CF = (T + δ T )(1 − t − δ t ) (21)
2 shafting efficiency ηS are assumed to remain constant.
we can estimate the variation of the coefficients δKQ, Three experiments are enough for the calculation of the
δKT, δw δt and δCF. updated values of RT, w, t. leading to the following sys-
tem:
The proposed full scale trials in calm water can be per-
formed either at one of the existing loading conditions,
available from the sea trials, for which the CWEEI(VS)
curve is defined, as stated in section 2, or at an interme-
diate condition. Furthermore, these trials should be
carried out after each dry-docking to update the per-
formance of the vessel.

6
K Q ( J1 ) + δ K Q =
SHP1 ηR ηS / (2πρ N13 D 5 ) The problem is set by the following equations:

KQ ( J 2 ) + δ KQ =
SHP2 ηR ηS / (2πρ N D ) 3
2
5
K Q ( J j ) = SHPj ηS / (2πρ N 3j D 5 )
KQ ( J 3 ) + δ KQ =
SHP3 ηR ηS / (2πρ N 33 D 5 ) ( RTj + X Wj ) / ( ρ N 2j D 4 (1 − t j ))
KT ( J j ) = (29)
J1= VS 1 (1 − w1 − δ w) / (N1 D) =
J j VSj (1 − w j ) / (N j D)
J 2= VS 2 (1 − w2 − δ w) / (N 2 D) Solving the system, the values of Jj, wj, RTj, correspond-
J 3= VS 3 (1 − w3 − δ w) / (N 3 D) ing to the ship speed VSj, are obtained.
KT ( J1 ) + δ KT = ( RT 1 + X W 1 + 1 ρ WS VS21 δ CF ) / (27)
2
/ ( ρ N1 D (1 − t1 − δ t ))
2 4
4. Fault Tree Analysis
KT ( J 2 ) + δ KT = ( RT 2 + X W 2 + 1 ρ WS VS22 δ CF ) /
2
On the basis of the results derived in the previous sec-
/( ρ N 2 D (1 − t2 − δ t ))
2 4
tion 3 the operator of a ship can:
KT ( J 3 ) + δ KT = ( RT 3 + X W 3 + 1 ρ WS VS23 δ CF ) / 1. Establish new set of propulsive performance data for
2
/ ( ρ N 3 D (1 − t3 − δ t ))
2 4 a ship assuming that the open water propeller char-
acteristics are available.
Subscripts “1”, “2” and “3” refer to the first, the second 2. In case the results of the sea tests after dry-docking
and the third experiment, respectively. deviate from the available original data systematical-
Solving the system, the values of J1, J2 and J3 and the ly, new KT and KQ curves may be established by car-
variation of the coefficients δw, δt, δCF, δKT, δΚQ are rying out four full scale trials and deriving new coef-
obtained. ficients for the KT(J) and KQ(J) polynomials.
3. Update the propulsive performance after a dry-
3.2 Propulsive Performance Evaluation at a New docking with clean hull and propeller surface. This
Loading Condition information is changing slightly with the ageing of
In this case the propeller open water characteristics the ship. However, this data will be later as reference
ΚT(J), KQ(J) are known. Since model tests give reliable to estimate the extent of fouling and its influence on
results for the thrust deduction fraction (the scale effects performance.
are assumed to be small), the thrust deduction fraction, 4. On regular intervals between successive dry-
expressed as a function of ship speed and draught, is dockings, e.g. every semester the company can up-
assumed to be a known function of ship speed at the date ship performance. In this way they will have
examined loading condition. better perception of the deterioration of the energy
efficiency of the ship and plan for an optimal strate-
However in this case if the ship is instrumented with a
gy of dry-docking and for maintenance, in general.
thrust meter, the shaft horsepower of the engine, propel-
ler revolutions, ship speed, propeller thrust as well as In the sequel some rules of thumb and recommendations
wind speed and direction are measured during full scale are provided regarding the variation of the parameters
experiments. The expression of the propeller character- involved in the propulsive efficiency of a vessel.
istics ΚT(J), KQ(J) and thrust deduction fraction t(VS) are
known. The value of shafting efficiency ηS is assumed 4.1 Specific Fuel Consumption
to be known. For each ship speed VSj one experiment is The specific fuel consumption equation, which is usu-
enough for the calculation of the values of RTj, wj, ηRj. ally supplied by the manufacturer of the Main Engine, is
K Q ( J j ) = SHPj ηRj ηS / (2πρ N 3j D 5 ) derived from shop tests at various loads and shaft
speeds. By means of a regression analysis, the specific
KT ( J j ) = T j / (ρ N 2j D 4 )
(28) fuel consumption can be expressed as:
=
J j VSj (1 − w j ) / (N j D)
SFC= SFC (SHP, N) (30)
RTj + X Wj = T j (1 − t j )
Use of this expression implies that the main engine is in
Solving the system, the values of Jj, wj, RTj, ηRj, corre- good working condition.
sponding to the ship speed VSj, are obtained.
However, in case this information is not available, the
In case the ship is not equipped with a thrust meter, the operator can refuge to the tables of average consump-
shaft horsepower of the engine, propeller revolutions tion based on the size and the date of manufacture off
and ship speed as well as wind speed and direction are the main engines as stated in a previous chapter.
measured during full scale experiments. The functions
of propeller characteristics KT(J), ΚQ(J) and thrust de- 4.2 Influence of Thrust Deduction Fraction on the
duction fraction t(VS) are known. Results
The value of shafting efficiency ηS is assumed to be Considering that the value of thrust deduction fraction is
known. The value of relative rotative efficiency is set not accurately known, the effect on the values of fouling
ηR=1.00. For each ship speed VSj one experiment is coefficients will be investigated using the propeller
enough for the calculation of the values of RTj, wj.

7
polynomials ΚT(J), ΚQ(J) and an erroneous value for t. These results may also be extended to ships fitted with
Assuming a value of t=0.25 (constant), an test case is controllable pitch propellers CPP, in case the reference
run for the same loading condition which was used with tests and the trials for the evaluation and the update of
the correct t value. Assuming that a thrust meter is in- the coefficients are carried out with the same propeller
stalled onboard, we enter the same measured values as pitch. However, the systems of equations to be solved
in the case of correct t value. are different in case pitch is varying during the tests.
The resulting values of RT, w, ηR based on the aforemen-
tioned erroneous t value are used along with the rest
coefficients to perform two new full scale measure-
5. Conclusions
ments, which provide the same results as with the cor-
rect t. In order to update, revolutions, propeller thrust,
In this paper a new Energy Efficiency Indicator applica-
brake power and wind data are to be measured at each
ble in Calm Water is proposed which depends on the
experiment.
ship speed VS. The new index is quite robust and flexi-
The values of δw, δKT and δKQ have not changed (pro- ble and allows for establishing the starting point for
peller fouling predictions are independent of thrust examination of the evolution of the efficiency of the
deduction values) and for an overestimation about 40% vessel with the time passing and in-between successive
of t, the hull fouling is underestimated by about 15%. dry-dockings. As stated in the paper, the proposed index
Needless to say that the predictions of shaft horsepower can only be used externally in order to compare the
are not affected by the values of thrust deduction frac- energy efficiency of its ships with other ships of the
tion. same or other company, if data derived using the afore-
mentioned methodology is available.
4.3 Effect of the Variation of the Propulsive Perfor-
mance Parameters Furthermore, a new methodology is presented, based on
the theory of ship hydrodynamics, which exploits the
To gain an insight in the effect of varying values of the interaction of the ship and the propeller to assess the
propeller-ship interaction factors w and t and the effect variation of the propulsive performance parameters and
of propeller and hull fouling on the propulsive perfor- eventually, the effect of fouling on it.
mance of the ship (or more accurately on the mathemat-
This methodology provides better insight into the evolu-
ical model describing this performance), a thorough
tion of ship consumption with ageing as well as between
examination was performed, using data for a container-
dry-dockings.
ship with fixed-pitch propeller FPP.
The proposed methodology can easily be incorporated
For this ship and the prescribed propulsion plant a sys- in a computer code using commercially available com-
tematic variation of δw, δt δKT, δKQ and δCF has been puter routines. It is based on data recorded by the crew
performed and the influence on the propulsion perfor- and not on information provided in the manuals. The
mance was plotted around an initial working point (i.e. implementation of the methodology doesn’t affect the
clean hull, clean propeller). charter plan and provides a realistic presentation of the
Each diagram shows the combined influence on perfor- actual ship condition.
mance for two parameters variation, i.e. (w, t), (ΚT, ΚQ), In case sufficient data for similar loading conditions
(CF, KQ), (w, KQ), (w, CF). Thus, five diagrams have (similar displacement) with varying trim are available, a
been constructed for the initial ship speed VS, the nomi- similar methodology incorporating SI techniques could
nal propeller revolutions N and the maximum continu- be applied to trace the effect of trim on the propulsive
ous rating on the shaft SHP. performance of the ship for the specific displacement.
On the basis of these diagrams the following trends The accuracy of the proposed methodology depends
were revealed: only on the accuracy of the indicators used for recording
the information. However, with the advance of technol-
• Increase of δΚQ (propeller fouling) increases the
ogy the instrumentation onboard is very accurate. Even,
shaft horsepower requirements but the relationship
the accuracy of the estimation of the ship speed via the
for speed vs. RPM remains almost the same.
GPS is quite satisfactory if the recorded time is at least
• Increase of effective wake fraction decreases shaft 15 minutes.
horsepower requirements and generally improves
propulsive efficiency.
• Increase of thrust deduction fraction and hull fric- 6. Acknowledgements
tional resistance coefficient CF increases shaft
horsepower requirements or decreases attainable Preliminary investigations into the propulsive perfor-
speed. mance of cargo ships were carried out on the way to
develop an Expert Voyage Pilot (EVP) during ESPRIT
These trends are more general and apply also in the case
Projects P-1074 and P-2163, partly funded by E.C. All
of a tanker or a bulk carrier, as it has been found by
pertinent research was performed in the Laboratory for
additional calculations for these types of ships. Thus, it
Ship and Marine Hydrodynamics of National Technical
is expected to be valid also for rest of the cargo ships
University of Athens.
with FPP.

8
 Meadows, D. (1998). “Indicators and Information Sys-
tems for Sustainable Development”, Hartland Four
References Corners VT, the Sustainability Institute.
Newman, JN. (1977). “Marine Hydrodynamics”, MIT
Blendermann, W. (1994). “Parameter Identification of Press
Wind Loads on Ships’, Journal of Wind Industrial RightShip, (2011). “GHG Emissions Rating, Calculat-
Aerodynamics, Vol. 51, pp. 339-351. ing and Comparing CO2 Emissions from the Global
Grigoropoulos, GJ. (2004). “Hull Form Optimization Maritime Fleet, RightShip Pty Ltd, Australia, Aug.
for Hydrodynamic Performance.” Marine Technol- http://site.rightship.com/environmental-rating.aspx
ogy, Vol. 41, No. 4, October Theodossiou, DKN and Grigoropoulos, GJ (2005).
Loukakis, TA, Bardis, L, Gelegenis, G., Katsaounis, G. “Optimal Routing Decision Support System”, Proc.
and Grigoropoulos, GJ (1992). “The Expert Voyage 1st Intl. Symp. on Ship Operation, Management and
Pilot. Response Function Updating”, ESPRIT Pro- Economics, SOME 2005, Athens.
ject 2163, Task 1.2.5, Technical Note 2163-TN-
MTUA-003, Athens.