7/23/2019 Reference Values for Ship Pollution Dnv http://slidepdf.com/reader/full/reference-values-for-ship-pollution-dnv 1/59 TECHNICAL R EPORTDET NORSKE VERITAS D ET N ORSKE V ERITAS , THE R ESEARCH COUNCIL OF NORWAY R EFERENCE VALUES FOR SHIP POLLUTION R EPORT NO. 99-2034 R EVISION NO. 0
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
This report presents models to calculate pollution emissions from ships during voyage (not includingaccidents or port operations). The models presented enable quantification of exhaust gas emissionsand toxic substances from marine antifouling systems. The models are based on a breakdown of the
fleet into segments (ship types and size categories.)
Exhaust gas emissions may be quantified in terms of carbon monoxide (CO), carbon dioxide (CO2),
particulate matter (PM), sulphur dioxide (SO2) and nitrogen oxide (N2O). Toxic compounds fromantifouling paint may be quantified in terms of tributyltin (TBT).
The models apply statistical relations for the installed engine power and wetted surface area asfunction of ship size for defined categories of ship types. The distribution of engine types is taken into
account.
The models are applied to quantify the total emissions from the commercial world fleet in 1996. Thecalculated pollution levels are compared with independent studies and methods. This validationindicate that the models are able to quantify the representative level of pollution from ships.
It should be noted that the models are applicable for the calculation of emissions from a fleet or a
segment of a fleet (numerous ships). The uncertainty increases significantly when applied to a singleshi .
Report No.: Subject Group:
99-2034 M21 Indexing terms
Report title:
Reference values for ship pollution Ship
Pollution
Model
GlobalWork carried out by:
Øyvind Endresen and Eirik Sørgård No distribution without permission from theClient or responsible organisational unit
Work verified by:
Tommy Johnsen
Limited distribution within
Det Norske VeritasDate of this revision: Rev. No.: Number of pages:
Page 1Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
1 INTRODUCTION
The ship transportation industry may be faced with more regulations to limit pollution. IMOMARPOL Annex VI address NOx and SOx emission control measures. IMO is presently
discussing Greenhouse Gas (GHG) emissions in the context of the Kyoto agreement. IMO has proposed a ban of tributyltin (TBT) in anti fouling paint.
In order to assess costs and benefits for the society by regulating ship transportation, arelationship between the fleet characteristics and the global or regional emissions may be useful.This reports documents the development of models that enable quantification of the emission
levels for the entire fleet or for selected elements of the fleet (e.g. ship type and size category).
The models may also be used to establish reference values for ship pollution that serve as typicalfigures for comparison.
The model is based on emission factors documented in /1/. Statistical relations to enableapplication of these factors to calculate emission levels for fleet segments are developed and
documented in this report. The models are applied to calculate the exhaust gas emissions andTBT input to the marine environment from the world fleet. A comparison of calculated emission
levels with available global statistics is made.
Sections 2-4 of this report presents the antifouling emission model. Sections 5-7 presents theexhaust gas emission model. The principle difference between the models is that the anti-fouling
model is related to the wetted surface area of a ship while the exhaust gas emission model is
related to installed engine power. The statistical relationships between ship type/size and wettedarea are given in Appendix A, and the statistical relationships between ship type/size andinstalled engine power are given in Appendix B.
2 THE ANTI-FOULING EMISSION MODEL
Ship hull fouling increases fuel consumption. Ship owners coat the wetted surface with paint to
prevent fouling. Present antifouling systems mainly contain toxic compounds like copper andTBT (tributyltin) and cotoxicants like zinc and organic compounds like triazines. 80 - 90 % of
these biocides will leach within 3 to 5 years into the sea. Self-polishing coatings (SPC) need flowvelocity to active the anti-fouling effect. Conventional coatings emit biocides constantly /2/.
TBT is an organic compound containing the metal tin. It is used as a broad-spectrum killer of
algae, fungi, insects, and mites. Since the 1960s, TBT has mainly been used as a marineantifouling agent. It leaches into the water from surface coatings on boats, aquaculture pens,moorings, and industrial cooling pipes. Other sources are yards, marinas, and municipal
wastewater and sewage sludge. It is estimated that TBT-based antifouling paint is used on 70-90 percent of the world's fleet /3/, /24/.
This report presents an antifouling emission model to estimate the yearly emitted amount of toxic compounds from the fleet. As an example of use, tributyltin (TBT) emissions from theworld fleet in 1996 is calculated (see Section 4).
Page 2Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
The antifouling emission model is based on (Figure 1):
• A breakdown of the fleet according to ship type and ship size /4/• Relations between wetted surface and DWT for each vessels category; Table 4
• Distribution of antifouling types (TBT, Copper and others) /2/, /3/
• Leaching rate /2/, /10/
A breakdown of the fleet according to ship type (i) and size (k ) and applied antifouling system(s) is made on three levels (Figure 1). Level three consists of the fraction of vessels withantifouling system s for a ship type i and of size x (k). The ship type breakdown according to /4/
is shown in Table 1.
Level 0:
Level 1:
Level 2:
Level 3:
Figure 1 System breakdown of the fleet
The total wetted surface area (WS) for ships of type i and size k and with antifouling systems of type s is calculated from:
• An established relation between the wetted surface area for a given ship type and the size of
the ship (Table 4).
• The number of ships in each ship type category i of size k /4/.
• The fraction of vessels with an antifouling system of type s ( ß?) /2/, /3/.
Page 3Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
Table 1 Ship type breakdown according to /4/.
Abbre-viation
Vessel types Number of size categories
Unit for size Reference
LGT Liquid gas tanker 14 DWT Table 7A in /4/CT Chemical tanker 14 DWT Table 8 in /4/
OT Oil tanker 21 DWT Table 9 in /4/
B Bulk 1) 21 DWT Tables 10 and 11 in /4/
GC General cargo2) 11 DWT Table 12 in /4/
RO RO-RO cargo3) 11 DWT Table 13 in /4/
C Container 14 DWT Table 14A in /4/
RC Refrigerated cargo 4 DWT Table 15A in /4/
P Passenger 4 GRT Table 16 in /4/1)
Dry and Dry/Oil2)
Including Passenger/General cargo3)
Including Passenger/RO-RO cargo
The average size within a ship type category i and size category k ( x ik ) is calculated by dividing
the total deadweight (W ik ) with the number of ships ( N ik ) in category ik :
ik
ik
ik N
W
x =
(1)
Equation (1) combined with the relation between wetted surface and ship size in Table 4, givesthe average wetted surface (WS ik ) for a ship type i of size k :
ii b
ik
ik i
b
ik iik N
W a xaWS )(⋅=⋅=
(2)
Where: N ik : the number of ships of type i and size k WS ik : the average wetted surface for a ship of type i of k W ik : the total deadweight for ships category ik
x ik : the average deadweight for a ship of type i and size k ai , bi : constants from the correlation analyses for a ship type category
The emitted amount of toxic compounds from an antifouling system is calculated on level 3 andaggregated up to fleet statistics at level 0, by the emission equations given below. The emissionsfrom ships of type i and size k using anti-fouling type s, M iks, is given by:
Page 4Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
where C s is the leaching rate and βs is the fraction of vessels with anti-fouling system of type s.The emissions from all vessels of type i and size k , M ik , is given by:
∑=
=S
siksik M M
1
(4)
The emissions from vessels of type i using anti-fouling type s, M is, may be calculated by:
∑=
= K
k
iksis M M 1
(5)
The emission from vessels in ship category i, M i, is given by:
∑=
=S
sisi M M
1
(6)
The total emission for the fleet, M , is given by:
∑=
= I
ii M M
1
(7)
3 WETTED SURFACE AREA AS FUNCTION OF SHIP SIZE AND TYPEThis section presents algorithms to calculate the wetted surface area. A correlation analysis is
performed for each ship type category to find statistical relations between wetted surface area
and ship size given by the deadweight tonnage.
The correlation analysis is based on information in /6/ which consists of 50446 vessel. For each
vessel in the database the following data was used:
a) Vessel codes according to “Ships Type Group and Main Type” /7/ b) Ship size given by DWT (deadweight tonnage) or GRT (gross tonnage)
c) Ship beamd) LBP (length between perpendiculars)e) Ship draft
The wetted surface area was calculated by means of the ship beam, draft and length between
perpendiculars. The estimated wetted area was the plotted as function of size (DWT/GRT) for each ship type to identify and establish correlation functions.
Page 5Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
3.1 Algorithms for wetted surface area
For a vessel floating at a given waterline, the total area of its outer surface in contact with thesurrounding water is known as its wetted surface. The following equations may be applied to
estimate the wetted surface area, WS, /2/, /8/:
LC WS ws ⋅∇⋅= (8)
)22( T BT L L BC AC WS bboxb ⋅⋅+⋅⋅+⋅⋅=⋅= (9)
3/11/3 )5.0(3.4 ∇⋅⋅+∇⋅= LWS (10)
Where:
WS : wetted surface area up to any waterline (m2)∇ : volume of displacement at that waterline (m3), i.e. the volume of the underwater portion
of a vessel, up to the waterline at which the vessel is floating (may be estimated byequation (11))
C b : block coefficient
Cws : wetted surface coefficientA box : wetted underwater box area (m2) of a vessel
L : length of a vessel (m)B : breadth of the ship (m)
T : mean molded draft to the prevailing waterline (m)
Other relations are described as well in the literature /9/. The block coefficient, C b, is defined asthe ration of the volume of displacement of the molded form up to any waterline to the volume
of a rectangular prism with length, breadth and depth equal to length, breadth and mean draft of the ship, at that waterline /8/. The relation is written as:
T B LC b ⋅⋅
∇=
(11)
Practice varies regarding L and B. Some authors take L as LBP (length between perpendiculars),some as LWL (length on waterline), and some as an effective length. B may be the taken as the
molded breadth at the design waterline and at amidships, the maximum molded breadth at aselected waterline (not necessarily at amidships), or according to another standard. Most
merchant ships have vertical sides amidships, with upper waterline parallel to the centreline,thereby removing possible ambiguity in B.
Page 6Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
Values of C b at design displacement may vary from about 0.36 for a fine high-speed vessel to
about 0.92 for a slow speed and full Great Lakes Bulk carrier. Typical rang and characteristicvalues for Block coefficient are given in Table 2.
Table 2 Rang and characteristic values for Block coefficient.
Ship type Cb -Rang Cb
Tank/Bulk 0.80-0.85 0.83
Container/Dry cargo 0.55-0.65 0.60
Passenger/RO-RO 0.50-0.60 0.57
Supply/Tug 0.70-0.80 0.75
Fishing craft 0.70-0.80 0.75
The wetted surface coefficient, C ws, is typically in the range between about 2.6- 2.9 according to
plots in /5/. Figure 2 shows wetted surface coefficient against B/T (breadth/draft) for differentships, based on extracted values from figure 38 (page 49) in /5/. The black lines are trend lines
for each ship type based on the extracted values (see Equation (12) and Table 3). In all cases C ws,reaches a minimum in the range of B/T from 2.5 to 3.75. The increase of C ws, towards lower B/Tvalues from the minimum point is assumed to reflect the greater submergence of the stern
overhang which tends to accompany draft increases beyond the design draft.
2.55
2.6
2.65
2.7
2.75
2.8
2.85
2.9
2.95
3
3.05
1 2 3 4 5 6 7 8
Breadth/Draft
C
w s
Tanker
Dry cargo
Passenger
Passenger-cargo
Figure 2 Wetted surface coeffisient as a function of the Bretadth/Draft factor/5/
Page 7Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
The wetted surface coefficient, Cws, may be estimated for different vessels using curve constants
given in Table 3.
d++ b+aC
32
ws
⋅
⋅
⋅=
T
Bc
T
B
T
B (12)
Table 3 Vessel type depended curve constants, equation (12).
Ship type a b c d
Tanker -0.418 0.095 -0.006 3.222
Dry Cargo -0.270 0.053 -0.003 3.045
Passenger -0.266 0.055 -0.003 2.987
Passenger-cargo -0.402 0.082 -0.005 3.283
3.2 Statistical relation between wetted surface and dead-weight
Equations (8)-(10) and the data provided in the database /6/ were used to calculate the wetted
surface area for each vessel type according to the predefined codes (last column, Table 4, /7/)
which are related to ship types (first column of Table 4) as given in Table 1. A correlationanalysis was then performed for each ship category to find statistical relations between wettedsurface and ship size (Table 4). The ship size is given by deadweight tonnage (DWT) for allships but passenger vessels where gross tonnage (GRT) is applied. Appendix A shows plots of
the wetted surface area as a function size for each vessel category.
The statistical results present in Table 4 shows agood correlation between wetted surface area
and size for all vessel categories. Using Equation (10) to estimate the wetted surface area seemsto give the best statistical results.
Tanker and Bulk gives the highest correlation, and RO-RO the lowest. A comparison of the
statistical relations in Table 4 give the following characteristics:
• Equation (9) is not recommended for RO-RO and Passenger vessels as the standard deviation
then is high• Nearly the same relations for Tanker and Bulk and low standard deviation using the relations
• The General Cargo relation is approximately the same as for Tanker and Bulk, but the
Page 8Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
Table 4 Relations for wetted surface (WS) in m2 based on dead-weight tonnes (DWT) and
gross tonnes (GRT).Type WS Algorithm DWT/GRT
(tons)
r2
SWS
(m2)
SWS/WS
(%)
N Code
LGT
LGT
LGT
(8)
(9)
(10)
WS= 12.89⋅DWT0.62
WS = 13.5⋅DWT0.62
WS = 13.08 ⋅DWT0.62
260 - 83228
260 - 83228
260 - 83228
0.98
0.98
0.98
648
743
608
11.4
11.6
11.2
882
882
882
17
17
17
OCT
OCT
OCT
(8)
(9)
(10)
WS = 11.62 ⋅DWT0.62
WS = 11.99 ⋅DWT0.62
WS = 11.30 ⋅DWT0.62
232-564650
232-564650
232-564650
0.99
0.99
0.99
380
387
369
7.8
7.6
7.5
6263
6263
6263
11,12,13
11,12,13
11,12,13
B
BB
(8)
(9)(10)
WS = 12.23 ⋅DWT0.61
WS = 13 ⋅DWT0.61
WS = 12.02 ⋅DWT0.61
562-364767
562-364767562-364767
0.99
0.990.99
388
386371
8.6
8.48.2
8453
84538453
21,22,34,36,39
21,22,34,36,3921,22,34,36,39
GC
GC
GC
(8)
(9)
(10)
WS = 13.92 ⋅DWT0.58
WS = 13.95 ⋅DWT0.57
WS = 14.22 ⋅DWT0.58
41-51880
41-51880
41-51880
0.90
0.89
0.90
378
344
371
27.1
28.4
26.9
6673
6673
6673
41,52
41,52
41,52
C
C
C
(8)
(9)
(10)
WS = 5.48 ⋅DWT0.69
WS = 6.15 ⋅DWT0.67
WS = 5.6 ⋅DWT0.69
1116-104886
1116-104886
1116-104886
0.97
0.96
0.97
595
654
577
10.9
13.6
10.7
2362
2362
2362
48
48
48
RO
RO
RO
(8)
(9)
(10)
WS= 40.17 ⋅DWT0.50
WS= 40.12 ⋅DWT0.47
WS= 40.1 ⋅DWT
0.50
15-42600
15-42600
15-42600
0.82
0.78
0.83
712
883
696
31.1
47.9
30.9
2558
2558
2558
43,53
43,53
43,53
P
P
P
(8)
(9)
(10)
WS= 7.88 ⋅GRT0.63
WS= 10.479 ⋅GRT0.58
WS= 6.28 ⋅GRT0.64
15-16606
15-16606
15-16606
0.96
0.88
0.96
568
738
572
24.8
50.5
27.4
613
613
613
51,59
51,59
51,59
R
R
R
(8)
(9)
(10)
WS= 17.61 ⋅DWT0.58
WS= 20.82 ⋅DWT0.55
WS= 18.17 ⋅DWT0.58
196-19286
196-19286
196-19286
0.91
0.85
0.91
321
351
317
16.4
20.6
16.4
1665
1665
1665
42
42
42
WS- equation number used to estimate wetted surface
N- number of vesselr - the correlation coefficient
Sws= The absolute standard error of the predicted ws-value for each DWT (equation (13))
(Sws/ws)= The relative standard error of the predicted ws-value for each DWT (equation (14))
The following equations are used to calculate of absolute and relative standard error in thestatistical relations given in Table 4 (and later on in Table 11 for engine power):
Page 9Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
2
2
12
)(
)(
2
12
)()()(11
∑∑==
−−
−−
= = N
ib
i
b
ii N
i i
ii y
ax
ax y
N x y
x y y
N y
S (14)
Where:
S y : The absolute standard error of the predicted y-value for each x. The standard error is ameasure of the mean deviation between given value and estimated value by means of the
correlation function.
(S y /y) : The relative standard error of the predicted y-value for each x. The relative standarderror is a measure for the relative deviation between given value and estimated value by
means of the correlation function.
N : number of ships in each category (see Table 4).
yi : wetted surface area (m2) (or engine power (kW), see Section 6) for an individual ship i.
x i : dead-weight in tons (or gross tonnage), for an individual ship i.
a,b : constants from the correlation analyses (Table 4 and Table 11).
4 EXAMPLE OF USE; EMITTED MARINE TBT IN 1996
A review of current antifouling regulations world wide (1996) is given in /11/. In addition,current legislative status regarding antifouling paints are given in /10/ and /12/. Application of
TBT-antifouling in Japan New Zealand is not longer allowed. EU, Canada and USA require arelease rate less than 4 µg/cm2/day and Australia require a release rate less than 5 µg/cm2/day.
IMO recommend a TBT rate below 4 µg/cm2/day /10/. The TBT leaching rate was assumed to be
2 µg/cm2/day in a previous work in the North Sea /2/. A leaching rate of 2 and 4 µg/cm2/day is
applied in the model.
It is believed that TBT-based antifouling paint is used on 70 % of the world's fleet /3/. The paint
manufacturers claim that between 70 and 90 % of antifouling paint sold contains organotin /24/.The fraction of TBT based antifouling systems is assumed equal 70 % in the model. It should be
noted that this fraction is applied to all vessel types although some types may have alower/higher rate because of the type of trade or service (e.g. cruise ships).
The wetted surface area for the world fleet in 1996 was calculated by the statistical relations
given in Table 4. From these assumptions, the amount of TBT emitted from the ship types wascalculated by equations (3)-(7).
The model results are presented in Table 5, Table 6, Table 7 and Figure 3. The emission of TBTworld wide is estimated to be in the range of 750-1500 ton/year and the total wetted surface area
about 148⋅10
6
m
2
. The main contributing vessel categories are Oil Tankers, Bulk carriers andGeneral Cargo vessels.
Page 12Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
The estimated wetted surface is in good agreement with the estimated wetted surface for the
1989 world fleet /2/ amounting to 136.3?106m². The estimate presented in /2/ is based on 16836ships, mainly above 300 GRT. The figures given in Tables 5-7 are based on ships above 100
GRT and will thus include more ships and the total wetted surface area should be higher thanreported in /2/. This is reflected by the difference in average wetted surface area which is 4100
m² in /2/ and 3400m² in our study. A slightly different breakdown on ship types is applied in /2/.However, the results from our study corresponds fairly well with the results in /2/ when takinginto account that more ships and less ships are included and that the number of ships have
increased from 1989 to 1996.
The amount of antifouling paint applied to ship hulls is calculated in /2/ by the average paint
thickness combined with the wetted surface area. By assuming a volumentric concentration of
BT equal to 30 % in the paint and an average coating thickness of 0.15 mm, the amount of TBTon ships is estimated to 7,2 tons (density of TBT is 1170 kg/m³). This will leach out into the
marine environment over a period for 3-5 years, resulting in an annual input amounting to 1400-2400 tons.
Although there is a considerable uncertainty with respect to the annual input of TBT to themarine environment, the limited information available seems to indicate that the model presentedherein seems to give realistic estimates.
Page 13Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
5 MARIN AIR EMISSION MODEL
Exhaust gas from ship engines contribute to pollution problems. IMO and the internationalcommunity have particularly addressed nitrous oxide (NOx) and sulphur oxide (SOx) emissions
(IMO, MARPOL Annex VI) contributing to acidification problems (NOx also contribute toozone formation), and green house gas (GHG) emissions in the context of the Kyoto agreement.
Carbon dioxide (CO2) is a particularly focused GHG. Exhaust gas from ships include all thesegases and others which contribute to pollution.
Sources and emission factors are summarised in /1/. Technologies for pollution reduction is
addressed in /13/. The model presented in this report enable quantification of world wide
emission levels. The model may also be used to quantify emissions from fleet segments (e.g.tankers). The model may be used to quantify the effect of pollution reduction measuresimplemented for the fleet or a fleet segment (ships above a defined size, specified ship types,etc.). Combined with impact and consequence models, the model described may form a basis for
cost/benefit analysis of pollution control options.
5.1 Model description
A breakdown of the world fleet according to ship type, ship size and engine type is made onthree levels (Figure 4). Level three consists of the fraction of vessels with engine type s for a ship
type i of size k .The model described below is based on the following data:
• Fleet number and tonnage distribution (DWT/GRT) statistics /4/
• Distribution of engine types on ship types and fleet categories /14/
• Established relations between installed engine power and ship size (DWT/GRT); Table 11
• Marine emission factors /1/
• Specific fuel consumption for an engine type category (g/kWh) /1/
• Operating profile for each vessel category /2/, /15/
Page 14Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
Level 0:
Level 1:
Level 2:
Level 3:
Figure 4 A breakdown of the fleet
5.2 Installed engine power estimated by deadweight tonnage
The installed engine power (IEP) in a ship type category i of size k and with engine type s was based on:• The relation between IEP and DWT, given in Table 11
• The number of ships of type i of size k /4/
• The fraction of vessels with engine type s (α ik s) /14/ , /16/
The average size of a ship within a ship type category i and size category k ( xik ) is calculated by
dividing the total DWT (W ik ) on the number of ships ( N ik ) in category ik :
Page 15Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
Equation (15) combined with the relations between IEP and DWT in Table 11, gives the average
IEP ( pik ) for a ship of type i and size k :
ii b
ik
ik
i
b
ik iik N
W a xa p )(⋅=⋅=
(16)
Where:i : ship type category
k : size category N ik : the number of ships in a specific ship type category i and size category k
pik : the average IEP for a ship type category i of size x (k) (kW)W ik : the total DWT for a ship type category i of size x (k) (tonne)
x ik : the average DWT for a ship type category i of size x (k) (tonne)ai , bi : constants from the correlation analyses for a ship type category i
To get the total installed engine power in a vessel, equation (16) also has to include the effect of auxiliary machinery. The equation for installed power related to auxiliary engines is similar to
equation (16), but has other correlation coefficients (see Section 6).
5.3 Number of vessels with a specific engine typeThe number of vessels with engine type s in a ship type category i and size category k (niks) iscalculated by multiplying the total number of vessels ( N ik ) with the fraction of vessels with
engine type s (α iks):
iksik iks N n α⋅= (17)
Where: s : engine type category
niks : the number of ships with a engine type s in a specific ship type category i and size
category k α iks : the fraction of vessels with engine type s for a ship type i and of size x (k)
5.4 Marine energy consumption
The energy consumption for an individual ship during one year may be estimated with:
Page 16Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
Where:
e : energy consumption during a year (kWh) p : installed engine power (kW)
m : average engine loadt : number of operating hours during a year (h)
The energy consumption for all vessels with engine type s in a ship type category i and sizecategories k ( E iks) is calculated by multiplying the individual ship consumption (equation (18))
with the number of vessels with engine type s (niks):
ii b
ik
b
ik iiksik ik iksik ik ik iks N W at mn pt m E
−⋅⋅⋅⋅⋅=⋅⋅⋅= 1)( α (19)
The total energy consumption for all ships of type i and size k ( E ik ) is given by summarising over all engine type s s:
∑=
=S
siksik
E E 1
(20)
Summarising the energy consumption in all size categories k gives the energy consumption in a
ship type category i ( E i):
∑==
K
k ik i E E 1
(21)
Finally, the marine energy consumption during a year ( E) for the fleet may be estimated as a sum
of the energy consumption in each ship type category i:
∑=
≈ I
ii
E E 1
(22)
Equation (22) may also be written as:
∑∑∑= = =≈
I
i
K
k
S
siks E E
1 1 1(23)
Where: E : the total energy consumption during a year (kWh) E i : the total energy consumption during a year for a ship type category i (kWh)
E ik : the total energy consumption during a year for a ship type category i and size category k (kWh)
E iks : the total energy consumption during a year for a ship type category i and size category k and engine type category s (kWh)
mik : average engine load for a ship type category i and size category k
t ik : the number of operating hours (time consumption) during a year for a vessel in shiptype category i and size category k (h)
Page 17Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
5.5 Marine fuel consumption
Using fuel consumption instead of energy consumption gives the same relations as equations(20)-(22) with E (energy consumption) replaced with F (fuel consumption), but equation (19)
changes to:
313 1010)( −−− ⋅⋅⋅⋅⋅⋅⋅=⋅⋅⋅⋅⋅= ii b
ik
b
ik iiksik ik siksik ik ik siks N W at mben pt mbe F α (24)
Where:
F : the total fuel consumption during a year (kg)
F i : the total fuel consumption during a year for ship type category i (kg) F ik : the total fuel consumption during a year for ship type category i and size category k (kg) F iks : the total fuel consumption during a year for ship type category i and size category k and
engine type category s (kg)
be s : specific fuel consumption for engine type category s (g/kWh) /1/.
5.6 Marine time consumption
Some emission factors are given per operational hour /1/. It is then usefully to have an estimate
for the total amount of operational hours for the fleet. The total operation time (T ik ) in a ship type
category i and size category k is calculated by multiplication of the typical individual operationtime (t ik ) for a vessel with the number of ships ( N ik ):
ik ik ik N t T ⋅= (25)
Multiplication of the individual operational time (t ik ) with the number of vessels with engine type s (niks), gives the engine type dependent amount of operation time (T iks) for all vessels of type i,
size k and engine type s:
iksik ik iksik iks N t nt T α⋅⋅=⋅= (26)
Using time consumption instead of energy consumption gives the same relations as equations(20)-(22) with E (energy consumption) replaced with T (time consumption), but equation (19)changes to equation (26).
Where:
T : the total time consumption during a year (h)T i : the total time consumption during a year for a ship type category i (h)T ik : the total number of operating hours during a year in a ship type category i and size
category k (h)
T iks : the total number of operating hours during a year in a ship type category i and sizecategory k and engine type category s (h)
Page 18Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
5.7 Marine exhaust gas emission
The marine exhaust gas emissions may be calculated from the fuel, energy or time consumptionand the emission factors. An aggregation on levels (Figure 4) may be made by using the
equations given below.
5.7.1 Based on energy consumptionEmission level 3:The emissions from vessels with engine type s in ship type category i and size category k are
given by:
iks s g iks g E C M ⋅= )()((27)
Emission level 2:The emissions from vessels in ship type category i and size category k are given by:
∑=
⋅=S
s
iks s g ik g E C M 1
)()(
(28)
The emissions from vessels with engine type s in ship category i are given by:
∑=
⋅= K
k
iks s g is g E C M 1
)()(
(29)
Emission level 1:The emissions from vessels in ship type category i are given by :
∑=
=S
sis g i g
M M 1
)()(
(30)
Emission level 0:The total emissions are given by:
Page 19Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
Where:
g : individual exhaust gas component (NOx, SOx, CO2, CO, HC and PM) M (g) : total emissions for the individual exhaust gas component g (kg)
M (g)i : total emissions for the individual exhaust gas component g for ship type category i (kg) M (g)ik : total emissions for the individual exhaust gas component g for ship type category i and
size category k (kg) M (g)iks : total emissions for the individual exhaust gas component g for ship type category i and
size category k and engine type category s (kg)
M (g)is : total emissions for the individual exhaust gas component g for ship type category i andengine type category s (kg)
C (g)s : emission factor (kg pollution per kWh)
5.7.2 Based on fuel consumption
The emission based on fuel consumption from vessels with engine type s in ship category i andsize category k is given by:
Level 3:
iks s g iks g F C M ⋅=
)()(ˆ (32)
Where:C= emission factor (kg pollution per kg fuel)
Level 2,1 and 0 is calculated in the same way as in section 5.7.1.
5.7.3 Based on time consumption
The emission based on time consumption from vessels with engine type s in ship category i andsize category k is given by:
Level 3:
iks s g iks g T c M ⋅=
)()((33)
Where:
c(g)s= emission factor (kg pollution per hour)
Level 2,1 and 0 is calculated in the same way as in section5.7.1.
Page 20Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
6 INSTALLED ENGINE POWER AS FUNCTION OF SHIP SIZE
This section presents algorithms to establish a statistical estimate for installed engine power asfunction of ship type and ship size. The algorithms are based on data from the DNV SPRINT
database which contains information on about 4500 ships. For each ship type, a correlationfunction is sought for installed engine power as function of ship size. The ship types considered
are given in Table 8. The ship size is given by deadweight tons (DWT) or gross tons (GRT). Toenable conversion of size from GRT to DWT and vice versa, a set of correlation functions (linear and power relations) were calculated from the data. The results are given in Table 9.
Table 8 Breakdown of the world fleet (Lloyd’s, 1996).
Abbreviation Vessel types Number of DWT size categories
LGT Liquid gastanker 14
CT Chemical tanker 14
OT Oil tanker 21
B Bulk 21
GC General cargo 11
RO RO-RO cargo 11
C Container 14
RC Refrigerated cargo 4
P Passenger 41)
1)
Number of Gross tonnage size categories
Table 9 Relations between DWT and GRT
Type Linear relation r (1)
Power relation r (1)
LGT DWT=1.23⋅GRT 0.99 DWT=1.21 ⋅GRT 1.06 0.99
OCT DWT=1.89⋅GRT 1 DWT=0.9⋅GRT 1.06 1
B DWT=1.81⋅GRT 0.99 DWT=0.61 ⋅GRT1.1 0.99
GC DWT=1.42⋅GRT 0.98 - -
C DWT=1.08⋅GRT 0.98 - -
RO DWT=0.56⋅GRT 0.74 DWT=2.44 ⋅ GRT0.82 0.69
P DWT=0.10⋅GRT 0.79 DWT=0.54 ⋅GRT0.85 0.91
(1) Correlation coefficient
This ship types defined in/16/ do not have the same coding system as for example the world fleet
statistics in /4/. Table 10 defines the relations between the coding systems and the codes appliedto establish the statistical relations between installed engine power and ship size for the specifiedship types.
Page 22Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
The data given in /16/ was adjusted as follows:
• Data records with zero value for kW or DWT were removed.• Four duplications were removed.
• Six fast speed passenger ships were removed from the general cargo category.
• One ship was removed from the container category because of a given engine power that was
very small compared with DWT.
• Only ships above 100 grt were considered.
• Fishing vessels were omitted.
The data analysed included about 2500 vessels in total (Table 11). The installed engine power was plotted against ship type and a correlation function was sought. A power function was found
to give the best fit between installed engine power and ship size. The result are given inTable 11. The plots are given in Appendix B.
Table 11 Relations for Engine power (IEP) in kW, based on DWT/GRT
SIEP= The absolute standard error of the predicted IEP-value for each DWT, defined in Section 3.2 (equation (13))(SIEP/IEP)= The relative standard error of the predicted IEP-value for each DWT, defined in Section 3.2 (equation (14))
The auxiliary engines may be a significant source for energy and fuel consumption. Only limited
data were available on auxiliary engines and a breakdown into ship types could not be made. 40vessel in the DNV SPRINT database were analysed. The average installed auxiliary engine
power was about 35 % of the installed main engine power for 5 reefers and 6 passenger vessels.
Most of the auxiliary engines are probably medium speed engines. Class rules requireredundancy in the auxiliary engine system. This results in multiple installation of auxiliary
engine power compared with actual use and needs. A rough estimate may therefore be that theenergy consume from the auxiliary engine system is of the order 10% of the main engine system.
Page 23Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
7 EXAMPLE OF USE; MARINE EMISSION IN 1996
The model described in Section 5 may be applied to quantify the marine exhaust gas emissions.The required input data to the model is given below. The number of ships within defined size
and ship type categories (see Table 8) is based on the Lloyd’s Register World Fleet Statistics for 1996 /4/. Thus, the estimates presented are valid for the world fleet in 1996.
The distribution of engine types on ship types and size categories is based on data from DNVSPRINT database containing about 4500 ships /16/. However, the results presented in Table 12is based on an analysis of 2440 ships (excluding ships below 100 GRT, fishing vessels and
passenger vessels as well as ships for which the coding was found incompatible) /14/.
Table 12 Distribution of engine types on ship types
Ship size categories (GRT)Ship Type
< 1,000 1000 – 6000 6000 – 30000 >30000
Tankers SS:
MS: 69%
HS: 31%
TU:
SS: 22%
MS: 78%
HS:
TU:
SS: 91%
MS: 9%
HS:
TU:
SS: 81%
MS: 2%
HS:
TU: 17%
Bulk SS:
MS: 100%
HS:
TU:
SS: 19%
MS: 78%
HS: 3%
TU:
SS: 90%
MS: 10%
HS:
TU:
SS: 96%
MS: 3%
HS:
TU: 2%
General cargo SS:MS: 72%
HS: 28%
TU:
SS: 12%MS: 85%
HS: 3%
TU:
SS: 52%MS: 48%
HS:
TU:
SS: 98%MS: 2%
HS:
TU:
SS: Slow Speed
MS: Medium Speed
HS: High Speed
TU: Turbine Machinery
The estimate installed engine power is given from the relations in Table 11, which enable
calculation of installed engine power for the given ship size and ship type. The effect and use of auxiliary engines are not included.
Data for the specific fuel consumption can be found in /1/. Representative figures used in this
example are given in Table 13. The operating profile in terms of number of operating hours per year and average engine load is given in Table 14. It should be noted that the operating profile is
very uncertain. The number of operating hours are to a large degree based on expert judgementwith the assumption that larger ships travel longer journeys and therefore have less time in portand more operating hours. The average engine load is also uncertain. In open sea, the ships will
probably run the engines on 85% MCR (Maximum Continuous Rating). This figure has previously been applied in /19/. However, the estimated 70% MCR represents an average figure
Page 24Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
Table 13 Specific fuel consumption.
Engine type Specific fuel consumption
(g/kWh)
Publication/reference
Slow speed 195
Medium speed 215
High speed 230
Turbine machinery 290
/17/, /18/, /1/(1)
(1) Based on testbed measurement and DNV ship onboard measurement (medium speed)
Table 14 Activity profile
Ship size Hours/year Publication/reference Average main engine
load* % MCR
<4999 DWT 4000 0,7
>4999 and <99999 DWT 5000 0,7
Over 99999 DWT 6000
/2/, /15/
0,7
MCR – maximum continues rating
Average main engine load, estimated based on the NOx wwight factors, duty cycle (ref: ISO 8178).
The data given above may be applied to calculate the fuel consumption for each ship type andsize category according to the model described in Section 5.5. The exhaust gas emissions are
then calculated by applying emission factors which expresses the amount of gas emitted (kg) per unit of fuel consumption. Emission factors are summarised in /1/ and applied factors for this
study are given in Table 15.
Table 15 Emission factors applied in this study
Emission factors for engine type /19/,/20/,/25/,/26/ (kg emitted per tonne fuel)Gas component
Slow speed Medium speed High speed Turbine machinery
CO 7.4 7.4 7.4 0.4 NMVOC 2.4 2.4 2.4 0.1
CH4 0.3 0.3 0.3 0.08
N2O 0.08 0.08 0.08 0.08
CO2 3170 3170 3170 3170
Residual 20 x S
(S=2.7%)
20 x S
(S=2.7%)
20 x S
(S=2.7%)
20 x S
(S=2.7%)
SO2
Distillate 20 x S
(S =0.5 %)
20 x S
(S =0.5 %)
20 x S
(S =0.5 %)
20 x S
(S =0.5 %)
NOx 87 57 57 7
PM(1)
7.6 1.2 1.2 2.5
S – sulphur content of oil fuel (% by wt)(1)
Average estimates with the effect offuel type distribution taken into account.
Page 25Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
The resulting amount of exhaust gas emissions from the world fleet from the main engine is
given in Table 16. The estimated fuel consumption is given as well. The total estimated fuelconsumption of the world fleet amounts to 132 Mton. We may assume that the fuel consumption
of auxiliary engines amounts to 10% of the main engine consume. The total estimate for fuelconsumption is then of the order 145 Mton, which corresponds to EIA fuel data indicating a
consumption amounting to 140 Mton /21/.
Table 16 Modelled emissions from world fleet in 1996. Exhaust gas emissions for main
engines only. Figures separated on vessel types and given in megatons (Mton: 106 tons) or
The estimated amounts of carbon (C), sulphur dioxide (SO2) and nitrous oxides (NOx) emissions
based on the model described above are compared with corresponding estimates provided in /22/and /23/ (Table 17). The estimates given by /23/ corresponds fairly well with the results from the
statistical model presented. However, the SO2 emission reported by /23/ are somewhat higher.The estimates presented by /22/ are higher. Both these estimates are based on world wide fuelconsumption for ship transportation and will thus include the auxiliary engines.
Table 17 Comparison of estimates for global exhaust gas emissions from ships.
Source Year C* (Mton) SO2(Mton) Nox(Mton)
Present study. Statistical model(1)
1996 109 5.5 9.9
UNFCCC, 1997/23/ 1994 109 7.5 – 11.5 9.3
Corbett, 1999/22/ 1992/1993 123.6 8.5 10.1
* Emitted amounts of carbon. Approximately 85 % (carbon content by weight) of marine fuel consumption
Page 26Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
8 CONCLUSION
Statistical models to calculate the fleet emissions of pollutants are described. The models cover emissions related to exhaust gas from the engines and the leaching of eco-toxic substances from
antifouling paint. The required input data to the model is identified and described. The modelsare applied to calculate exhaust gas tributyltin emissions for the world fleet of commercial
vessels. The estimated emissions seems realistic because the estimated amount of fuel consumedcorresponds to data provided by the Energy Information Administration /21/, and because theestimated amounts of carbon, sulphur dioxide and nitrous oxides corresponds to other recent
studies /22/, /23/.
The estimated amount of tributyltin in this study is of the order 750 -1500 tons per year. A study
from 1989 /2/ based on paint thickness and volumetric concentration of TBT indicates an annual
input of 1400 – 2400 tons. All the TBT in the applied coating is then assumed to leach out beforedocking. A docking interval of five years and a leaching rate of 4 µg/cm2/day /10/ (IMO
recommended limit) will give corresponding results between our study and /2/, i.e. about 1500tons per year. The estimated total wetted surface area in our study corresponds with the results in
/2/ when taking into account that more and less ships are included in the study presented herein.The model presented then seems to give realistic results.
Page 27Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
9 REFERENCES
/1/ Endresen Ø., Mjelde A., Sverud T. and Sørgård E., Data and models for quantification
of ship pollution, Det Norske Veritas, Technical Report No. 98-2059, Rev. 01, 1999.
/2/ J. Isensee et. al , Emission of Antifouling – Biocides into the North sea- An estimate,
German Journal of Hydrography, Vol. 46, No. 4, 1994.
/3/ Mayell H. and Swanson C., Ban considered on anti-fouling ship paint, article inENVIRONMENTAL NEWS NETWORK, Thursday, December 24, 1998.
/4/ Lloyd's Register of Shipping, World Fleet Statistics, 1996.(http://www.lr.org/information/publications)
/5/ Saunders, H.E., Hydrodynamics in Ship Design, VOL. II, SNAME, 1957.
/6/ Det Norske Veritas, DNV internal database on ship characteristics, 1999.
/7/ Det Norske Veritas, DNV internal memo describing classification codes for ship types
applied by various databases, 1999.
/8/ Lewis E.V. (Editor), Principles of naval architecture, Stability and strength, Volume 1,
2nd revision, Jersey City N.J., SNAME, pp. 310, 1988.
/9/ NTH, Analysis of ship dimensions and ship sizes, University of Trondheim, Chapter 12.8, 1981 (in Norwegian).
/10/ IMO, Committee agrees to ban toxic anti-fouling paint, Marine Environment ProtectionCommittee (MEPC) - 41st session: 30 March - 3 April, 1998.
(http://www.imo.org/news/298/mepc.htm)
/11/ Hunter J.E. and Cain P., Antifouling Coatings in the 1990’s – Environmental,Economic and legislative Aspects, paper presented at the IBC UK Conference:
Managing Environmental Risk in the Maritime Industry, 1997.
/12/ Lintu S., The use of antifouling for ships, legislative status to date and the build up toMEPC 42., paper presented at the IBC UK Conference, Marine Environmental
Regulation: The Cost of the shipping Industry, 1998.
/13/ Sørgård E., Mjelde A., Sverud T. and Endresen Ø., Technologies for pollution
reduction from ship transportation, Det Norske Veritas Technical Report No. 99-2033,1999.
/14/ Hansson L. and Kiær E., Technical Failures - System Criticality Ranking, SAFECO
Work Package II.6, MARINTEK and Det Norske Veritas, Norway, MARINTEK reportno. MT23 F96-0360/233509.00.01, 1997.
/15/ Oftedal S., Air Pollution from sea Vessels, European Federation for Transport andEnvironment, Secretariat: Rue de la Victoire 26, 1060 Brussels, Belgium, 1996.
Page 28Reference to part of this report which may lead to misinterpretation is not permissible.
26 March 2001, M/dnv techn rep 99-2034.doc
/16/ Det Norske Veritas, SPRINT database containing information on DNV classed ships,
1999.
/17/ Klokk S.N., The Norwegian Green Ship Program – low emission diesel engines,
Advanced study Workshop, Air Pollution from Marine Engines, Athens, 20 January1994.
/18/ Harrington. R.L., Marine Engineering, pp. 953, Society of Naval Architects and MarineEngineers, Jersey City, NJ, 1992
/19/ Lloyd’s Register, Marine Exhaust Emissions Research Programme, Lloyd’s Register
Engineering Services, London, 1995
/20/ Intergovernmental Panel og Climate Change (IPCC), IPCC Guidelines for National
Greenhouse gas Inventories, OECD, 1997. (http://www.iea.org/ipcc.htm)
/21/ Energy Information Administration (EIA), International Energy Annual 1997, WorldEnergy database, 1999. (http://www.eia.doe.gov/emeu/iea/main1.html)
/22/ Corbett. J.J., Fischbeck P.S., and Pandis S.N., Global Nitrogen and Sulfur EmisionsInventories for Oceangoing Ships, Journal of Geophysical Research, 104 (D3), 3457 –
3470, 1999
/23/ United Nations Framework Convention on climate Change (UNFCCC), Special issuesin carbon/energy taxation: Marine bunker charges, Working paper 11, 1997.
(http://www.oecd.org/env/docs/cc/gd9777.pdf)
/24/ The Motor Ship, Article: IMO to ban TBT antifouling, May 1998
/25/ E&P Forum (Exploration and Production), Methods for estimating atmosphericemissions from E&P operations, Ref. HN88-001.REP, 1993
/26/ EMEP/CORINAIR, EMEP Co-operative Programme for Monitoring and Evaluation of the Long Range Transmission of Air Pollutants in Europe, The Core Inventory of Air Emissions in Europe (CORINAIR), Atmospheric Emission Inventory Guidebook,
This appendix presents plots of estimated wetted surface area versus ship size for eight defined
ship types. Plots are presented based on two alternative algorithms for the calculation of thewetted surface area. Correlation functions as given in Table 4 are given by solid lines.
This appendix presents plots for installed engine power as function of ship size (given byDWT/GRT) for defined categories of ship types. A further breakdown is made according to
DNV reference codes for ship types (see Table 10). The plots include the correlation line asgiven in Table 11. For each ship type, the distribution of actual installed engine power ( yi) versustheoretical estimates given by the correlation function ( yt ) is given by the error probability
distribution and the cumulative error probability distribution. Actual installed engine power istaken from the database /6/ and the theoretically estimated installed engine power is based on
relations given in Table 11. These distributions indicate the uncertainty for a given ship withrespect to estimated installed power.