Life Cycle Assessment of Pollutants from Ships: The case of an Aframax tanker Anastasios Mountaneas Master Thesis Report no.: SDPO.14.002.m Delft, 24 th January 2014 Anastasios Mountaneas Student no.: 4188349 [email protected]+30 6 974 320 226 Delft University of Technology Mechanical, Maritime and Materials Engineering (3ME) Marine Technology – Shipping Management Thesis committee: Prof. dr. E.M. Van de Voorde, TU Delft Assist. Prof. ir. J.W. Frouws, TU Delft Prof. dr. ir. J.G. Vogtländer, TU Delft Dr. Nikolaos Kakalis DNV GL Dr. George Dimopoulos DNV GL Delft University of Technology
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.
Transcript
Life Cycle Assessment of Pollutants from Ships: The case of an Aframax tanker
Appendix A – Alternative data sources ................................................................................................... 1
Appendix B – Inventory of the case study (115000DWT tanker) ........................................................... 1
Chapter 1 Introduction
1
1 Introduction In this chapter, the background of the thesis is given with respect to the sustainability notion and the
current approaches to ship-generated pollution. The problem to be addressed is defined and specific
objectives are set. Finally, the possible contribution of the thesis outcome to the society and the hosting
company is outlined and the thesis structure is presented.
1.1 Background Sustainability is becoming key topic of the ongoing discussion about the impact of human activities on
the environment. Strategic goals of EU maritime transportation policy for 2009-2018 explicitly mention
sustainability and call maritime industry to start acting under this notion (European Commission, 2009).
However, all recent attempts towards this direction were fragmented and missed the big picture by
focusing on specific areas like energy efficiency and air emissions alone. Other approaches considered
sustainability identical with its environmental dimension and incorporated some economic aspects. All
these drew the attention to particular environmental problems which can be solved with certain
technological solutions (hardware) and misinterpret the broad view of sustainability. Omitted social and
economic dimensions could lead to a total understanding of environmental problems and provide
combinations of solutions from different fields in wider perspective. This becomes apparent when
examining ship’s life cycle; decisions in design and construction have direct impact during shipbreaking,
where non industrial practices are the standard in some sites around the world.
Ships generate different waste streams which can have a global or local effect on the environment. Air
emissions during operation are the most typical example of global impact, which shows the contribution
of shipping to the greenhouse effect and to the ozone depletion phenomenon. On the other side, local
effects are exemplified in the case of ship generated waste. A cruise ship with 2000-3000 passengers
produces daily 550,000-800,000 litres of grey water, 100,000-115,000 litres of black water, 7,000 –
10,500 litres of garbage and 60-130 kg of toxic waste (Oceana, 2008). According to MARPOL 73/78, the
aforementioned waste together with oily waste are firstly treated on board and then delivered to port
facilities for further treatment and final deposition. However, it is not rare that port waste facilities do
not have the appropriate equipment or size, and major problems are caused to the local community.
Ship related environmental impact is regulated by a complex of international conventions and national
laws and rules. Although this legal framework affects a wide spectrum of activities throughout ship’s life
cycle – from the shipyard’s location to the final dismantling practices, it could be described as
fragmented and overlapping. There is not a systematic record of pollutants while discussions to regulate
some of them e.g. particulate matter have just started. Additionally, the main focus is on the quantities
of certain waste streams and not on the actual impacts, e.g. on human health. Thus, any magnifying and
cumulative effect of pollutants throughout lifetime is neglected.
DNV GL, a leading classification society and a global provider of services for managing risk, has a
constant focus on sustainable shipping and innovation. It is an important part of the DVN strategy to
perform research and innovation within key technology areas. DNV GL Research and Innovation (R&I)
Chapter 1 Introduction
2
serves the purpose of exploring and testing new technologies and building new knowledge within
selected technology areas that are believed to be of particular significance for DNV GLs own
development and business activities in the future. In this context, DNV GL R&I would be interested in
having a tool offering a holistic view of the aforementioned topics that could assist in discovering
possible areas for further research. Life cycle assessment (LCA), an ISO standardized methodology, can
be employed for the development of such a tool and provide an alternative method to assess
continuously emerging technologies. Taking advantage of its early involvement in LCA for maritime
applications, DNV GL can develop a specialized tool that makes a step towards sustainability by using its
augmented knowledge on all ship’s lifetime phases.
1.2 Sustainability measures - Environmental assessment Sustainability is a notion with no universally accepted definition yet. However, it is based on a simple
principle: the survival and wellbeing of human race depends directly or indirectly on the natural
environment. Thus, the fulfilling of social, economic and other requirements of present and future
generations are coupled with the condition of the nature. Sustainability provides the conditions under
which human race and nature exist in harmony without the development of one inducing the
degradation of the other.
This approach of sustainability implies that the aforementioned conditions can take the form of
quantifiable limits. However, such limits have not been discovered or invented yet, mainly due to the
concept of sustainability itself. At the 2012 UN RIO+20 conference (UN, 2012), it is stated that
sustainable development “…rests on integration and a balanced consideration of social, economic and
environmental goals and objectives in both public and private decision-making”. Hence, it is evident that,
various disciplines and contexts over different scales of time and space must be considered when
applying the concept of sustainability. For example, sustainable agriculture would include the actions of
the farmer at a local level, the global balance of production and consumption and the interaction with
present and future population. Various decisions and actors are involved together with their individual
perception of well-being.
Several sustainability measures have been developed to account for this inexistence of generally
accepted limits, especially for the environmental dimension of sustainability. The measures have the
form of indicators, indices and methods that attempt to describe the impacts between the human
activities or interventions, ecology and environment. While they do not suggest criteria for sustainability,
they offer a manner to benchmark current practices/products and measure the effect of different
alternatives. Among these measures are included Life Cycle Assessment (LCA), Environmental Impact
1.3 Problem definition Increasing attention has been put in the recent years on the environmental impact resulting from the
ship’s construction, operation and end-of-life practices. However, the focus is usually on the
quantification of certain pollutants (e.g. air emissions from engines) and not on their real impact on the
environment. Following this path, the consequences of a certain pollutant to other impact categories
may be overlooked, as well as the consequences of less ‘popular’ pollutants. Additionally, while
Chapter 1 Introduction
3
decisions with respect to the ship’s characteristics and operational profile may have consequences for
several years ahead and span a wide geographical area, no consistent methodology has been applied to
assess the impact of these decisions on the environment. The life cycle assessment (LCA) methodology is
a promising sustainability metric that provides the framework to develop a specialized tool that records,
quantifies and assesses environmentally all the pollutants generated by a ship. All previous attempts of
LCA for ships had shortcomings (data quality problems, perceived complexity of the method, limited
understanding of the results and mal definition of the problem) that hindered the wide application of
the methodology.
1.4 Objectives The main objective of this study is the development of a DNV GL tool for recording all pollutants
generated by ship related factors and assessing their subsequent environmental impact by analysing the
entire life cycle of a ship.
The following goals should be accomplished:
Systematic recording of pollutants generated during ship construction, operation, maintenance
and scrapping.
Collection and compilation of ship relevant data regarding the interaction with the environment
in every life cycle phase.
Connection of the aforementioned environmental data with ship related parameters. In this way,
results for a particular ship can be generated and correlation between specific environmental
impacts and ship characteristics can be investigated.
1.5 Contribution to the company and the society The developed tool will enable DNV GL to improve its competence and competitive advantage in terms
of environmental performance estimation, eco-labelling and environmental product declaration (EPD)
for seagoing vessels. Such services could be useful for self-assessment, benchmarking, improvement and
branding purposes of of ship operating companies and shipyards. They can also differentiate the price of
similar vessels at the second hand market. Additionally, the tool could be used for crosschecking the
results of other LCA studies or tools, when certification of ISO compliance and results validity is required.
With respect to society, environmental organizations can use the methodology proposed in this thesis to
develop a scheme for shipping companies and shipyards to report their pollutants, as in case of US
Environmental Protection Agency (EPA). The results of this reporting, together with their impact
assessment can be used for environmental benchmarking within the sector and if further elaborated,
they can establish a metric for the environmental performance of the whole sector. Another possible
benefit with respect to society and the environment would be to integrally assess all processes related
to ship’s lifetime, in order to identify “red areas” at the material and energy inflows and outflows, where
sustainability metrics are low and there is potential for performance improvements.
Chapter 1 Introduction
4
1.6 Thesis outline The rest of this thesis is organised as follows. Chapter 2 presents the fundamentals of the LCA, and the
methodological procedure and requirements as set by the relevant ISO standards. Additionally, previous
attempts to apply LCA for ships are reviewed and their main aspects are identified. In chapter 3, the
initial specifications of the methodology are defined and documented; the system, its boundaries and
the assessment method are set. The modelling of the system is described in detail in chapter 4.
Chapter 5 contains a case study and the interpretation of the results. The conclusions derived from this
study and review against the initial objectives is presented in chapter 5. Recommendations for future
work are suggested in chapter 7.
Chapter 2 The LCA Methodology
5
2 The LCA Methodology In this chapter, the general principles of the LCA methodology are presented and the technique is
described according to the ISO standards. Modelling frameworks, methodological aspects and notions
are then discussed with respect to the scope of the present study. Previous attempts to apply LCA for
ships are also investigated and discussed.
2.1 General description of the LCA methodology Life cycle assessment is a technique to address environmental aspects and assess environmental
impacts related to all the stages of a product’s life from raw materials extraction to production, use,
maintenance and disposal or recycling. Natural resources and pollutant emissions are identified and
described quantitatively with processes composing the life cycle model. LCA usually demands extensive
data collection and processing concerning materials, compositions, products, manufacturing procedures,
energy use and corresponding environmental impacts.
Apart from calculating material and energy inflows and outflows, LCA can help in identifying areas for
improvement during a product’s life cycle, comparing alternative process technologies for the same
product or different materials while the product has the same function. Additionally, it suggests relevant
indicators and metrics of environmental performance and can be used as a tool for policy-makers and
marketing (e.g. eco labelling schemes, environmental claims, environmental product declaration). Its life
cycle perspective encourages preventative and proactive environmental management rather than
reactive end-of-pipe approach. In a broader view, LCA itself is a sustainability tool as it directly connects
to environmental impact and makes benchmarking possible.
2.2 The LCA method according to ISO 14040/14044:2006 The ISO standards 14044 and 14040 describe the typical procedure to perform a LCA study. More
specifically, the general framework and principles are provided in standard 14040, while detailed
requirements and guidelines can be found in standard 14044. The four distinct phases comprising a LCA
study are described hereafter.
Goal and scope definition
The product under study, the purpose of the study, how and to whom the results are to be
communicated is explicitly described in the goal and scope definition. The goal of a LCA affects
significantly the extent and the level of detail of the study, which are explicitly defined in the scope. This
specification of the modelling to be performed includes, according to ISO 14040, the following choices
that guide the subsequent work:
Decision of the function of a product or system. Multiple functions are possible even for a single
product and the one selected for study should be described in goal and scope.
Functional unit: expresses the function of a product system in quantitative terms and provides
the reference unit where all inputs and outputs are related. It, actually, defines precisely what is
being studied and forms a basis for comparability of LCA results.
Chapter 2 The LCA Methodology
6
System boundary: defines which processes are included in the system that performs the
function. The criteria for the choice of the processes vary according to goal and audience of the
study, assumptions and data availability. The inventory analysis that follows and the degree of
confidence of the final results depend significantly on the criteria set and the decisions made at
this step.
Environmental impacts considered: early selection of the environmental impacts of the product
system under study is important as it guides process modelling and the collection of relevant
data in inventory analysis. Usually “default” lists of environmental impacts are used including,
e.g. global warming, acidification, toxicity, but not always all of these are relevant to the product
system or the scope.
Data quality description. Data should be accompanied with a general characterization of their
quality in order for the final outcome to be interpreted to correct extend.
Life cycle inventory analysis (LCI)
Life cycle inventory analysis includes data collection relevant to the process model defined in goal and
definition, and calculation procedures that link these data with relevant input, output of the system, the
functional unit and reference flows. In principle, it is a modelling procedure of the processes of the
product system. Usually, the result of the LCI is an incomplete mass and energy balance system, as only
environmentally interesting flows are considered.
Most times LCI starts with building a flow model depicting all processes and material/energy exchanges
between them according to the system boundary set in goal and scope. Then, data are collected for
every process’ energy inputs, raw material usage, products and wastes, emissions to various
compartments and other environmental aspects. After all these data are connected to reference flows
and the flow model is completed, the usage of resources and the emissions can be calculated with
respect to the functional unit.
LCI might seem a straightforward process of collecting data; however it can become very time-
consuming and sometimes complicated due to techniques such as allocation. Allocation is a technique of
partitioning environmental impacts to different products and procedures, when not all of these are
within the scope of the study. Multiple functions can also complicate the partitioning of the
environmental impacts of even one product.
Life cycle impact assessment (LCIA)
LCIA aims to evaluate the significance of potential environmental impacts using LCI results (Baumann &
Tillman, 2004). It attempts to translate inventory data to more environmentally relevant information by
associating them with impact categories and indicators. Thus, another target is to aggregate LCI
extensive information in fewer parameters. The LCIA comprises of three mandatory and four optional
steps (ISO, 2006b):
Mandatory:
1. Selection of impact categories, category indicators and characterization models;
Chapter 2 The LCA Methodology
7
2. Classification: Assignment of LCI results to the selected impact categories;
3. Characterization : Calculation of category indicator results;
Optional:
4. Normalization: Calculation of the magnitude of category indicator results relative to reference
information;
5. Grouping: Sorting and possibly ranking of the impact categories;
6. Weighting: Aggregating indicator results across impact categories using numerical factors based
on value-choices;
7. Data quality analysis: Understanding the reliability of the collection of indicator results.
Life cycle interpretation
Life cycle interpretation is the process of assessing the results from LCI and/or LCIA to draw conclusions.
It is the phase, where raw numbers from inventory and impact assessment calculations, are refined to
provide meaningful results. Interpretation should be in accordance with goal and scope of the study and
should include check procedures to understand the uncertainty of the results. However, LCA is an
iterative process and in case unexpected results occur not in line with goal and scope, they should not
be disregarded. Instead, goal and scope could be reformulated (Baumann & Tillman, 2004).
According to ISO 14044, interpretation comprises of three elements:
1. Identification of significant issues;
2. Evaluation that considers completeness, sensitivity and consistency checks;
3. Conclusions, limitations and recommendations.
2.3 LCA methodological concepts and modelling frameworks The ISO standards provide the necessary steps and the general directions on how to conduct a thorough
and coherent LCA study. However, the provisions of the standards can be interpreted in different ways
as various approaches can be followed to define the boundaries and the modelling components
according to the initial goal and scope of a LCA study. This could result in LCAs with misleading
conclusions, yet compliant to ISO standards. Thus, a robust procedure is needed to interpret properly
the primary goal of a LCA study and specify in detail the procedural steps.
In the following sections, two significant LCA system modelling frameworks are presented. The selection
of the appropriate framework depends largely on the LCA goal and the decision context that this poses.
The specification for the present study is conducted in chapter 3.
2.3.1 Attributional and consequential system modelling frameworks
In attributional LCA, the system under study is modelled as it is or forecasted to be, and it is considered
to be placed into a static technological and economic environment. The environmental impacts
attributed to the system can be calculated over its total life cycle and thus can include all upstream and
downstream value chains, and its end-of-life. The data used to describe the processes comprising the
system can be average (aggregated), generic or process specific depending on the magnitude of the
Chapter 2 The LCA Methodology
8
averaging effect across the value chains. For example, the environmental impact of a single supplier
delivering a significant material to the production facility of the product under study should be ideally
based on specific data, whereas the impact of the electricity consumption of the facility could be based
on average data. As far as the origin of the data is concerned, historical, fact-based or measured data
can be used within the attributional framework.
Consequential LCA attempts to describe the consequences of a decision in the foreground system to
other systems and processes both inside and outside the life cycle of the product and the economy. In
contrast with the attributional LCA, no actual or forecasted models of the supply and value chains with
respect to the product systems under study are used, but instead, generic supply chains that interact
with the market mechanisms, political decisions etc. are modelled. A typical example of consequential
LCA application is the changing of the level of output, consumption and disposal of a product. The
modelling of the causal relationships originating from this change can include expected measures such
as material bans, green incentives, opening (or closing) of production plants, displacement of competing
products etc. The data used in consequential LCA are of higher uncertainty than in attributional, as this
the modelling depends heavily on economic models, demand and supply and market mechanisms.
2.3.2 Foreground and background systems
The modelling of the product system heavily depends on the goal of the particular LCA study and it can
differ considerably between studies of different goal, even though the same product system is of
interest. The processes of the product system can be differentiated into two levels: the processes of the
foreground system and those of the background system. The foreground system refers to the system of
primary concern while the background system provides energy and materials to the foreground system
as aggregated data sets in which individual plants and operations are not identified. The differentiation
between foreground and background systems is based on two different perspectives (EC-JRC-IES, 2010b):
(1) the specificity perspective, where the system processes are characterised with respect to whether
process specific data or average/generic data are needed, and (2) the management perspective, where
the system processes are characterised on whether they can be managed by “direct control or decisive
influence” within the decision context of the LCA study.
With respect to the specificity perspective, the foreground system contains the processes that are
specific to it and thus specific data (manuals, suppliers etc.) are the most appropriate and must be used.
For example, the foreground system for a product LCA study conducted for its producer contains all the
processes in the production facility and the specific upstream and downstream processes that cannot be
sufficiently described by the market average data. On the other side, the management perspective
identifies a wider set of processes during the lifecycle as foreground, depending on the varying impact of
the decisions made. It includes all in-house processes of the producer/ operator of the system under
study, the processes downstream and upstream which the producer/ operator can influence or specify
and processes during the usage and end-of-life of the product that can be influenced by either direct
decisions or design characteristics of the product. In the case of ship recycling for example, all related
processes are considered foreground since their environmental impact largely depends on the material
selection made by the ship owner at the design stage.
Chapter 2 The LCA Methodology
9
The background system in the specificity perspective contains all the processes that can be
appropriately represented by average and generic data due to the average effect, for example, across
the suppliers. A theoretically homogenous market is assumed to provide the processes to the
foreground system. Under these terms, the usage and end-of-line phases of a product are background
processes in a LCA study conducted on behalf of the producer. The management perspective identifies
as background processes the part of the system that no sufficient control or influence by the
producer/operator/user exists. For example, the steel production for steel parts is a background process
for the steel parts purchaser.
2.4 Literature review: ship focused LCAs Application of the LCA method in the maritime industry was attempted in the past by joint efforts of
companies and research institutes and proved that its application is possible with relative success. The
main problem in all studies was the lack of ship relevant data. Simplifying assumptions and use of data
from other sectors for similar processes did approach the magnitude of the impact, however with high
uncertainty. Herein, most thorough LCA applications for the shipping sector are presented in
chronological order, to enable observing the different approaches of researchers through time.
2.4.1 LCA of “M/V Color Festival”
LCA on the shipping sector was used for the first time on the passenger ship “M/V Color Festival”
(Johnsen & Fet, 1999). Actually, it was a screening of LCA and its purpose was to demonstrate and
evaluate the methodology on this sector. The main conclusion was that dividing ship in different systems
and subsystems can greatly facilitate data collection and process modelling. Data for the construction
phase originated from actual reports from Norwegian shipyards and where no appropriate data were
available, databases provided by the LCA software SimaPro (PRé Consultants, n.d.) were used. Generally,
this data is not specific for ships or maritime related processes, as they intent to describe the market
average or common practice in many industries.
The functional unit for the study was the “ton-km transported per year between Oslo and Hirshals”. The
vessel was considered to transport a certain number of passengers, cars and trailers per year between
Oslo and Hirshals and her life time was assumed 20 years.
Results showed that different environmental impacts were considered important at different life phases
of the ship:
Global warming, acidification, eutrophication, smog and energy consumption for the operational
phase.
Solid waste from the scrapping phase.
Local impacts like toxicity for humans and ecology for construction and maintenance.
The impact analysis concluded to the most important processes per life phase of the ship. For example,
processes related to oil combustion and antifouling paint leaching was considered as being important
during the operation phase. Processes related to painting were considered as being important during
maintenance. Finally, the recycling of materials was important for the scrapping phase assessment.
Chapter 2 The LCA Methodology
10
Using the impact assessment method Eco indicator ’95, impacts to the categories of human toxicology
and acidification were also identified. Finally, the main conclusion was that using system approach, i.e.
dividing the ship in different systems and subsystems can greatly facilitate data collection and process
modelling.
2.4.2 EVEA ‘LCA-Ship’ tool
“LCA-SHIP” (Jivén et al., 2004) is one more custom made LCA software developed by a Swedish
consortium of shipping companies, authorities and institutions. The interesting feature of this tool is the
modelling of the ship’s energy systems, energy flow and exchanges during the operation phase of the
ship.
The first step at the energy systems modelling is the calculation of the required propulsion power. The
ITTC-78 ship performance prediction method is used with most input data having default values
according to regression of all displacement ships tested at SSPA Sweden AB. The modelling of the energy
systems on board is based on a systems approach. The main components of the system, e.g. main
engine, shaft generators, economizers, are presented as interconnected blocks where the user defines
the percentage of energy consumed with respect to the energy entering the system. Types of energy
considered are: mechanical, thermal and electrical. The purpose for this kind of modelling is to allow for
checking saving strategies and optimization procedures. Yet, no optimization procedure is provided by
the tool and it should be noticed that the aforementioned user-defined percentages are directly
translated to the fuel consumption of the power production system in the beginning of its lifetime (main
engine, generators and boilers).
The inventory contains data from all the life cycle phases. However, losses and recycling rates of
materials used at the construction and scrapping phase must be defined by the user. An effort with
relative success was put in order to include the impact from cradle-to-grave of the important
materials/equipment used during the life cycle phases. The numerous assumptions made towards this
direction led to inconsistences with respect to the analysis boundary and the quality of the final results.
For example, the engines and boilers are considered to have the same environmental impact with that
of equal weight steel, aluminium’s production impact is the average of EU whereas the steel is
considered produced in Sweden, the air emissions of electricity production are based on that of the EU
at 1994. Additionally, several data of known poor quality from previous LCA studies (e.g. M/V Color
Festival) were used (e.g. the welding length during construction is based on a very small vessel). Some
other limitations of the inventory include the use of only two types of fuel (heavy and diesel oil) and that
drydockings involve only sandblasting and no painting. Sandblasting is prohibited by most countries
nowadays.
With respect to impact assessment, five impact categories are evaluated within the tool: acidification,
eutrophication, global warming, ozone depletion and photochemical ozone production. The results
obtained are endpoint impact indicators and thus, the optional elements of normalization, grouping and
weighting are implemented. The main impact assessment method is the Ecoindicator ’99, while the
following three are also available: EDIP, Tellus, EPS2000 and EPS money
Chapter 2 The LCA Methodology
11
More comprehensive and complete methods have been developed however nowadays (EC-JRC-IES,
2010a). The Eco indicator is a damage-oriented method, mainly developed for assessing industrial
products and not assemblies like ships. The EDIP method is a midpoint method which applies weighting
factors on the basis of political environmental targets set by the Danish Government or by various
international protocols(Wenzel, Hauschild, & Alting, 2000). The Tellus evaluating scheme was developed
with respect to the Swedish context and uses data on society's willingness-to-pay to calculate the
weighting factors. They use both data on emission taxes (the Swedish CO2-tax) and marginal costs for
reducing emissions down to decided emission limits for certain criteria pollutants (CO, NOX, PM10, SOX,
VOC and lead). The EPS system is mainly aimed to be a tool for a company's internal product
development process. The impact of materials used is expressed as an index which represents money
and it is linked to society’s willingness to pay for the protection of endpoint categories.
2.4.3 NMRI ‘LCA for ship’ tool
The National Maritime Research Institute of Japan (NMRI) has performed a long research program
(NMRI, 2006), which its main outcome was a ship specific LCA software focused on the Japanese
maritime context. A database was compiled by collecting data from shipping companies, shipyards and
on site investigations. More specifically, data for the LCA inventory of the construction phase were
developed from the construction of a 76000 DWT bulk carrier in a Japanese shipyard. Data for the
operation of the ship (average operating profile, trip pattern, engine load) were investigated with use of
the navigational logbooks of five specific ships: oil tanker, bulk carrier, container ship, LNG carrier,
RO/RO carrier. Shipbreaking operations in China were also surveyed, however no data are referenced
(M. Kameyama, Hiraoka, Sakurai, Naruse, & Tauchi, 2005). Additionally, LCA was performed for the
construction of a marine main engine and was incorporated in the software. The definition of the ship to
be analysed in the tool includes, apart from the main particulars, a specification list (weight, on board
position) of parts, hazardous materials etc. that has to be prepared by the user. The quantification of the
environmental loads and LCI analysis are carried out according to the matrix method, meaning that data
are treated only with linear relations. Data for background procedures were retrieved from the Japanese
LCA database
The life cycle impact assessment is performed with the LIME model, which is an indicator especially
created for the Japanese conditions (geography, population, current environmental load). The functional
unit used is ton-mile for all ship types. The software also provides indicators represented by the ratio of
the value obtained with the load on the environment products and services through its life cycle. These
indicators are intended for use in the Japanese environmental product declaration which is an eco-
labelling scheme.
Let us notice that all published literature (apart from two papers) from the aforementioned project and
the software itself are in Japanese.
2.4.4 ‘Sustainable Ship Design (SSD)’ tool
The SSD tool (SSD) is the most recent LCA application developed by a French consortium of
environmental consultants, design offices, shipyards and suppliers. The tool is an add-on for SimaPro
Chapter 2 The LCA Methodology
12
and has two versions: a complete tool for use by building sites and architectural offices and a tool
adapted to suppliers and subcontractors.
The complete tool intends to measure the environmental impacts of all the ship throughout the entire
life cycle. The modelling of the life cycle includes six stages: materials for construction (raw material
extraction, production of ship equipment, transportation to the yard), assembly on site (consumptions
and wastes at the ship yard), ship operation and maintenance, ship dismantling and end-of-life (fate of
dismantling wastes). Each life cycle step is modelled as a set of inventories of inflows and outflows of
every material used, process, consumption, machinery component etc. Thus, life cycle steps are
modelled to a large extend according to a typical bill of materials and not as expressions with respect to
ship parameters. This means that the tool user should explicitly know or approximate the types and/or
values for the inflows and outflows of all ship systems and components.
The tool for suppliers and subcontractors aims to measure impacts of their products which have an
adapted life cycle, from raw material extraction to end product life. Again here, an inventory of inflows
and outflows must be established by the user.
The assessment of the environmental impacts at both tools is accomplished by using the impact
assessment module of SimaPro and applying existing LCIA methods. The following impact indicators
have selected to describe the environmental impacts (Tincelin, Mermier, Pierson, Pelerin, & Jouanne,
2007):
Global warming – IPCC 2007 (CO2 equivalent)
Eutrophication (PO4 equivalent)
Atmospheric acidification (SO2 equivalent)
Ozone layer depletion (CFC11 equivalent)
Human toxicity (1.4-DB equivalent)
Fresh water aquatic eco toxicity (1.4-DB equivalent)
Marine aquatic ecotoxicity (1.4-DB equivalent)
Terrestrial ecotoxicity (1.4-DB equivalent)
Respiratory effects (PM2.5 equivalent)
Abiotic depletion (Antimony equivalent)
Water (m3)
Energy consumption (MJ equivalent)
Bulk waste production (kg)
Hazardous waste production (kg)
Within the SSD tool, a simplified methodology has been developed based on the Admiralty’s constant
and a simple life cycle model (Tincelin et al., 2007). The methodology uses a criterion to provide a first
evaluation of the effect of a green technology on an existing ship design and to compare green
technologies, given their LCAs.
Chapter 2 The LCA Methodology
13
2.4.5 BAL.LCPA
BAL.LCPA (BAL, 2013) is a decision making tool for shipyards and ship operators, developed under the
EU FP7 Collaborative Project BESST (Breakthrough in European Ship and Shipbuilding Technologies, 2009
– 2013). The BESST project focused at the Life Cycle Performance (LCP) assessment of ships concepts,
demonstrating the life cycle impact of maritime technological solutions compared to current designs.
The project was kicked off on 2009 and ended in February, 2013. The consortium consisted of 50
partners, including shipyards, manufacturers, universities and research institutes, classification societies
and maritime industry stakeholders. The total project cost was approximately 28 million euros.
The BAL.LCPA tool assesses the life cycle performance of vessels and their subsystems by combining
financial, environmental, safety and other societal aspects within a single methodology. The most
common application is the comparison of different technical options in the early design phase, in order
to determine the one(s) which is (are) most likely to deliver the best performance over time.
Additionally, it can be used to measure the performance of an already existing system.
Apart from modelling the technical performance of components, BAL.LCPA also considers risk of loss
and hidden costs such as repair and maintenance costs. Additionally, different scenarios with
parameters which evolve through time can be defined by taking into account environmental taxation,
fuel prices, currency exchange rates, discount rates etc.
The comparison and assessment of solutions is based on a list of Key Performance Indicators. Some of them are:
Net present Value (NPV) is a commonly used economic measure to describe the incoming and
outgoing cash flows over time at today’s value. It enables the comparison of economic
profitability among alternative solutions/investments.
Expected NPV, the NPV enhanced with the probability factors of certain accidents and the
connected costs.
Environmental indicators which measure the impact of relevant emitted gas on certain
categories. Equivalent CO2, SOX, NOX and PO4 quantities of gases are used to estimate the impact
on climate change, acidification, photochemical oxidation and eutrophication respectively.
Social Welfare Index (SWI) assesses the effects and the perception of technical components on
the human nature.
2.4.6 Other attempts
Other LCA applications in shipping sector limit the life cycle perspective to either specific environmental
impacts or specific ship type. Norwegian University of Science and Technology (NTNU) developed a LCA
application based on SimaPro especially for fishing vessels. Energy consumption and air emissions were
compared between different ship types and other transport modes in the study of (Kristensen, 2000).
Chapter 2 The LCA Methodology
14
2.4.7 Discussion on ship-focused LCAs
In this paragraph, all previous LCA studies/tools that their product system was a ship will be compared
and discussed. The target is to identify the points that the LCA practitioners agree and that would help
our research, and to locate areas which could be elaborated in the present study.
The main characteristics of the previous attempts are summarised in Table 1. In total, five documented
applications of the LCA method for ships were found. The first three , LCA of “M/V Color Festival”
(Johnsen & Fet, 1999), ‘LCA for ship’ (NMRI, 2006) and ‘LCA-Ship’ (Jivén et al., 2004), were conducted by
academia and thus, were followed by detailed documentation. The most recent, ‘Sustainable Ship
Design’(SSD, 2010) and ‘BAL.LCPA’ (BAL, 2013), were mainly driven by consultants which probably
explains their poor documentation.
In all studies, the product system defined is a vessel of particular type. The functional unit adopted by
academia studies expresses the supply of transport and hence ton-km is used for all ship types.
Alternatively, the results are presented per transported cargo unit (i.e. ton, TEU). The most recent
studies use a custom environmental index or the time dependence of the environmental assessment
results during the life time.
The tools developed cover all common ship types, although not tested. (Johnsen & Fet, 1999) is a
screening study for a specific RoPax vessel and part of the data used were collected from the ship’s
constructor and operator. The documentations of (Jivén et al., 2004), (SSD, 2010) and (BAL, 2013)
suggest that all ship types are within the scope, given that appropriate data are available. However, (SSD,
2010) is tested on special ship type, e.g. frigate and sailboat. On the other side, the tool of (NMRI, 2006)
was successfully tested for the types shown in Table 1.
The different applicability of the tools on various ship types is explained by examining the modelling of
the system for the inventory analysis and the data used. In case of extensive use of databases, either
very few or extensive details of the materials and energy used during the ships life cycle are required.
This depends on the modelling level of detail in terms of processes considered. For example, (Jivén et al.,
2004) focuses on the energy consumption on board by providing an extensive, yet simplistic, model of
the on board machinery and thus, requires details with respect to the energy flow among the different
components. The environmental impact of the shipbuilding phase is calculated based on data from the
previous study of (Johnsen & Fet, 1999). In this way, data from a RoPax vessel are extrapolated for
modelling other vessel types without considering possible discrepancies. The most consistent modelling
approach was made by (NMRI, 2006), where actual data from a panamax bulker were extrapolated for
similar ship types and databases were used only for background processes.
Every study uses different impact assessment method. The reason is that all studies were conducted at
different years and used the best available, or in other words, the most popular at that time. None of
the methods was selected based on criteria related to the maritime context. The LIME method used by
the study (NMRI, 2006) is an monetary assessment method developed specifically for the Japan,
introducing a locality of the LCA results. In contrary, all other assessment methods have a global scope.
Chapter 2 The LCA Methodology
15
Table 1 Main characteristics and comparison of previous LCA tools and studies for ships. Each row of ‘environmental impact category considered’ contains identical, similar or linked environmental phenomena. Empty cells indicate that no similar phenomenon to the other attempts was considered (Source: own composition).
This decision context determines the LCI modelling frame (attributional or consequential) and in turn,
the processes included within the system boundary, the processes modelling and corresponding data
collection, and other decisions made during the inventory modelling and the impact assessment.
According to our goal definition, the study has a descriptive character and does not support any decision
directly. We are interested in the environmental impact occurring within a certain temporal window (i.e.
the ship’s life cycle) based on decisions already taken. Therefore, the system should be modelled as it
could be measured and the attributional modelling framework (see section 2.3.1) is the most
appropriate.
The ship is considered operating in a static economic and technological environment. Its technical
characteristics and equipment are unaltered since its construction (apart from time degradation and
maintenance substitutions). The models and parameter values used for describing the operational
profile and practices should be representative of the actual conditions during the life cycle. In this sense,
present and past data of adequate quality can be used for depicting/forecasting the real behaviour of
the ship from cradle-to-grave. Similarly, economic parameters, market mechanisms, future legislation
impact and interaction among them are not modelled and thus, not interacting with the life cycle phases
of the ship.
The attributional modelling framework adopted in this study introduces certain requirements for the
type (specific, average or generic) of data specifying the processes included within the system boundary.
This topic will be addressed in the chapter 4 where the system model is analysed in detail.
3.2.4 System Boundary
Ship’s life cycle is considered to start at keel laying and finish just after the final dismantling is completed.
The lifecycle is divided in three major phases: construction, operation and shipbreaking. The operation
phase consists of sailing and maintenance stages which are consecutive and repeated over time.
Therefore, the boundary of the product system is set on the fence line of the shipyard and shipbreaking
yard, and on the port. Raw materials acquisition, transportation of materials to the shipyard, production
of goods (e.g. fuels, consumables etc.) and every other process that takes place outside the shipyard’s
fence line is omitted in the present LCA analysis. Figure 1 illustrates the system boundary, the basic
processes and material/energy flows under study (within red frame).
Chapter 3 LCA initial specifications
20
Energy Sources
Materials (e.g. coatings)
Ship parts(e.g. machinery)
Shipbuilding Consumables
Electricity Production
Shipbuilding Ship
Electricity
Electricity
Sailing
Maintenance
Waste
Port Reception Facilities
Ship Shipbreaking
Recycled materials
Reused materials
Deposited materials(e.g. landfill, dump)
Figure 1 Boundary of the system under study. Framed with red colour are the main life cycle phases and product/energy flows (Source: own composition).
Chapter 3 LCA initial specifications
21
As far as the operation phase is concerned, the boundary is set exactly on the port quay. The interaction
between ship and port occurs through the delivery of wastes (oily wastes, garbage etc.) to the port
facilities as mandated by the MARPOL 73/78 convention. Normally, these wastes should be treated
before their final disposal in a landfill or an incinerator. Several treatment technologies are
available(Hess, Snuverink, Schoof, & A.M de Leeuw, 2004) with different applicability depending on the
port profile (Wolterink, Hess, Schoof, & Wijnen, 2004). However, the existence of such facilities varies
considerably according to local policies (Carpenter & Macgill, 2005; EPE, 2003).
Ideally, the life cycle of the system in attributional modelling should include the whole supply chain and
downstream, i.e. from raw materials extraction to end-of-life treatment and return of substances to
earth(ISO, 2006a). It is evident that the boundary of this study includes only the foreground processes
(see sec. 2.3.2) with respect to the ship’s total supply/value chain. The rationale behind this boundary
selection resides in the objective of this thesis to calculate and assess the interaction between the
environment and a ship of specific characteristics. This suggests that the point of view of the shipowner
and/or the shipbuilder is of interest for this LCA analysis.
With respect to the management perspective described in section 2.3.2, the boundary should include
those processes controlled directly or influenced decisively by these two actors. Thus, these processes
should be for the present study: all in-house functions (e.g. shipyard processes, ship operation profile),
processes at suppliers that are influenced by choice (e.g. blasting material type at repair yard, fuel type),
end-of-life processes since they are influenced by decision and design of the ship (shipbreaking assumed
as in third-world countries). The processes considered as background typically for attributional
modelling (EC-JRC-IES, 2010b) are those of tier-two suppliers and long term contractors/suppliers that
cannot be affected considerably.
Since the analysed ship has specific characteristics, the processes that particularise her life cycle should
also be included in foreground modelling according to the specificity perspective of section 2.3.2. Again
all in-house processes and tier-one suppliers should be considered as foreground and specific data
should describe them. In contrast, the processes far downstream and upstream the supply/value chain
(i.e. tier two suppliers) of the ship belongs to the background and averaged data can be used.
3.2.5 Basic assumptions
The modelling of the three phases composing ship’s lifetime is based on several assumptions with main
criterion the availability and quality of data with respect to ship-related processes. Only major
assumptions affecting the whole model are referred in this paragraph. Minor ones affecting only certain
processes (e.g. painting) will be mentioned in chapter 4.
At the shipbuilding phase, the environmental impact from conveyance within shipyard is not
considered. The number and capacity of cranes, trucks and forklifts vary significantly from
shipyard to shipyard and for different reasons making even estimations not a safe option.
The ship-port interaction is taken into account only as the delivery of bulk amounts of waste and
garbage. Their chemical composition has large variation depending on ship type, on board
treatment equipment (e.g. incinerator) and operator’s policy.
Chapter 3 LCA initial specifications
22
The impact of the emissions to local level (e.g. the vicinity of the shipyard or occupational
exposure) will not be assessed.
The end-of-life scenario assumes the current scrapping practices of the so-called shipbreaking
nations (India, Pakistan and Bangladesh). These countries account for more than 90% of the
industry while the Hong Kong Convention which regulates environmental, health and safety
issues of shipbreaking is estimated to come into force after 2020 (see chapter 0).
3.3 LCIA method selection During the ship’s lifecycle phases pollutants are released in all three compartments –air, soil and water-
and contribute to environmental impacts with different extent. Various impact assessment methods
exist that differentiate in the number of impact categories addressed and the models to assess them. In
this section, an attempt is made to determine the most relevant set of impact categories and the most
appropriate LCIA method for the life cycle of a ship from the recommended methodologies presented in
the ILCD handbook (EC-JRC-IES, 2010c).
Generally, the release of a substance to the environment triggers a series of consecutive phenomena
that finally affect three areas of protection: human health, natural environment and natural resources.
Most LCIA methods use a cause-and-effect approach to model the chain of the phenomena. Depending
on the position along this chain, the impact categories are distinguished in midpoint and endpoint. The
endpoint categories coincide with the aforementioned areas of protection and express the relative
importance of emissions and their consequences at the end of the cause-and-effect chain. Midpoint
impacts are defined as the link between the initial releases and final consequences, and each one can
cause more than one endpoint impact. The ILCD handbook proposes ten midpoint impact categories to
be checked for relevance in all sectors: climate change, ozone depletion, human toxicity, respiratory
inorganics, ionising radiation, photochemical ozone formation, acidification in land and water,
eutrophication, ecotoxicity, land use and resource depletion.
In line with the objectives of the thesis, the environmental impact assessment of a ship should include
the widest possible set of impact categories. Review of the categories considered in the previous ship
relevant LCA attempts (see section 2.4) and communication with industry professionals revealed that
the following impacts should be at least assessed: global warming, acidification, eutrophication,
photochemical oxidation, ozone depletion, toxicity for humans and ecology. The set of the midpoint
categories is a trade-off between the available categories of each LCIA method and the quality of the
underlying models, since different methods categorise the environmental impact differently. In the
present study, the ReCipe 2008 method was selected (see below) and the following midpoint categories
will be assessed:
Climate change;
Ozone depletion;
Terrestrial acidification;
Freshwater eutrophication;
Human toxicity;
Photochemical oxidant formation;
Chapter 3 LCA initial specifications
23
Particulate matter formation;
Terrestrial ecotoxicity;
Freshwater ecotoxicity;
Marine ecotoxicity.
It should be noticed that no model assessing the impact of invasive species due to ballast water
transportation could be found. The ionising radiation impact, although available by ReCiPe, will not be
assessed since it is found irrelevant with ship’s lifecycle (only smoke detectors contribute insignificantly
if not recycled/reused after ship dismantling).
The models used to calculate the midpoint impacts of the pollutants are less complex than those for
endpoints, as the amount of forecasting and effect modelling is reduced (Bare, 2002). Less
environmental mechanisms have to be modelled and thus assumptions and value choices are minimised.
While endpoint modelling allows for aggregation across impact categories (e.g. comparison climate
change and ozone depletion categories on human health), the requirement for reliable data and robust
models, which are often not available, reduces the level of comprehensiveness (Bare, 2002) and
increases the uncertainty of the results (Baumann & Tillman, 2004). This is the reason that impact
assessment at midpoint level is implemented at the present study.
Numerous LCIA methods are available with modelling at midpoint, endpoint or both levels. The EC-JRC-
IES provides the most recent attempt (EC-JRC-IES, 2010c) to distinguish the most up-to-date, thorough
and reliable methods used currently, based on a comparison methodology including a set of uniform
criteria including geographic differentiation, i.e. varying the impact of a substance with the region’s
characteristics ,such as population density, type of use (urban, rural) etc. To select the appropriate LCIA
method among the recommended by EC-JRC-IES with respect to the objectives of the present thesis, the
following criteria were set:
Range of impact categories and number of relevant substances covered;
Appropriateness of impact categories covered for ship’s LCA study;
Quality of modelling at midpoint level.
Based on these, the ReCiPe 2008 method was proved to be the most appropriate. It covers all the
midpoint impact categories selected above and provides the most extensive list of assessed substances
(covers 3000 substances while the next most extensive covers 1000 approximately). Additionally,
groups of pollutants that are difficult to speciate are used as reference substance in certain categories
and hence, the uncertainly is reduced. For example, NMVOCs, which are a collection of numerous
organic compounds is used as reference substance in the photochemical ozone formation category.
With respect to midpoint impact modelling, ReCiPe provides a harmonised approach in terms of
modelling principles and value choices (EC-JRC-IES, 2010c; M. Goedkoop et al., 2009). Thus, the
calculated results refer to the same point of the cause-and-effect chain and are uniform as far as the
regional and compartment conditions are concerned. This approach increases the validity of the results
and allows for selecting the location of e.g. the shipyard, between urban or rural area. The main
characteristics of the selected ReCiPe 2008 method are presented in the following section.
Chapter 3 LCA initial specifications
24
3.3.1 The ReCiPe 2008 impact assessment method
The ReCiPe method(M. Goedkoop et al., 2009) is a life cycle assessment model which incorporates the
advantages of two other models: the CML2001 (Guinée, 2002) and Eco-indicator ’99 (M. Goedkoop, R.
Spriensma, 1999). It uses the approach of the CML2001 for assessing the midpoint impacts and the main
characteristics of the endpoint impact assessment of the Eco-indicator ’99. Contrary to the other models,
ReCiPe provides 18 midpoint impact categories and three endpoint indicators, in order to evaluate the
life cycle impact.
Figure 2 Flowchart of ReCiPe model. Left: LCI parameters, middle: midpoint categories, right: endpoint categories (M. Goedkoop et al., 2009).
Figure 2 illustrates the categories and the connections among different impact levels. At the first level,
the results of the life cycle inventory are connected to the 18 midpoint categories and they are
converted to equivalent amount of the reference substance per category. At the second level, each
midpoint category is connected to the relevant endpoint categories, using the same categorization as in
the Eco-indicator 99 model. It should be noticed that a midpoint category can be connected with more
than one endpoint category. Where no adequate assessment method exists for linking the categories
between the two impact levels, the corresponding impact is not taken into account in the endpoint
results. An important category for the present study which falls within this case is marine eutrophication.
Another advantage of the ReCiPe model is that it provides factors for each level category (midpoint,
endpoint) with respect to three cultural perspectives:
Chapter 3 LCA initial specifications
25
Individualist: short term perspective, optimism that technology will ameliorate many problems
in the future
Hierarchist: mid-term perspective, consensus scientific model
Egalitarian: long-term perspective, precautionary attitude for the future
These perspectives represent a set of choices for time perspective and for expectations for the
environmental effect of management and future technology development. Table 3 presents the
quantitative connection between the midpoint and endpoint categories. The second column contains
the reference unit used for the quantification of the corresponding category. And the next three
columns contain the factors for translating midpoint results to the endpoint categories with respect to
the three different cultural perspectives. For example, 1kg of CO2 at the climate change midpoint
category can be converted to human health endpoint category by multiplying with 1.19x10-6 year/kg for
the individualist perspective, 1.40x10-6 for the hierarchist perspective and 3.51x10-6 for the egalitarian
perspective.
Table 3 Midpoint to endpoint factors (M. Goedkoop et al., 2009). The (-) mark indicates that no connection is established between the categories. The (*) mark indicates that the value depends on the type of substance or land use. “I, H, E” correspond to individualist, hierarchist, egalitarian perspective respectively. Up-to-date values are presented (M. Goedkoop et al., 2013).
Midpoint impact category
Unit
Endpoint impact category
Human health (DALY)
Ecosystems (species.yr)
Resources ($)
Climate change kg (CO2 eq. to air)
1.19E-06 (I) 8.73E-09 (I) -
1.40E-06 (H) 8.73E-09 (H) -
3.51E-06 (E) 1.87E-08 (E) -
Ozone depletion kg (CFC-11 eq. to air) * - -
Terrestrial acidification kg (SO2 eq. to air)
- 1.52E-09 (I) -
- 5.80E-09 (H) -
- 1.42E-08 (E) -
Freshwater eutrophication kg (P eq. to freshwater) - 4.44E-08 -
Human toxicity kg (1,4-DB eq. to air) 7.00E-07 (I,H,E) - -
Photochemical oxidant formation kg (NMVOC to air) 3.90E-08 - -
Particulate matter formation kg (PM10 eq. to air) 2.60E-04 - -
Terrestrial ecotoxicity kg (1,4-DB eq. to soil) - 1.51E-07 (I,H,E) -
Freshwater ecotoxicity kg (1,4-DB eq. to freshwater) - 8.61E-10 (I,H,E) -
Marine ecotoxicity kg (1,4-DB eq. to marine water)
- 1.76E-10 (I,H,E)
-
Ionising radiation kg (U235 eq. to air) 1.64E-08 - -
Agricultural land occupation m2/year - * -
Urban land occupation m2/year - * -
Natural land transformation m2 - * -
Fossil depletion kg (oil eq.) - - 5.17E-02 (I)
- - 1.65E-01 (H +E)
Metal depletion kg (Fe eq.) - - 7.15E-02
Chapter 3 LCA initial specifications
26
3.4 Chapter conclusions The first two steps of the standardised LCA procedure as described in chapter 2 were carried out in the
previous paragraphs. The goal of the study was set and its key points are the quantification of pollutants
during ship’s life cycle, their assessment and the compilation of ship-relevant data wherever possible.
Within the scope definition, two significant characteristics that influence the rest of the study were
made.
The first is the selection of the attributional modelling framework which determines how the system
should be modelled. According to this, the system should be modelled and appropriate data should be
used as it can be measured now, i.e. no future changes in technology, legislation, or market are taken
into account. This facilitates the models of the next chapter, since they can be static (i.e. time invariant),
and the respective data which can be past of present and still used for forecasting/depicting the
behaviour of the ship over its life cycle, although its spans many years ahead. For example, the currently
common ship breaking practices can be used for calculating the impact of ship’s end-of-life, although
new legislation might come into force and thus, change the final impact.
The definition of the system boundary in section 3.2.4 is the second item which is critical for the
selection of the relevant processes to be modelled in the next chapter. The boundary includes the
construction, operation and dismantling of a ship and is set on the fence line of the respective facilities.
The processes included are the so called “foreground” in paragraph 2.3.2, and are those that are under
the direct control of the ship owner or the shipyard management. This is a clear criterion for selecting
the processes to be modelled per life cycle phase in the next chapter.
The conclusion that our study includes the attributional modelling of the foreground processes out of a
ship’s life cycle also suggests which data types should be used for modelling in the following chapter.
Specific data, i.e. data either collected from a specific case or specific to maritime processes, are the
most appropriate, and should be used when available. Average data representing the current situation
should be the second option. Finally, generic data (e.g. combustion pollutants calculated through
stoichiometry) can also be used when no other data is available.
In this chapter, the impact categories considered and the assessment were also selected. All categories
addressed in previous studies (see sec. 2.4) are included, apart from resource depletion since raw
material extraction is outside the boundary of our study. The assessment method to be used is the
ReCiPe and its application will be demonstrated through a case study in chapter 5.
All the decisions taken and the specifications of the LCA method made in this chapter are of paramount
importance for the rest of the study. The objectives of the thesis were interpreted in the definition of
LCA goal and the boundary of the life cycle. The next chapter implements the inventory analysis step of
the LCA methodology according to process selection criterion, attributional modelling framework and
the data requirements set in the previous paragraphs.
Chapter 4 Life Cycle Inventory (LCI) analysis
27
4 Life Cycle Inventory (LCI) analysis The life cycle of the ship is divided in three distinct phases: construction, operation and dismantling.
Further, operation comprises of sailing and maintenance occurring in regular periods.
Modelling of each phase was based on how inflow of materials and energy are converted to products
and releases to the environment. This approach, apart from the apparent physical meaning, facilitates
data collection and discovering the relation between inputs and outputs. Thus, construction and
dismantling are modelled as sets of interconnected work processes, while, for the operation phase, a
system based approach is followed. However, transition from one phase to another does not affect final
results as the functional unit remains the same and reference flows are process specific.
For each process, material and energy inputs and outputs are identified and then a transfer function
that links them is constructed, usually involving several parameters that vary with the ship type, size,
materials used etc. The identification and specification of these parameters introduces the problem of
data quality and results uncertainty, when applying LCA in the shipping sector, something already
diagnosed in previous LCA attempts (see chapter 2.4). Stakeholders in the three phases have different
interest in the processes and, therefore, information is organized in a different way than that required
by the LCA methodology. Shipyards use semi-empirical ways of generating production material
information (Roh & Lee, 2007) from concept design level information, in order to plan block erection
and materials procurement, and to start construction as soon as possible. Only after detailed design,
accurate information for all construction stages becomes available. Of course, this information is not
openly available. Finding generic correlation functions is also very difficult as production material
information is mainly based on shipyard practice and not on ship parameters.
In operation phase, coupling of environmental impacts and ship characteristics is more straightforward,
as most releases to the environment are system-bound and have been thoroughly researched. However,
certain releases, such as coatings leaching, are product specific and depend on shipping companies’
policy making their association with ship parameter difficult.
Dismantling phase data gathering exhibits one major problem: data scarcity. Most of the global
dismantling activity takes place in third world countries, where collection of data appropriate for LCA is
the least concern.
4.1 Construction Construction includes all work within the shipyard’s area. Environmental impacts related to materials
entering or leaving this boundary are not considered part of this phase. Therefore, six main processes
were identified: welding, cutting, painting, conveyance, sea trials and overhead (lighting, heating, office
work etc.). However, sea trials and conveyance are not included in the analysis. Sea trials have a minor
contribution to environmental impact, while relations between conveyance and ship parameters could
not be found in the literature.
Chapter 4 Life Cycle Inventory (LCI) analysis
28
4.1.1 Electricity Consumption
Electricity is used in most shipbuilding processes. While not producing direct environmental impact
within the shipyard area, its consumption can be accounted for aerial pollutants at the power plants.
The quantity and mix of the pollutants depend on the energy mix of the country where the shipyard is
located. For example, countries with extensive renewable energy usage produce less greenhouse gases
per kW. For this reason, data for three representative countries (Japan, USA, EU-27) can be selected
depending on where the ship was build. Total emissions of CO2, CO, CH4, N2O, NOX, NMVOC, SO2 from
total electricity and heat production for the year 2010 were taken from national reports submitted to
UNFCCC (UNFCCC). Allocation of the total electricity output and heat emission was based on the
International Energy Agency (IEA) 2010 reports. 2010 is the most recent year that data were available
for all three countries both in UNFCCC and IEA. The emission factors calculated are shown in Table 4.
Table 4 Emission factors for greenhouse and acidification gases due to electricity production in three representative geographical areas.
Emission Type Emission Factor (gr/kWh)
Japan USA EU-27
CO2 3.408E+02 5.058E+02 3.156E+02
CH4 1.217E-03 4.850E-03 3.460E-02
N2O 5.279E-03 1.325E-02 6.840E-03
NOX 2.468E-01 3.993E-01 3.615E-01
CO 6.923E-02 1.427E-01 1.354E-01
NMVOC 3.276E-03 1.041E-02 1.544E-02
SO2 1.767E-01 1.160E+00 6.499E-01
4.1.2 Welding
Four welding methods are the most common in shipbuilding worldwide (Eyres & Bruce, 2012a; Jackens,
2012): flux core arc welding (FCAW), shielded metal arc welding (SMAW), gas metal arc welding
(GMAW), submerged arc welding (SAW). All methods incorporate electrodes (wire or rod), electricity
and possibly shielding gas, apart from the base metal to be welded.
All welding calculations for a specific ship are based on the amount of consumed electrodes and their
allocation per method. Yet, this information is derived more accurately only after the detailed design
stage. Estimation in earlier stages is possible by semi-empirical methods developed by each shipyard
according to its practices and experience (Roh & Lee, 2007). In our study, welding material is a function
of steel weight. According to a study for a panamax bulk carrier (Hiraoka et al., 2001), the actual welding
length to steel weight ratio is 0.0323 km/ton. The total amount of welding material accounted for a
panamax bulker at DSME shipyards is 140 tons, according to the actual documents delivered by the
shipyard. Based on that, the welding material consumed over the welding length can be derived; the
calculated value is 0.0112 ton/m. The allocation of this material to the welding methods is shown in
Table 5 (Jackens, 2012).
Chapter 4 Life Cycle Inventory (LCI) analysis
29
Table 5 Common welding methods mix in shipbuilding industry (Jackens, 2012).
Welding method
Allocation of weld material consumed
FCAW 82.00%
SMAW 7.00%
GMAW 1.00%
SAW 10.00%
Figure 3 shows the sequence for calculating air and health hazardous emissions from welding processes
in a shipyard.
WELDING
Wsteel
Electricity
PM10
Metal emissionsWeld method mix
Air emissions
Welding lenth
Welding consumables
Figure 3 Calculation sequence for emissions from welding process.
Metal substances are normally contained in the electrode to improve the quality of the weld. When the
electrode melts, part of them is released in the air together with particulate matter with diameter less
than 10 μm (PM10). Table 6 shows the emission factors of most significant metals (chromium, hexavalent
chromium, manganese, nickel, lead) and particulate matter. The emission factors express the quantity of
the pollutant released per consumed weld material.
Table 6 Emission factors for common shipbuilding welding processes (US-EPA, 1995).
Shielding gas is used in FCAW and SMAW methods to isolate the weld area from atmospheric gases and
thus, prevent lowering the weld quality and facilitate the procedure. Pure noble gas (usually argon), CO2
or more often in shipbuilding, a mixture of both is used as shielding gas. The amount of CO2 released to
atmosphere can be derived by the optimal volumetric flow rate of the shielding gas, its CO2 content and
the weld material consumption rate that are prescribed by the manufacturers. Noble gases are not
considered having significant environmental impact and thus are not measured. Review of data sheets
of major manufacturers (ESAB, ELGA and Hyundai) showed that the aforementioned parameters do not
alter considerably for different weld materials. Additionally, the common parameter values in practice
(Weisman & Kearns, 2001) are in accordance with the values from manufacturers. Table 8 contains the
most usual in practice shielding gas flow rate, CO2 volumetric content and the derived CO2 quantity
released in the atmosphere per unit weight of consumed weld material.
Table 8 Common values in shipbuilding practice for shielding gas (mix) flow rate, its CO2 content and the respective CO2 release in atmosphere per weld material consumed (Weisman & Kearns, 2001).
Parameter Unit Welding method
FCAW SMAW GMAW SAW
Shielding gas flow rate litres/min 18 23 not appl. not appl.
CO2 content %(volume) 25% 20% not appl. not appl.
CO2 release per weld material consumed
litres/kg 21 145 not appl. not appl.
4.1.3 Cutting
Most common methods in use for forming plates into the required shapes in shipbuilding are plasma arc
and oxy-fuel flame (Eyres & Bruce, 2012a, 2012b). For most steel parts, numerically controlled cutting
Chapter 4 Life Cycle Inventory (LCI) analysis
31
machines with plasma arc are used, whereas oxy-fuel (natural gas or propylene) is used for planning and
minor cutting done with hand.
In plasma-arc method, an electricity arc turns a high flow rate inert gas into plasma with enough
temperature to melt the steel plate. Considerable amount of particulate matter is released in the air, if
cutting is performed in open air. However, underwater cutting can decrease this amount significantly.
Table 9 shows the emission factors for particulate matter production per meter cut in open air and
underwater.
Table 9 Particulate matter (PM10) emission factors for plasma cutting in different conditions. The values are given per length cut (Steiner, Bach, Windelberg, & Georgi, 1988) .
Plasma cutting method
PM10 Emission Factor [gr/m]
Minimum value Maximum
Open air 10 1000
Underwater 0.03 10
To calculate the total emissions from the cutting process per ship, estimation for the total cutting length
is required. In this study, cutting is considered correlated with welding. Based on actual data (M.
Kameyama, Hiraoka,K.,Tauchi,H., 2007), cutting per weld length ratio has a value of 0.482 m/m and
electricity consumed per meter cut is 2.222 kWh/m. Figure 4 illustrates the sequence to calculate the
pollutants from cutting.
CUTTING
Welding length
Cutting method mix
Electricity
PM10
Air emissions
Cutting length
Figure 4 Calculation sequence for air emissions and PM10 from cutting process.
Total cutting work depends on ship size and shipyard organization. Several parts can be cut from a single
steel plate (nesting). However, the size of the plate is affected by several factors, e.g. crane capacity,
cutting machine size and capacity, ship’s plate thickness, making accurate calculations per ship very
difficult.
4.1.4 Painting
Material inputs for painting are primarily paints and solvents. Paints contain the pigments and binder
which provide the desired characteristics to the final coating. Solvents are used to dilute the paint,
Chapter 4 Life Cycle Inventory (LCI) analysis
32
facilitate the process and clean the equipment. The solvent portion of the paint is released into the
atmosphere through diffuse emissions during painting, paint drying and the use of cleaning solvents.
Table 18 presents the emission factors per pollutant type and engine type. The amount of a particular
pollutant produced by the combustion in the main and auxiliary ship machines depends on various
parameters, with most significant being the fuel type, engine type, ambient conditions and specific fuel
consumption. To directly estimate the air emissions of a particular ship, data from on board
measurements of exhaust gas emissions and fuel flow rate are needed (Salonen, Heikkinen, & Ilus, 2012).
However, these measurements are not typically present on board and are rarely published. Thus, only
indirect estimation methods based on average data compiled from emission inventories. In the present
study, data from three different studies were compiled to produce the emission factors of Table 18.
Table 18 Emission factors for gas compounds related to the engine types (European environment agency, 2000; Lloyd's Register, 1995; Statistics Norway, 2000).
Gas Component Slow Speed
Medium Speed
High Speed
Turbine machinery
Carbon monoxide (CO) kg/kg fuel
7.4 7.4 7.4 0.4
Non Methane Volatile Organic Compounds (NMVOC)
kg/kg fuel
2.4 2.4 2.4 0.1
Methane (CH4) kg/kg fuel
0.3 0.3 0.3 0.08
Nitrous oxide (N2O) kg/kg fuel
0.08 0.08 0.08 0.08
Carbon dioxide (CO2) kg/kg fuel
3170 3170 3170 3170
Sulphur dioxide (SO2)
Residual Fuel 2.7% sulphur content
kg/kg fuel
54 54 54 54
Distilate Fuel 0.5% sulphur content
kg/kg fuel
10 10 10 -
Nitrogen oxides (NOX) kg/kg fuel
87 57 57 7
Particulate Matter (PM10) 7.6 1.2 1.2 2.5
Finally, Table 19 contains the ratio of auxiliary over main engine installed power per ship type.
Chapter 4 Life Cycle Inventory (LCI) analysis
40
Table 19 Auxiliary engine (AUX) power versus main engine (ME) power (DNV, 2002; Trozzi, 2010).
Ship type AUX/ME (%) Ship type AUX/ME (%)
CT 30 R 24
LGT 20 RO 20
OT 30 P 45
B 30 OSV 35
C 25 OOA 35
GC 23 OA 10
4.2.1.2 Hull performance
Hull and propeller performance refers to the relation between the condition of the underwater hull and
propeller and the propulsion power required to move the vessel at a given speed. Generally, the
performance deteriorates over time due to the increase of the hull and propeller surface roughness.
Roughness is a key parameter that affects frictional resistance and it can be categorized in two types
(Willsher, 2008):
Physical or basic roughness that results from coatings build up, coatings cracking, detachment,
corrosion, repeated spotblasting, welds, mechanical damages etc.
Biological roughness that the accumulation of unwanted living organisms on underwater
surfaces (bio-fouling).
Bio-fouling is a biological phenomenon whose type, severity and extent vary greatly depending on the
type of the antifouling coating, ship’s trading pattern and operational profile (i.e. vessel speed and
activity) (T. Smith, 2013; Willsher, 2008). At dry-docking, the ship’s underwater hull and propeller are
blasted and repainted and thus bio-fouling is totally removed. Biological roughness is eliminated,
however the physical roughness is not (even the blasting for preparing the hull surface for painting
creates some) and it accumulates over ship’s life time. Therefore, the deterioration of the hull
performance during the total life time can be assumed as the superposition of two phenomena:
1. The constant accumulation of mechanical damage over lifetime.
2. The periodical effect of bio-fouling over a sailing interval (the interval between dry dockings)
that builds over the abovementioned mechanical damage.
From an operational perspective, the permanent accumulation of mechanical damage onto the hull and
the respective physical roughness built up, is observed as the increase in fuel consumption measured
exactly after the ship leaves the drydock (bio-fouling is removed) with respect to the previous drydock.
According to discussions with two ship-operating companies the increase is approximately .
This suggests that the phenomenon follows a geometric progression pattern over life time with common
ratio ( ) . The fuel oil consumption exactly after the th dry-docking ( )can be
calculated by applying basic geometric progression formulas:
, Εq.5
Chapter 4 Life Cycle Inventory (LCI) analysis
41
Where,
is the number of drydockings during life cycle
is the common ratio
is the initial fuel oil consumption of the unfouled newbuilt ship
is the fuel oil consumption exactly after the th dry-docking (bio-fouling totally removed)
Due to the small value of the common ratio, the increase of fuel consumption follows an almost linear
pattern with respect to time for the longest part of life cycle (see Figure 6). However, the increase
becomes sharper and deviates from linear as the ship approaches her last dry-docking and end-of-life.
This is in accordance with the studies of Townsin (R. L. Townsin, 2000; R. L. Townsin, Byrne, D., Svensen,
T.E. and Milne,A., 1986), which proposed a linear increase of the average hull roughness over time (until
approximately 20 years) due to mechanical damage and thus an almost linear increase in fuel
consumption. It should be noticed that no reblasting was considered in the studies of Townsin.
Reblasting plays a major role in the accumulation of mechanical damage over life time and it is taken
into account in our case.
Figure 6 Geometric and linear evolution of fractional FOC increase for due to mechanical damages during lifetime. The ship is scrapped after 27.5 years of service. The vertical red lines indicate the drydockings. Notice the increasing difference between the two curves after the 17
th year.
The cyclical effect of bio-fouling between dry-dockings is complex phenomenon influenced by various
parameters such as coating type, ship type , speed, intermediate hull washings, geographical locations
of ports of call (Chad Hewitt, 2011). With respect to life cycle calculations and current market average
conditions, it can be adequately modelled as a linear increase of the fuel oil consumption exactly after
the th drydocking ( ). This assumption can be verified by observing published data by coating
manufacturers and measurement equipment suppliers (ENIRAM, 2012; International Paint, 2004;
0.00%
1.00%
2.00%
3.00%
4.00%
5.00%
6.00%
7.00%
8.00%
9.00%
0 5 10 15 20 25 30
Frac
tio
anal
FO
C in
cre
ase
Years after launch
Geometric Linear
Chapter 4 Life Cycle Inventory (LCI) analysis
42
Kjølberg, 2013; Wallentin, 2012). According to this data, the fuel consumption increase ( ) to maintain
a certain speed before a 5-year dry-docking, can range between 5-30%, with most common values being
between , with lower values observed for new coating types. It should be noticed that
this measured fuel consumption increase includes the effect from both mechanical damage and bio-
fouling. The fuel oil consumption at a given time is calculated as follows:
( ) ( )
( )
(
)
Εq.6
With
The fuel consumed during a sailing interval , - can be calculated using the Εq.8:
( ) ∫ ( )
( ) (
) Εq.7
The total fuel oil consumed during lifetime is ( and denote the scrapping and last drydock time
respectively):
∫ ( )
∫ ( )
∫ ( )
Εq.8
For convenience, the two integrals are treated separately:
∫ ( )
( ) (
)
( ) (
)
Εq.9
∫ ( )
∑ ∫ ( )
∑ , ( ) ( ) -
( ) ∑ [( ) ]
Εq.10
Chapter 4 Life Cycle Inventory (LCI) analysis
43
If the drydockings occur at constant sailing intervals and
, then the time
period between the last drydocking and the scrapping, and the total life cycle duration can be written
respectively:
Εq.11
( ) Εq.12
Hence, the equations Εq.9 and Εq.10 yield to:
∫ ( )
( )
Εq.13
∫ ( )
( ) ∑ [ ]
( )
Εq.14
Therefore, the total fuel oil consumed during lifetime is:
[ ( ) ( )
]
Εq.15
Or
[ ( )
( )
]
Εq.16
From the above equation, the total fractional fuel oil consumption increase ( ⁄ ) over a ship’s
life time can be derived:
[ ( )
( )
]
Εq.17
For a ship undergoing drydockings every 5 years, scrapped after 27.5 years and mechanical damage and
biofouling impact to fuel oil consumption equal to and respectively, equation
Εq.17 gives a 9.38% total fuel oil increase over life time .
The total fractional fuel oil consumption increase ( ⁄ ) according to Εq.17 is plotted in Figure
7 for (green line). The total fractional FOC increase (green line) follows a linear pattern
between the FOC increase at the start and the end of the sailing interval (blue and green marks
respectively). The FOC increase at the end of the sailing interval is calculated as an increase ( ) of the
FOC at the beginning of the interval, as shown by the stepwise curve (blue line). The dashed line (light
Chapter 4 Life Cycle Inventory (LCI) analysis
44
blue) is the FOC growth due to the physical roughness of the hull over time (see equation Εq.5 and
Figure 7).
Figure 7 Total fractional FOC increase (green line) during a 27.5 year life cycle. Red vertical lines indicate the drydockings. Green marks and blue marks indicate respectively the FOC increase just before and exactly after drydockings. The rightmost green point corresponds to the end of operational life (27.5 years) and thus, the FOC increase has not reached the maximum possible value, like in the previous drydock intervals.
4.2.1.3 Ballast water
Total annual ballast quantity transported by a ship depends on her type and size and can be calculated
by equation Εq.18:
Εq.18
where Εq.19
Table 20 explains the symbols and subscripts of equations Εq.18 and Εq.19 and enlists the tables with
relevant data. The number of ballast trips is assumed equal to the number of load trips, and thus
can be taken from Table 22. Utilization rate is assumed equal to 0.86. According to equation Εq.19,
average ballast capacity and DWT are correlated in a linear fashion, where the slope is given in Table
21. The correlation coefficient has high values, confirming the validity of the linear relation. The results
also agree with other studies (Carlton, Reid, & Leeuwen, 1995; D. Smith, Wonharn, McCann, Reid, &
Carlton, 1996).
0.00%
5.00%
10.00%
15.00%
20.00%
0 5 10 15 20 25 30
Frac
tio
nal
FO
C in
cre
ase
Years after launch
Total FOC fractional increase FOC fractional increase growth after drydocks Fractional FOC after drydocks
σ2
σ1
Chapter 4 Life Cycle Inventory (LCI) analysis
45
Table 20 Explanation of symbols and subscripts in equations Εq.18 and Εq.19.
Subscripts
Ship type, * + Table 14
Ship size category Table 14
Symbols
Total annual ballast water per ship type i and ship size k , -
Average ballast capacity for ship type i and ship size k , -
Annual average ballast trips for ship type i and size k Table 22
Average ballast capacity utilization rate for ship type i and size k
Relative ballast capacity for hip type i Table 21
Average DWT of ship type k and size i -
Table 21 Correlation between DWT and normal ballast capacity for different ship types. αi denotes relative ballast capacity for ship type i and r denotes correlation coefficient.
Ship type Number of ships
Minimum DWT Maximum DWT Mean DWT
Oil tanker 0.34 0.97 25 5121 312638 120993
Bulk ships 0.35 0.97 21 25439 224222 95275
Container 0.37 0.95 13 5500 51880 27487
Chemical tanker 0.43 0.90 13 9176 45000 24126
Liquefied gas tanker 0.40 0.97 12 3451 77591 39521
Antifouling coatings gradually release toxic substances into the water to kill off the fouling and after a
period of 3-5 years they become depleted and need replacement. After the ban of the tributyltin (TBT)
based coatings by IMO (IMO, 2001), the most common biocides used copper based (mainly cuprous
oxide Cu2O) and in some cases zinc oxides (ZnO). Because copper and zinc based coatings are not as
effective as the previous TBT based, co-biocides (or boosters) are added as supplementary to make
them effective against a broader variety of aquatic species. Combinations of different co-biocides in the
same coating are common, usually containing two biocides (the main one and the co-biocide). Most
common co-biocides are diuron, irgarol 1051, chlorothalomil, seanine 211 and zinc pyrithione. Cuprous
oxide and co-biocides are harmful to the marine environment and through the food chain can be
hazardous to humans too. Thus, different countries have banned different components, leading to a
fragmented legislation at a worldwide level with significant differences in approved products. The paint
manufacturers report this only to the institutions charged with approval of the specific product in their
country and the information of the market share of a specific product is company confidential.
Fouling release coatings, also known as “non-stick”, are not using biocides to prevent biofouling, but
they are forming a very slick surface onto which organisms cannot easily adhere and in case they do, the
surface can be washed off by the ship’s motion when sailing or by cleaning equipment. As FR coatings
are not leaching biocides to prevent biofouling, they are considered environmentally safe, although
some minor complications might exist (Hydrex, 2012).
Extensive literature exists for the calculation of the release rates of the aforementioned biocides.
However, the results vary widely due to differences between the measurement methods and the
representiveness of the measurements with respect to actual leaching in practice. For the present study,
the arguments for the selection of emission factors for the Netherlands emission registry (Cotteleer,
2012) are adopted. Table 23 shows the emissions factors of Cu2O and co-biocides for copper based
antifouling coatings. Fouling release coatings are considered to have marginal releases to the
environment and thus their emission factor is zero.
Table 23 Leaching rates (emission factors) of cuprous oxide and co-biocide for antifouling and foul release coating types (Cotteleer, 2012). The leaching rate unit express the biocide released per painted unit area per day.
Coating type Component Leaching rate [μg/cm2/day]
Sailing At port
Antifouling (AF) Cuprous oxide (Cu2O) 6.0 4.500
Co-biocide 0.9 0.675
Foul release (FR) - 0 0
In the absence of appropriate release rates for the different types of co-biocides, it is assumed that each
antifouling coating contains the main biocide (Cu2O) and one co-biocide. Table 24 lists the most
Chapter 4 Life Cycle Inventory (LCI) analysis
47
common co-biocides in use worldwide (Dutch national waterboard, 2008a) and their CAS (chemical
abstracts service) numbers.
Table 24 Common co-biocides and corresponding CAS numbers.
Co-biocide CAS number
Irgarol 1051 28159-98-0
Zineb 12122-62-7
Diclofluanid 1085-98-9
Zinc pyrithion 13463-41-7
Seasine 211N 64359-83
Tolyfluanide 731-27-1
4.2.1.5 Sacrificial anodes
All metal surfaces that are in contact with sea water are susceptible to corrosion, i.e. dissolution of the
metal into the water. Corrosion leads to severe degradation of the strength properties and thus, can
have destructive results. The areas of the ship that are the most susceptible to corrosion are the
underwater part of the hull and the ballast tanks. Coatings are used to prevent corrosion, yet the
protective layer is not always sufficient as damages can occur and certain areas (e.g. the propeller) are
not coated. For this reason, two types of cathodic protection (passive and active) are also used to
prevent corrosion.
Passive cathodic protection makes use of sacrificial anodes, which are metallic parts fitted on the hull
and ballast tanks and dissolve into the water, protecting steel of the ship’s structure. Zinc and aluminium
are the most common anode metals for marine application (Willems, Schouten, & Heidbuurt, 2003).
Active cathodic protection uses impressed current (IC) to protect the metal. The system includes a
transformer, a rectifier, power source and non-consumable anodes. The IC system is considered to have
no direct environmental emissions, since it does not dissolve into the water, unlike the sacrificial anodes.
Active cathodic protection is not used in ballast tanks due to the risk of fire/ explosion from the
formation of hydrogen gas in combination with the electrical system. For the exterior surfaces of the
ships, IC systems are often used together with sacrificial anodes in order to protect certain areas that
are more susceptible to corrosion, such as the bow thruster tunnel and the rudder. Table 25 shows the
application percentages of the protection methods for the seagoing vessels. It is noticeable that 60% of
the ships are only coated and they do no use cathodic protection at all.
Table 25 Application percentages of the corrosion protection methods per ship area (Willems et al., 2003).
Protection system Ship area
Underwater hull Ballast tanks
Zinc anodes 70% 10%
Aluminium anodes 12.5% 30%
IC system 17.5% 0%
No anodes 0% 60%
The amount of the anodes consumed over a period of time can be calculated by the equation Εq.20.
Chapter 4 Life Cycle Inventory (LCI) analysis
48
Εq.20
Table 26 provides the explanation of symbols and subscripts of equation Εq.20, and enlists the tables
and sections with relevant data.
Table 26 Explanation of symbols and subscripts in equation Εq.20.
Subscripts
Protection system: zinc or aluminium anode -
Ship type, * + Table 14
Ship size category Table 14
Operating mode: Sailing =1, At port =2
Ship area: Underwater hull or ballast tanks
Symbols
Amount of anode dissolved of type g for a ship of type i of size k with
protection system s at mode m, -.
Average utilisation factor for the ship area s for a ship of type i.
Surface of protected ship area s for a ship of type i , - Sec. 4.1.4
Corrosion rate for protection system g at mode m
, -. Table 27
Average number of operating days during a year at mode m for protection
system s installed in ship of type i and size k [days].
Equation Εq.20 can be used to calculate the amount of anode dissolved into the water of a particular
ship area. It should be reminded that the IC system does not produce direct releases to the environment.
The final amount depends on the time that the anode actually works, i.e. it is submerged into the water.
For the underwater hull, the working time is considered that of the total operating days per year, as it is
given in Table 17 per ship type and size category. For the ballast tanks, the working time can be taken
equal to half of that for the underwater hull, based on the assumption that the ship operates on round
trips. An average utilisation factor for the surfaces is also introduced, in order to compensate for the
actual part of the ship area that is submerged into seawater. Again the anodes protect only the surfaces
that are fully submerged. For the underwater hull, the utilisation factor should account for the variation
of draught during the considered time period. Yet, no correlation formulas for the draught variation with
the operating profile of ships could be found in the literature. For ballast tanks, the average ballast
capacity utilization introduced in paragraph 4.2.1.3 could be used as an approximation.
The corrosion rates of both anode types change considerably when the ship sails at open sea and when
at berth, as it can be seen in Table 27. Therefore, the total operating time must be partitioned between
Chapter 4 Life Cycle Inventory (LCI) analysis
49
the two different operating modes. The annual port time has already been calculated based on Table 17
and on the assumptions b) and e) mentioned in paragraph 4.2.1.1.
Table 27 Corrosion rates of zinc and aluminium anodes per ship area and per ship type for two different situations: sailing and when at berth. Elaborated from (Dutch national waterboard, 2008b).
Ship area Ship type
Corrosion rate *μg/cm2/day]
Zinc Aluminium
Sailing At port Sailing At port
Underwater hull
Bulk Ship 46.2 11.5 13.8 3.5
Container Ship 46.2 11.5 13.8 3.5
Chemical Tanker 46.2 11.5 13.8 3.5
General Cargo 61.5 15.4 18.5 4.6
Liquefied Gas Tanker 46.2 11.5 13.8 3.5
Other Activities 67.7 16.9 20.3 5.1
Offshore Other Activities 67.7 16.9 20.3 5.1
Offshore Supply Vessel 67.7 16.9 20.3 5.1
Oil Tanker 46.2 11.5 13.8 3.5
Passenger Vessel 61.5 15.4 18.5 4.6
Reefers 46.2 11.5 13.8 3.5
RoRo vessel 61.5 15.4 18.5 4.6
Ballast tanks All types 276.9 - 83.1 -
The corrosion rates presented in Table 27 have been derived using the Dwight’s formula, the required
electrical current density and the electrical capacities of the anodes’ materials(Dutch national
waterboard, 2008b). Zinc has a considerably lower capacity than the aluminium (780 Ah/kg versus 2,600
Ah/kg) that results in higher corrosion rate, in order to achieve required electrical densities for
protecting the steel of the ship structure. The high values of the rates for the ballast tanks are due to the
conditions met there. The tanks contain stationary water for long time periods and air which is
extremely humid with poor ventilation; hence, a highly corrosive environment is created.
4.2.1.6 Produced Waste
A general model for estimating the amounts of produced waste from different subsystems on board as
ship is described in the present chapter (Ø. Endresen & Sørgård, 1999; NMD, 1994; Schnitler, 1995).
Equation Εq.21 gives a general description of the model:
∑
∑
∑
∑
Εq.21
where
Εq.22
Additionally, equation Εq.23 should be used for sewage and garbage production:
Chapter 4 Life Cycle Inventory (LCI) analysis
50
Εq.23
Symbols of equations Εq.21, Εq.22 and Εq.23are explained in Table 28.
Table 28 Explanation of symbols in equations Εq.21, Εq.22 and Εq.23.
Subscripts
Waste type -
Ship type, * + Table 14
Ship size category Table 14
Operating mode: Open sea =1, Harbouring =2, Other=3
Waste system type
Symbols
Waste production of type g from a ship, -.
Amount of waste produced of type g for a ship of type i of size k with waste systems s at
mode m , - .
Waste based production factor for waste type g in relation to ship system s at mode
m,( ) ⁄ .
Number of persons on board (passengers + crew) at mode m for a ship of type I of size k
(persons).
Waste production rate at mode m for a ship of type I of size k ,( )
-
Average number of operating days during a year at mode m for waste system s installed
in ship of type I and size k [days].
Equations Εq.21, Εq.22 and Εq.23 need detailed data for waste production rates for different operating
modes and subsystems. These data are not always available and for this reason average production
rates are used which are not always varying with operating mode and subsystem type.
4.2.1.6.1 Sludge system
Sludge system waste cannot be discharged in sea according to MARPOL Annex I. After separation,
residues are stored in special tanks or barrels and are delivered in appropriate port reception facilities.
The amount of sludge system waste can be estimated by equations Εq.21, Εq.22 ( ) and
data of Table 29. Let us notice that these numbers are based on studies of Norwegian ports (NMD, 1994;
NSFI, 1977) and include lubricating oil residues too.
Chapter 4 Life Cycle Inventory (LCI) analysis
51
Table 29 Estimates for produced amount of sludge per day as function of ship type and ship size category (NMD, 1994).
This category contains oily wastes from waste oil collectors, lubricating oil basins, hydraulic oil changes,
oil separators etc. The quantities that are not collected end in the engine room bilge spaces. The biggest
part of liquid oily waste is either incinerated or delivered at reception facilities ashore. The annual
production of liquid oily waste can be calculated through equations Εq.21, Εq.22 ( ) and
Table 31.
Chapter 4 Life Cycle Inventory (LCI) analysis
52
Table 31 Average quantities of liquid waste oil as a function of ship type and ship size category (NMD, 1994).
Ship size [GT]
Liquid oily waste , -
CT LGT OT B C GC R RO P OSV OOA OA
<999 - - 35 35 - 35 35 35 40 10 25 35
1000-4999 75 75 75 75 75 75 75 75 75 20 50 75
5000-9999 80 80 80 80 80 80 80 80 100 30 60 80
10000-24999 80 80 80 80 80 80 80 80 100 30 60 80
25000-49999 80 80 80 80 80 80 - - 100 - - 80
50000-99999 80 80 80 80 - - - - 100 - - 80
>100000 - - 80 80 - - - - - - - 80
4.2.1.6.4 Solid oily waste
Solid oily waste contains materials and parts which are contaminated with oil residues (e.g. oil filters).
Additionally, sediments from fuel tanks comprising of a mixture of oil, rust, sand etc. are included in this
category. Annual solid oily waste can be estimated by equations Εq.21, Εq.22 ( ) data from
Table 32 .
Table 32 Average quantities of solid oily waste as a function of ship type and ship size category (NMD, 1994).
Ship size [GT]
Solid oily waste , -
CT LGT OT B C GC R RO P OSV OOA OA
<999 - - 6 6 - 6 6 6 7 5 4 4
1000-4999 12 12 12 12 12 12 12 12 15 12 10 12
5000-9999 15 15 15 15 15 15 15 15 20 15 12 15
10000-24999 15 15 15 15 15 15 15 15 20 15 12 15
25000-49999 15 15 15 15 15 15 - - 20 - - 15
50000-99999 15 15 15 15 - - - - 20 - - 15
>100000 - - 15 15 - - - - - - - 15
4.2.1.6.5 Wastewater (black and grey water)
Ship produced wastewater consists of the black and grey water. Black water is defined as the sewage
from toilettes and medical facilities, while grey water is the effluent from kitchen, pantries, laundries,
galleys, baths and showers. Wastewater can either be treated on board or stored untreated in
appropriate tanks. According to MARPOL Annex IV, untreated black water can be discharged into the sea
only at distance greater than 12 nm from shore and when ship sails at a minimum speed of four knots.
Treated black water can be discharged anywhere, provided local rules are followed. Untreated grey
water can be discharged overboard at distance greater than 3nm from shore, apart from protected
areas such Alaska and Carribean.
Wastewater quantity can be calculated by equations Εq.21, Εq.22 and Εq.23. However, a more
convenient expression Εq.24 can be derived by simplifying equation Εq.23:
Chapter 4 Life Cycle Inventory (LCI) analysis
53
⌊
( ) ⌋
⌊
( ) ⌋
Εq.24
where
( ) Black water production rate , - in mode m for ship type I and
size category k. ( ) Grey water production rate , - in mode m for ship type I
and size category k.
Average produced quantities of black and grey water per person per day , - are
presented in Table 33. The two flushing technologies (conventional or vacuum system) common in use
today, consume considerably different quantities of water and thus different production rate should be
used (US-EPA, 2008a). On many ships there is a connection between black and grey water systems and
thus, an allocation between the two different systems must be made (HELCOM, 1990). Passenger ships
considered, 65% uses vacuum systems while 35% uses conventional technology.
Table 33 Amounts of black and grey water produced per person per day (DNV, 2002; HELCOM, 1990; NMD, 1994).
Ship type
Wastewater production [litres/person/day]
Black water, Grey water,
Conventional system Vacuum system Conventional system Vacuum system
Passenger/cruise vessels 70 25 230 185
Non passenger vessels 70 25 180 135
4.2.1.6.6 Garbage
Ship-produced garbage is either incinerated or delivered at ports or discharged in sea. MARPOL Annex V
specifies certain criteria for permitting discharge in sea. Paper, food etc. can be discharged only when in
a greater than a minimum distance from shore. In special areas, only food can be discharged in sea and
not within a distance of 12 nm from shore. Garbage quantity is a direct function of total people on board.
Table 34 contains garbage production rates (kg/person/day) which agree with other studies (Lloyd's
Register, 1995).
Table 34 Amounts of garbage produced per person per day (NMD, 1994).
Ship type Dry garbage [kg/person/day]
Food waste [kg/person/day]
Total [kg/person/day]
Passenger/cruise vessels 1.26 0.84 2.10
Non passenger vessels 0.90 0.60 1.50
4.2.1.6.7 Crew and Passenger numbers
Table 35 presents data for the average number of people on board per ship type and size category,
which are needed for estimating the quantities of garbage and black/grey water per ship.
Chapter 4 Life Cycle Inventory (LCI) analysis
54
Table 35 Average numbers of personnel and passengers (DNV, 2002; NMD, 1994).
Ship size [GT]
Number of personnel and passengers
CT LGT OT B C GC R RO P OSV OOA OA
<999 - - 4 6 - 4 5 6 200 8 8 4
1000-4999 11 11 11 11 11 11 11 11 500 10 15 12
5000-9999 20 20 20 18 20 20 19 20 800 15 15 40
10000-24999 25 23 25 25 25 25 25 25 1200 20 20 50
25000-49999 27 25 27 27 25 25 - - 1500 - - 50
50000-99999 30 30 30 30 - - - - 2000 - - 50
>100000 - - 30 30 - - - - - - - 50
4.2.1.7 Slop system – cargo residues from oil tankers
The cargo oil tankers need to be washed at least in two occasions: (a) changing the type of cargo after a
trip, (b) prior to maintenance being carried out on tanks, pipes, or valves either on board or in shipyards.
Normally, the tank washings should be collected and settled in slop tanks. Then, these slops should be
separated via water or oil separator with oil being retained and water disposed of. The amount of this
water is estimated at 4-8% of DWT (Schnitler, 1995) and its contamination with oil residues varies
considerably with cargo type and washing system type. The frequency of tank washings depends mainly
on operating policy and cargo variation.
According to MARPOL Annex I Reg.9., the total quantity of oil discharged into the sea shall not exceed
1/30000 of the total quantity of the particular cargo (1/15000 for tankers built before 1980) (see eq.
Εq.25). The tanker must proceed en route at distance greater than 50 nm from nearest land and not
within a special area. The rate of discharge of oil content shall not exceed 30 litres per nautical mile (see
eq. Εq.26). The final discharged amount (see eq. Εq.27) is the minimum of the two values calculated by
equations Εq.25 and Εq.26.
( )
Εq.25
( ) Εq.26
( ( ) ( )) Εq.27
Table 36 explains the symbols and subscripts of equations Εq.25 and Εq.26 and enlists the tables with
relevant data. Annual cargo transported , - for a ship of size can be estimated from equation
Εq.18, and using the data of Table 17 and assuming a utilization rate , (Johnsen, 2000;
Wijnolst & Wergeland, 1997). An estimation of ship’s speed is provided in Table 37, if no real data are
available. The main assumptions of aforementioned model are: (a) all ships discharge the maximum
allowed quantity and (b) rules are followed in any occasion.
Chapter 4 Life Cycle Inventory (LCI) analysis
55
Table 36 Explanation of symbols and subscripts in equations Εq.25 and Εq.26.
Subscripts
Ship type, * + Table 14
Ship size category Table 14
Symbols
Total annual production, ⁄ - of oil from slop operations for oil tankers of size
-
Annual cargo transported , ⁄ - by an oil tanker of size Εq.18
Maximum cargo fraction allowed to be discharged at sea, i.e. 1/30000 for tankers delivered after 01.01.1980 and 1/15000 for tankers delivered before this date
Specific gravity of oil, on average 0.85 , - -
Operational hours , - at sea Table 17
Average operating speed Table 37
Average factor of mileage conducted in areas where discharge is legal. This factor will be low for small ships engaged in short sea and coastal transportation and bigger for large ship conducting transcontinental voyages.
Table 37
The mean speed of a ship is a critical parameter that is often used as a basis for calculating the average
load of the ship engines and hence, the fuel consumption. However, the mean speed given in Table 37 is
used only for the calculations of this paragraph in the present thesis. The coupling of the speed with the
fuel consumption would introduce complexity that is not desired for life cycle modelling and
assumptions that would increase the uncertainly of the final results. For this reason, the engine load
factor was introduced in paragraph. 4.2.1.1.
Table 37 Average operating speed and fraction of distance travelled within areas where discharge is legal (DNV, 2002).
Ship size [GT]
Mean speed [knots]
Fraction of distance travelled in “legal area”
<999 11 0
1000-4999 13 0
5000-9999 13 0.1
10000-24999 15 0.3
25000-49999 14 0.4
50000-99999 15 0.5
>100000 15 0.5
4.2.1.8 Volatile Organic Compounds (VOCs) from oil tankers
Emissions to the atmosphere occur during the carriage of organic cargo by oil tankers. The emissions are
mainly volatile organic compounds (VOCs) that arise from the evaporation of the volatile compounds of
the cargo being transported. The quantity and composition of the VOCs varies with the cargo type and
Chapter 4 Life Cycle Inventory (LCI) analysis
56
the stage of the tanker’s trip, i.e. loading, transit, unloading and ballast leg. Approximately 0.1% of
payload can be lost and emitted to the atmosphere as VOCs (Ø. Endresen & Sørgård, 1999; European
environment agency, 2000; Martens, 1993).
The loading stage is the most severe with respect to the amount of VOCs released to the atmosphere, as
organic vapours in "empty" cargo tanks are displaced to the atmosphere by the liquid being loaded into
the tanks. These vapours are a composite of (1) residual vapours from previous loads, (2) vapours
generated from the new cargo being loaded, and (3) vapours transferred to the tank in vapour balance
systems. Thus, the quantity of evaporative losses from loading operations depends on the following
parameters (MEPC, 2013; US-EPA, 2008b):
Physical and chemical properties of the previous cargo;
Method of unloading the previous cargo;
Operations to transport the empty carrier to a loading terminal;
Method of loading the new cargo;
Physical and chemical properties of the new cargo.
Significant amount of vapours are also emitted during the transit stage of a trip due to the heating of the
cargo to preserve it at an adequate viscosity and avoid phase separation, and due to the sloshing of the
tanks depending on the sea state(MEPC, 2013).
The large variation of the aforementioned parameters with the numerous cargo types and the
operational pattern of the ship does not allow for a universal modelling of the VOC generation
mechanism within the scope of the present study. Thus, emission factors for the two most severe
carriage stages (loading and transit) are adopted for the six most important cargo types in terms of
annual volume transported. The emission factors can be seen in Table 38:
Table 38 VOC emission factors for crude oil losses during the loading and transit stages (US-EPA, 2008b).
Cargo type Loading
[mg/litre] Transit
[mg/week/litre]
Gasoline 215 320
Crude oil 73 150
Jet Naphtha 60 84
Kerosene 0.63 0.60
Distillate 0.55 0.54
Residual 0.004 0.003
It should be noticed that the quantity of methane and ethane is negligible in VOC emissions of all
products other than crude oil (US-EPA, 2008b). For the crude oil, ethane and methane account for 15%
of the amount emitted.
4.2.2 Maintenance
The maintenance part of the operational phase normally includes all the processes carried out during
the drydockings and the minor repairs needed between the drydocking intervals, i.e. steelwork,
Chapter 4 Life Cycle Inventory (LCI) analysis
57
machinery and equipment replacement, painting and respective surface preparation. However, only
painting and surface preparation operations are considered to have significant environmental impact
(US-EPA, 1997b) and thus are taken into account in the present life cycle modelling. Only the modelling
of the surface preparation (blasting) process will be presented below, since the painting process model
is identical to that described in section 4.1.4 for the shipbuilding phase.
4.2.2.1 Surface Preparation (Blasting)
Surface preparation methods are used to remove impurities such as rust, corrosion, and old coatings
from a substrate and create a profile with better adhesion for new coating. Swedish standards are often
quoted, SA 3 being white metal and SA 2.5 being near white metal. SA 2.5 is common as a reasonably
achievable standard for shipyards (Eyres & Bruce, 2012b).Dry abrasive blasting is the most widely used
method in shipbuilding and ship repair. In this method, an abrasive material is mixed with compressed
air and this mixture is projected onto the surface. Traditionally, sand was used as the abrasive, but it is
being replaced by other materials due to the adverse health and environmental effects of silica dust
(silicosis) (NPI, 1999). Nowadays, common abrasive materials used in shipbuilding are barshot
(hematite), coal slag, copper slag, garnet, steel grit/shot, and specialty sand (Kura, 2005). The process
waste is a mixture of used abrasive, paint chips and eroded material. Total suspended particles, PM10
and metal emission per unit blasted area can be calculated by using the emission factors of Table 39.
Table 39 Total suspended particles (TSP), PM10 and metal emission factors per blasted unit area for six blasting materials (Kura, 2005).
Apart from IHM items, ship’s LDT consists of steel, furniture, machine parts etc. which might not be
hazardous, yet they constitute significant material flows. Several breakdowns of LDT can be found in
literature (DNV, 1999; Melissen & Molemaker, 2005; Sarraf et al., 2010; Srinivasa Reddy, Basha, Sravan
Kumar, Joshi, & Ghosh, 2003; Tilwankar et al., 2008), with most of them based on experience rather
than actual measurements. Table 42 shows the estimations per ship as provided by the Indian
Shipbrokers association.
Table 42 Estimation of lightship displacement breakdown. Derived from (Melissen & Molemaker, 2005) .
Items Oil Tanker Bulk Carrier General Cargo
Recycled Steel 86.50% 83.00% 78.00%
Non-ferrous metals 1.00% 1.00% 1.00%
Machinery 3.00% 4.00% 6.00%
Accommodation equipment 1.50% 3.00% 5.00%
Waste 8.00% 9.00% 10.00%
4.4 Chapter conclusions In this chapter, all the particular processes of the three life cycle phases were modelled and data were
collected in order to calculate the amount of releases to the environment. The current technologies and
materials used for the construction, operation and dismantling of ships were identified and linked to
characteristics of the ship and its operation pattern. The modelling of the processes has two stages:
firstly, calculating the quantities of materials and energy as functions of the product system
Chapter 4 Life Cycle Inventory (LCI) analysis
61
characteristics; then, this input of materials and energy is converted by the process models to the final
environmental releases, or in other words, to the life cycle inventory of the system. The inventory serves
in turn as the input to the impact assessment method. It is converted to the potential environmental
impact with respect to the categories selected in section 3.3.
Problems in data collection were encountered and revealed gaps in the literature. More specifically,
correlations of the superstructure and cargo space areas could not be found, meaning that the impact
related with this data can only be calculated for specific cases. Additionally, no emission factors for oxy-
fuel cutting were located in the literature, despite the fact that it is the main process in ship breaking. It
was also concluded that the hazardous materials found on board and released after ship dismantling are
very different from ship to ship and cannot be easily calculated.
The procedure described above is exemplified in the next chapter, where the inventory of an aframax
tanker is calculated using the models of the present chapter. Its potential environmental impact is
quantified using the ReCipE LCIA method, as well as the contribution of each particular process to the
total result.
Chapter 5 Case study: Inventory analysis and impact assessment of an Aframax tanker
62
5 Case study: Inventory analysis and impact assessment of an Aframax
tanker According to the specifications set in chapter 3 and implementing the models developed in chapter 4, a
computational tool was developed. The tool has a spreadsheet form and was built on the MS Excel
software. In this section, the environmental impact of an existing 115000 DWT oil carrier are calculated
and assessed using the tool. Results concerning the total life cycle are presented, as well as the
contribution of particular processes and life cycle phases.
5.1 Case description The parameters defining the ship and its life cycle are shown in Table 43. The LCA analysis and impact
assessment results are presented hereafter.
Table 43 Main particulars and machinery specification for an 115000 DWT oil tanker.
Parameter Unit Value
Main particulars:
Deadweight tons 115000
Gross tonnage GT 62216
Steel weight tons 12502
LPPxBxDxTmax m 239x44x21x14.8
Crew persons 30
Machinery specs.: Main engines Auxiliary engines
Installed power kW 14342.0 2868.4
Type - slow speed medium speed
Fuel type - residual fuel oil (2.7% sulphur)
marine diesel oil (0.7% sulphur)
SFOC kg/kWh 0.195 0.215
Load factor (sailing) - 0.70 0.25
Load factor (at berth) - 0 0.35
In the next paragraphs the life cycle inventory results and their impact assessment is demonstrated. For
each phase, process key parameters calculated within the model are presented together with the
material outflows and energy consumption.
5.2 Case results In this section, the environmental releases per life cycle phase are explained and their impact
assessment per impact category is illustrated with charts. The results of the total life cycle and the
phases/ processes contributing the most are identified. The full life cycle inventory of the case study can
be found in the appendix B.
Construction phase
The construction of the ship under investigation is assumed to take place in Japan. Therefore, the air
emissions due to electricity consumption are those of the Japanese energy mix. For the ship under
investigation, the total length of welds was 404.6 km, 4531 kg of welding consumables (rod/wires) are
Chapter 5 Case study: Inventory analysis and impact assessment of an Aframax tanker
63
used and the corresponding electricity consumption was 4342 kWh. The carbon dioxide emitted from
the welding shielding gas is estimated to 1618 kg. A total of 195km of various thickness steel is cut, of
which 60% was cut with underwater plasma cutting and the rest with atmospheric plasma cutting. The
electricity consumed due to cutting was 433575 kWh. All internal and external surfaces of the ship were
painted, resulting in the release of 32111 kg non methane VOC in the atmosphere. The overhead
electricity consumption which accounts offices, ventilation, air conditioning etc. was calculated
683042.5 kWh. Table 44 presents the amounts of the materials released directly into the air in the
shipyard area and outside the shipyard due to electricity consumption.
Table 44 Direct (within shipyard) and indirect (electricity consumption) material releases during the construction phase.
Air pollutant Unit Quantity Process
Direct emissions (within yard):
Chromium kg 0.09 welding
Manganese kg 33.43 welding
Nickel kg 0.36 welding
PM10 kg 61.63 welding
CO2 kg 1674.51 welding
PM10 kg 47986.21 cutting
NMVOC kg 32111.33 painting
Indirect (outside the yard):
CO2 kg 177292.406 electricity consumption
CH4 kg 0.633 electricity consumption
N2O kg 2.746 electricity consumption
NOX kg 128.392 electricity consumption
CO kg 36.014 electricity consumption
NMVOC kg 1.704 electricity consumption
SO2 kg 91.933 electricity consumption
Figure 8 shows the impact assessment with the ReCiPe methodology for the releases of the construction
phase. The results are normalized with respect to the total equivalent amount of each category. The
same approach will be followed for all impact assessment diagrams hereafter. The welding process
dominates the ecotoxicity impact categories due to the release of heavy metals in the air. The high
number of particulate matter produced during the cutting process explains its large effect to the
relevant category. The painting, being the only process which produces VOCs prevails at the
photochemical oxidant formation. At the climate change, marine eutrophication and terrestrial
acidification categories, the overhead electricity consumption plays a significant role as expected, as it
produces the highest amount of combustion gases among the other processes.
Chapter 5 Case study: Inventory analysis and impact assessment of an Aframax tanker
64
Figure 8 Share of different construction processes to the total per impact category, for the oil tanker case study.The calculations are based on the ReCiPe method.
Operation phase: Sailing
The ship is considered to operate over a life cycle of 26 years and complete a total number of 260
roundtrip voyages at an average speed of 15 knots. Annually, the ship spent 3600 hours at sailing and
manoeuvring , 4920 hours at berth and 240 hours was off-hire for repairs, inspections etc.
The cargo type is assumed crude oil with 0.851 ton/m3 density and the total amount transported per
year was 861218 tons. For ballast legs, the utilization rate of the ship’s ballast capacity was 89%.
Annually the ship consumed 12245 tons of residual oil (main engines) and 1487 tons of low sulphur
marine diesel oil (auxiliary engines). The degradation of the hull over life time and the periodic effect of
biofouling between the drydockings (see paragraph 4.2.1.2) has been taken into account and resulted in
9.73% increase in fuel consumption). Table 45 shows the air emissions (annual and lifetime) due to the
engines’ combustion.
Table 45 Annual and lifetime air emissions from ship engines' (main and auxiliary) operation.
Air pollutant Unit Annual Lifetime
Carbon monoxide (CO) tons 101.62 2642.14
Non methane VOCs tons 32.96 856.91
Methane (CH4) tons 4.12 107.11
Nitrous oxide (N2O) tons 1.10 28.56
Carbon dioxide (CO2) tons 43072.08 1119874.16
Sulphur dioxide (SO2) tons 676.11 17578.78
Nitrogen oxides (NOX) tons 1030.98 26805.36
Particulate matter (PM10) tons 83.68 2175.63
Triggered by the ship’s operation, three types of material outflows are continually produced: ballast
water, underwater coating leaching, anode corrosion and various wastes. For ballast water, the
0%10%20%30%40%50%60%70%80%90%
100%
Cutting Overhead Painting Welding
Chapter 5 Case study: Inventory analysis and impact assessment of an Aframax tanker
65
cumulative effect over time to certain geographical areas has been proven to cause the appearance of
invasive species (D. Smith et al., 1996). The underwater coating is assumed to be of the antifouling type
and thus cuprous oxides and irgarol are leached into the sea. As far as corrosion protection is concerned,
zinc anodes are considered for the total life time. In the case of oil tankers, an extra pollutant enters the
marine environment due to the cargo tanks washings before cargo type change or before repairs: slop
tanks wash water. The quantities of the aforementioned continually produced pollutants are shown in
Table 46. The litre is used as unit for those wastes that their composition and density are unknown.
Table 46 Lifetime amounts of coating leakage, ballast discharged at sea, slop tank water and wastes produced on board.
Pollutant Unit Annual Lifetime
Ballast water tons 386340 10044843
Cuprous oxide tons 0.36 9.4
Irgarol tons 0.054 1.4
Zinc tons 41.3 1073.9
Slop tank water tons 28.7 746.4
Wastes:
Oil sludge litres 177500 4615000
Bilge water litres 426000 11076000
Liquid oily waste litres 28400 738400
Solid oily waste tons 5.3 138.5
Black water litres 314175 8168550
Grey water tons 2209.9 57456.8
Garbage tons 15.9 415.4
VOCs are emitted to the atmosphere during the cargo loading, unloading and transit of oil/product
carriers. Table 47 shows the total VOCs produced and their breakdown to methane, ethane and non-
methane VOCs.
Table 47 Annual and lifetime quantity of VOCs emitted during cargo handling operations of an oil tanker.
Unit Annual Lifetime
Total VOC amount tons 5233.5 136070.4
Breakdown:
Ethane (C2H6) tons 461.0 11987.2
Methane (CH4) tons 324.0 8423.4
NMVOCs tons 4448.5 115659.8
The environmental impact assessment of the pollutants is conducted by applying the ReCiPe method
(see chapter 3.3.1). Three problems were met:
The composition of the wastes discharged into the sea is unknown and their impact could not be
assessed.
The introduction of invasive species due to the ballast water transportation has not been
modelled yet and thus it is not included in ReCiPe.
Chapter 5 Case study: Inventory analysis and impact assessment of an Aframax tanker
66
No characterization factors in certain impact categories are provided for some pollutants (e.g.
VOCs are not assessed at the ecotoxicity categories, although it is a reference unit in
photochemical oxidant formation category)
Thus, the impact assessment step of the LCA method cannot be carried out for the pollutants of the
above list, leading to a biased assessment to the ecotoxicity categories. Figure 9 illustrates the impact
assessment of the sailing part of the operation phase.
Figure 9 Share of the different processes of the sailing phase to the total impact of that phase. Calculations are based on ReCiPe method.
The processes of the sailing part of the operation phase have impact in all categories apart from the
ozone depletion. This is due to the fact that ozone depletion is mainly affected by the fugitive emissions
of refrigerants which were not taken into account in the inventory analysis. The thesis scope does not
include ship types such as reefers and fishing vessels which use considerable amounts of refrigerants for
their operation.
Operation phase: Maintenance
The repair stage includes two processes (painting and blasting) which are performed with different
frequency on ship’s surfaces during her life cycle, depending on the condition of the ship and the policy
of the operating company. In our case, the hull’s bottom, topsides and cargo tanks were blasted
completely and repainted every 5 years. The weatherdeck, superstructure and ballast tank surfaces
were blasted and repainted every 10 years.
The total emissions due to the repair painting over the life cycle are 95.7 tons of non-methane VOCs.
Blasting with steel grit as abrasive material accounts for 463.2 tons of total suspended particles, of
which 463.2 kg are particulate matter with diameter less than 10μm. The air compressors used at
blasting are equipped with high speed diesel engines producing the following air emissions over life
cycle:
0%10%20%30%40%50%60%70%80%90%
100%
Air Emission Anodes Paint leaching VOC_tankers
Chapter 5 Case study: Inventory analysis and impact assessment of an Aframax tanker
67
Figure 10 Air emissions from diesel oil combustion air compressors for blasting.
Air pollutant Unit Lifetime
Carbon monoxide (CO) kg 208.6
Non methane VOCs kg 67.7
Methane (CH4) kg 8.46
Nitrous oxide (N2O) kg 2.26
Carbon dioxide (CO2) kg 89941.9
Sulphur dioxide (SO2) kg 282.0
Nitrogen oxides (NOX) kg 1691.7
Particulate matter (PM10) kg 31.0
The contribution of the two processes of the repair stage are shown in Figure 11. Non methane VOCs
from the painting processes contribute only to the photochemical oxidant formation category. Releases
from blasting contribute to four categories due to the particulate matter speciation and the air
emissions from the air compressors diesel fuel compustion.
Figure 11 Share of the blasting and painting processes to the total impact of the repair phase. Calculations are based on ReCiPe method.
It should be noted again that the appearance of zero impact on the ecotoxicity categories due to the
maintenance properties is not actually true. The lack of speciation profile for the painting NMVOCs
hinders the assessment for these categories.
Shipbreaking phase
The shipbreaking is assumed to be conducted with the practices described in paragraph 0. The
environmental impact comes from two processes: the air emissions from the acetylene cutting torches
and the releases to the environment from the untreated/unrecycled materials abandoned on the beach
or at the unformal waste disposal sites. Table 48 presents part of the materials that can be found on
board prior to demolition and their fate after the ship is completely dismantled. The total carbon dioxide
coming from the acetylene torches was estimated at 1960 tons.
0%10%20%30%40%50%60%70%80%90%
100%
Blasting Painting
Chapter 5 Case study: Inventory analysis and impact assessment of an Aframax tanker
68
Table 48 Materials quantities before and their fate after ship dismantling.
Apart from the PCBs and the heavy metals releases to the environment, the impact of the rest of the
materials could not be assessed due to the lack of characterisation factor (e.g. asbestos) or the lack of
their composition. The results of the assessment are shown in Figure 12. Two processes are shown
(torch cutting and dumbing), since they are the only having impact within the boundary of the product
system, i.e. the fence line of the ship breaking yard.
Figure 12 Share of the two shipbreaking processes to the total of the phase. Cutting accounts only for CO2 release and thus affects the climate change category solely. Calculations are based on ReCiPe method.
0%10%20%30%40%50%60%70%80%90%
100%
Cutting Dumping
Chapter 5 Case study: Inventory analysis and impact assessment of an Aframax tanker
69
Aggregated results
The impact assessment of all the pollutants produced during the life cycle is present in Figure 13 per
phase:
Figure 13 Share of the three life cycle phases to the total per impact category, for the oil tanker case study.The calculations are based on the ReCiPe method.
It can be seen that the operational phase is dominant at the categories of climate change, marine
eutrophication, particle matter formation, photochemical oxidant formation and terrestrial acidification,
mainly due to the air emissions produced by the ship’s engines during the life. Shipbreaking prevails at
terrestrial ecotoxicity and ozone depletion due to the heavy metals that remain onto the soil and the
ozone depleting substances that are released in the air respectively. The release of heavy metals is also
the reason that the shipbreaking and construction phases dominate the ecotoxicity categories. More
specifically, the heavy metals that contaminate the soil at the scrapping are responsible for human and
freshwater toxicity while the direct releases from the welding process at the shipyard are responsible for
marine ecotoxicity.
0%10%20%30%40%50%60%70%80%90%
100%
Construction Operation Scrapping
Chapter 6 Conclusions
70
6 Conclusions The target of this thesis was to assess the environmental impacts caused by the pollutants generated
during a ship’s life cycle using the LCA methodology. Relevant modelling concepts and LCA approaches
were investigated to determine the most appropriate one for applying the LCA method in the maritime
context. Correspondingly, the life cycle was divided in three phases (shipbuilding, operation and
dismantling) that were recognised as those directly related to the ship as product system and include
the foreground processes that the shipowner/shipbuilder can control or decisively influence. In line with
initial thesis objectives, a record of the total pollutants was created and their quantities were
parameterised with respect to ship parameters. Thus, the crucial parameters that control certain
environmental consequences were revealed.
The environmental impact from the ship construction phase was found to be strongly influenced by the
ship’s size, as this defines the key parameters for main processes conducted within the shipyard. For the
welding and steel cutting operations, the ship size in terms of lightship steel weight defines the total
weld material consumed and the total cut length. The welding emissions affect mainly the toxicity
impact categories (human, marine, freshwater and terrestrial), while cutting processes generate mainly
for the AMVER vessels: provided by Carroll, E., US Coast Guard. Babcock, R. E. (2004). Data for design of vapor recovery units for crude oils stock tank emissions. BAL. (2013). BAL.LCPA Retrieved 11/10, 2013, from http://lcpa.bal.eu/L3C/index.xhtml Bare, J. C. (2002). TRACI, The Tool for the Reduction and Assessment of Chemical and Other
Environmental Impacts. Journal of Industrial Ecology, 6(3-4), 49-78. doi: 10.1162/108819802766269539
Baumann, H., & Tillman, A.-M. (2004). The Hitch Hiker's Guide to LCA. An orientation in life cycle assessment methodology and application: External organization.
Carlton, J. T., Reid, D. M., & Leeuwen, H. (1995). The Role of Shipping in the Introduction of Nonindigenous Aquatic Organisms to the Coastal Waters of the United States (other than the Great Lakes) and an Analysis of Control Options Shipping Study I (pp. 75): USCG
Carpenter, A., & Macgill, S. M. (2005). The EU Directive on port reception facilities for ship-generated waste and cargo residues: The results of a second survey on the provision and uptake of facilities in North Sea ports. Marine Pollution Bulletin, 50(12), 1541-1547. doi: http://dx.doi.org/10.1016/j.marpolbul.2005.06.021
CCPA. (2008). Guidline for quantifying emissions from chemical facilities: Canadian Chemical Producers' Association.
Celebi, U. B., & Vardar, N. (2008). Investigation of VOC emissions from indoor and outdoor painting processes in shipyards. Atmospheric Environment, 42(22), 5685-5695. doi: http://dx.doi.org/10.1016/j.atmosenv.2008.03.003
Chad Hewitt, M. C., Ashley Coutts, Nick Rawlinson. (2011). Short Report: Vessel Biofouling Risk Assessment: The Department of Agriculture, Fisheries & Forestry (DAFF).
Chatzinikolaou, S., & Ventikos, N. P. (2008). Hazardous waste management and ship recycling: friends or foe? HazWasteManagement, A5(1).
Cotteleer, A. (2012). Coating emissions of sea shipping for 2010. Netherlands continental shelf,dutch port areas and OSPAR region II: MARIN.
Demaria, F. (2010). Shipbreaking at Alang–Sosiya (India): An ecological distribution conflict. Ecological Economics, 70(2), 250-260. doi: http://dx.doi.org/10.1016/j.ecolecon.2010.09.006
DNV. (1999). Decommissioning of ships: Environmental Protection and Ship Demolition Practices. In D. N. Veritas (Ed.).
DNV. (2000). Decommissioning of Ships: Environmental Standards, Ship-Breaking Practicies, On-Site Assessment: Bangladesh–Chittagong. In Det Norske Veritas (Ed.).
DNV. (2002). Det Norske Veritas, Internal databases, containing information on DNV classed vessels. Dua, S. K., Aponte, C. A., Srinivasamurthy, A. K., Rupani, K., Allen, M., Rose, R. W., & Ebadian, M. A.
(2000). Size distribution and rate of production of airborne particulate matter generated during the cutting of metals. Paper presented at the International Decommissioning Symposium.
Dutch national waterboard. (2008a). Antifoulants in marine coatings, merchant shipping and fisheries (Version dated June 2008 ed.): Netherlands national water board - Water unit in cooperation with Deltares and TNO.
Dutch national waterboard. (2008b). Sacrificial anodes, merchant shipping and fisheries: Netherlands national water board - Water unit in cooperation with Deltares and TNO.
EC-JRC-IES. (2010a). International Reference Life Cycle Data System (ILCD) Handbook—Framework and Requirements for Life Cycle Impact Assessment Models and Indicators. Joint Research Centre—
Institute for Environment and Sustainability. Publications Office of the European Union, Luxembourg.
EC-JRC-IES. (2010b). International Reference Life Cycle Data System (ILCD) Handbook—general guide for life cycle assessment—detailed guidance. Joint Research Centre—Institute for Environment and Sustainability. Publications Office of the European Union, Luxembourg.
EC-JRC-IES. (2010c). International Reference Life Cycle Data System (ILCD) Handbook: Analysing of existing Environmental Impact Assessment methodologies for use in Life Cycle Assessment. Joint Research Centre—Institute for Environment and Sustainability. Publications Office of the European Union, Luxembourg.
Endresen, Ø., & Sørgård, E. (1999). Reference values for ship pollution: Det Norske Veritas. Endresen, Ø., Sørgård, E., Sundet, J. K., Dalsøren, S. B., Isaksen, I. S. A., Berglen, T. F., & Gravir, G. (2003).
Emission from international sea transportation and environmental impact. Journal of Geophysical Research: Atmospheres, 108(D17), 4560. doi: 10.1029/2002jd002898
ENIRAM. (2012). Study of hull fouling on cruise vessels across various seas. EPE. (2003). Activity A: Collection and treatment of solid and liquid wastes. In REMPEC (Ed.), Port
reception facilities for collecting ship-generated garbage, bilge waters and oily wastes: Environmental protection engineering S.A.
European Commission. (2009). Strategic goals and recommendations for the EU’s maritime transport policy until 2018 COM(2009) 8 final.
Eyres, D. J., & Bruce, G. J. (2012a). 9 - Welding and cutting processes used in shipbuilding Ship Construction (Seventh Edition) (pp. 81-101). Oxford: Butterworth-Heinemann.
Eyres, D. J., & Bruce, G. J. (2012b). 13 - Plate and section preparation and machining Ship Construction (Seventh Edition) (pp. 135-146). Oxford: Butterworth-Heinemann.
FIDH. (2002). Where do the 'floating dustbins' end up? Labour Rights in Shipbreaking Yards in South Asia The cases of Chittagong (Bangladesh) and Alang (India). (Vol. 2). Paris, France: International Federation for Human Rights.
Frey, R. S. (2013). Breaking Ships in the World-System: An Analysis of Two Ship Breaking Capitals, Alang India and Chittagong, Bangladesh. CSSJ Working Papers, 13(1).
Goedkoop, M., Heijungs, R., Huijbregts, M., de Schryver, A., Struijs, J., & van Zelm, R. (2009). ReCiPe 2008. A life cycle impact assessment method which comprises harmonised category indicators at the midpoint and the endpoint level. (Vol. 2013). The Hague, The Netherlands: VROM.
Goedkoop, M., Heijungs, R., Huijbregts, M., Schryver, A. D., Struijs, J., & Zelm, R. v. (2013). ReCiPe Retrieved 21/11/2013, 2013, from http://www.lcia-recipe.net/
Goedkoop, M., R. Spriensma. (1999). The Eco-indicator 99. A damage oriented method for Life Cycle Impac Assessment. Methodology Report.: Publikatiereeks produktenbeleid.
Guinée, J. B. G., M.; Heijungs, R.; Huppes, G.; Kleijn, R.; Koning, A. de; Oers, L. van; Wegener Sleeswijk, A.; Suh, S.; Udo de Haes, H.A.; Bruijn, H. de; Duin, R. van; Huijbregts, M.A.J. (2002). Handbook on life cycle assessment. Operational guide to the ISO standards. I: LCA in perspective. IIa: Guide. IIb: Operational annex. III: Scientific background.
Harrington, R. L. (1992). Marine Engineering. Jersey City, NJ: Society of Naval Architects and Marine Engineers.
HELCOM. (1990). HELCOM recommendation 11/10 Guidelines for capacity calculation of sewage system on board passenger ships: HELCOM, Helsinki Convention.
Hempel. (2011). The Coating Reference Handbook. Hess, M., Snuverink, A., Schoof, A., & A.M de Leeuw, A. M. (2004). Activity D: Standard designs for (A)
oily wastes reception, treatment, storage and dewatering facilities and (B) garbage collection,
treatment and disposal facilities. In REMPEC (Ed.), Port reception facilities for collecting ship-generated garbage, bilge water and oily wastes. Hague, the Netherlands: Tebodin B.V.
Hiraoka, K., Kameyama, M., Senda, T., Kiriya, N., Naruse, K., Sakurai, A., . . . Kihara, T. (2001). Application of Life Cycle Assessment to Shipping. Comprehensive Research on Waste Disposal and Recycling of Wastes 2001, 68.61-68.17.
Hossain, K. A., Iqbal, K. S., & Zakaria, N. G. (2010). Ship recycling prospects in Bangladesh. Paper presented at the Proceeding of International Conference on Marine Technology (MATEC2010).
Hossain, M., & Islam, M. M. (2006). Ship Breaking Activities and its Impact on the Coastal Zone of Chittagong, Bangladesh: Towards Sustainable Management: Advocacy & Publication Unit, Young Power in Social Action (YPSA).
Hydrex. (2012). Ship hull coating systems simplified (white paper) Hydrex white papers series (Vol. 4.): Hydrex N.V.
IMO. (2001). International Convention on the Control of Harmful Anti-fouling Systems on Ships. International Paint. (2004). Hull roughness calculator: International Paint Ltd. ISO. (2006a). Environmental management-life cycle assessment-principles and framework International
Organization for Standardization (Vol. 14040:2006). ISO. (2006b). Environmental management—life cycle assessment—requirements and guidelines (Vol.
14044:2006). Jackens, J. (2012). Presentation of Shipyard Welding Data. Paper presented at the National Shipbuilding
Research Program Environmental Panel Meeting, St. Augustine, FL. Jivén, K., Sjöbris, A., Nilsson, M., Ellis, J., Trägårdh, P., & Nordström, M. (2004). LCA-ship, design tool for
energy efficient ships–a life cycle analysis program for ships. Final report, 08-27. Johnsen, T. (2000). Environmental comparison of alternative transport chains for passengers. A case
study.: Det Norske Veritas. Johnsen, T., & Fet, A. M. (1999). Screening life cycle assessment of M/V color festival: Report. Kameyama, M., Hiraoka, K., Sakurai, A., Naruse, T., & Tauchi, H. (2005). Development of LCA Software for
Ships and LCI Analysis based on Acutual Shipbuilding and Operation. Paper presented at the Proc. 6 th Int. Conf. on Ecobalance.
Kameyama, M., Hiraoka,K.,Tauchi,H. (2007). Study on Life Cycle Impact Assessment for Ships. National Maritime Research Institute, Tokyo, Japan.
Kjølberg, S. (2013). High Performance Hull Coatings: Market view. Paper presented at the Maritime Week Americas, Miami, USA.
Klokk, S. N. (1994). The Norwegian Green Ship Program – Low emission Diesel engines. Paper presented at the Advanced study Workshop, Air Pollution from Marine Engines, Athens, Greece.
Kristensen, H. O. (2000). Energy consumption and exhaust emissions for various types of marine transport compared with trucks and private cars. Paper presented at the ENSUS 2000 Conference in Newcastle.
Kura, B. (2005). Residual Risk from Abrasive Blasting Emissions: Particle Size and Metal Speciation. New Orleans, USA: University of New Orelans.
Lloyd's Register. (1995). Marine Exhaust Emissions Research Programme Lloyd's Register Engineering Services. United Kingdom, London: Lloyd's Register of Shipping (LR).
Lloyd’s List. (2002). The shipping movements database Seasearcher. Lloyd's Maritime Information Services (LIMIS), 2002, from http://www.seasearcher.com/
Malherbe, L., & Mandin, C. (2007). VOC emissions during outdoor ship painting and health-risk assessment. Atmospheric Environment, 41(30), 6322-6330.
Martens, O. (1993). Solutions to hydrocarbon gas emission from tankers, Marine System Design and operations. Paper presented at the ICMES 93, Marine System Design and Operation, Hamburg, Germany.
Melissen, P., & Molemaker, R. J. (2005). The Ship Recycling Fund. Financing environmentally sound scrapping and recycling of sea-going ships. Rotterdam: ECORYS
Mener, W. C., Rosen, P. L., Austin, D. M., & Holt, W. S. (1999). Shipyard welding emission factors development National Shipbuilding Research Program.
MEPC. (2013). Mechanisms for VOC formation and estimates of global emissions. Misra, H. (2005). Status of Ship-Breaking Industry in India. The IUP Journal of Managerial Economics, 7(3-
4), 109-116. NIOSH. (1999). Evaluation of Substitute Materials for Silica Sand in Abrasive Blasting. Pittsburgh, USA:
National Institute for Occupational Safety and Health (NIOSH). NMD. (1994). Mottak av avfall fra skip (Reception facilities for garbage from ships). In N. M. Directorate
(Ed.): Norwegian Ministry of Environment (Miljøverndepartementet). NMRI. (2006). LCA site for ship. Retrieved 11/10, 2013, from http://www.nmri.go.jp/env/lca/ NPI. (1999). Emission Estimation Technique Manual for Shipbuilding Repair and Maintenance. Australia. NSFI. (1977). Kartlegging av oljeholdig avfall fra skip med anløp i norske havner (Mapping oily waste
from ships calling at Norwegian port): Norsk Skipsforskningsinstitutt (Norwegian ship research institute).
Oceana. (2008). Report on Contamination by cruise ships. Protecting the World's Oceans. Brussels, Belgium: Oceana.
OSHA. (2001). Ship Breaking Fact Sheet: US Department of Labor:Occupational Safety and Health Administration (OSHA).
PRé Consultants. (n.d.) Retrieved October 12, 2013, from http://www.pre-sustainability.com/ Roh, M.-I., & Lee, K.-Y. (2007). Generation of production material information for a building block and
simulation of block erection for process planning and scheduling in shipbuilding. International Journal of Production Research, 45(20), 4653-4683.
Rudd, H. J., & Hill, N. A. (2001). Measures to Reduce Emissions of VOCs during Loading and Unloading of Ships in the EU.
Salonen, S., Heikkinen, A., & Ilus, T. (2012). Estimating Vessel Emissions and Efficiency – A Data Driven Approach. Liege, Belgium. Paper presented at the COMPIT 2012 - IT Trends in Design and Operation for the Maritime Industry retrieved from
Sarraf, M., Stuer-Lauridsen, F., Dyoulgerov, M., Bloch, R., Wingfiel, S., & Watkinson, R. (2010). The Ship Breaking and Recycling Industry in Bangladesh and Pakistan: World Bank.
Schnitler, P. (1995). Environmental Indexing of ships – operational discharges: DNV. Serageldin, M. (2005). Proposed emission factors for: stainless steels, mild steels, alloys steels. Paper
presented at the Shipbuilding and Ship repair - Residual Risk. Smith, D., Wonharn, M. J., McCann, L. D., Reid, D. M., & Carlton, J. T. (1996). Biological invasions by Non-
indigenous Species in United States Water: Quantifying the Role of Ballast water and Sediments Part I and II Shipping Study II.
Smith, T. (2013). Method for establishing fuel savings. Paper presented at the Hull and Propeller Workshop, London. UK. http://www.bellona.org/filearchive/fil_8_Tristan_Smith_lUCL.pdf
SRIA. (2006), 2013, from http://www.sriaindia.com/home.html Srinivasa Reddy, M., Basha, S., Sravan Kumar, V. G., Joshi, H. V., & Ghosh, P. K. (2003). Quantification
and classification of ship scraping waste at Alang–Sosiya, India. Marine Pollution Bulletin, 46(12), 1609-1614. doi: http://dx.doi.org/10.1016/S0025-326X(03)00329-1
SSD. (2010). Sustainable Ship Design Retrieved June 26, 2013, from http://sustainableshipdesign.com/ Statistics Norway. (2000). Manning of Norwegian registered vessels (Arbeidstakere på norskregistrerte
skip, etter skipets fart, personenes nasjonalitet og kjønn. 1993-1999). Oslo.
Steiner, H., Bach, F. W., Windelberg, D., & Georgi, B. (1988). Aerosol generation during cutting of various materials with plasma, laser and consumable electrode. Journal of Aerosol Science, 19(7), 1381-1384. doi: http://dx.doi.org/10.1016/0021-8502(88)90179-6
The Oil Industry International Exploration & Production Forum, E. P. (1994). Method for estimating atmospheric emissions from E&P operations.
Tilwankar, A., Mahindrakar, A., & Asolekar, S. (2008). Steel recycling resulting from ship dismantling in India: implications for green house gas emissions. Paper presented at the Proceedings of second International Conference on “Dismantling of Obsolete Vessels”, during.
Tincelin, T., Mermier, L., Pierson, Y., Pelerin, E., & Jouanne, G. (2007). A life cycle approach to shipbuilding and ship operation. Stirling Design International and Ship studio.
Townsin, R. L. (2000). Workshop - Calculating the Cost of Marine Surface Roughness on Ship Performance. Paper presented at the WEGEMT School on Marine Coatings at the University of Plymouth, Plymouth, UK.
Townsin, R. L., Byrne, D., Svensen, T.E. and Milne,A. (1986). Fuel Economy due to Improvements in Ships Hull Surface Condition 1976-1986. International Shipbuilding Progress, 33(383).
Trozzi, C. (2010). Update of Emission Estimate Methodology for Maritime Navigation: Techne Consulting. Trozzi, C., & Vaccaro, R. (1998). Methodologies for estimating air pollutant emissions from ships Techn.
report MEET RF98. UN. (2012). Paper presented at the United Nations Conference of Sustainable Development, Rio de
Janeiro. http://www.uncsd2012.org/index.php?menu=62 UNFCCC. (2010). National Reports on implementation of the Conference of the Parties (COP), from
http://unfccc.int/national_reports/items/1408.php Unknown. from http://www4.hcmut.edu.vn/~dantn/WT/ US-EPA. (1995). Electric Arc Welding. AP - 42: Compilation of Air Pollutant Emission Factors, Volume I:
Stationary Point and Area Sources 5th Edition(Section 12.19). US-EPA. (1997a). Abrasive Blasting. AP - 42: Compilation of Air Pollutant Emission Factors, Volume I:
Stationary Point and Area Sources 5th Edition(Section 13.2.6). US-EPA. (1997b). EPA Office of Compliance Sector Notebook Project: Profile of the shipbuilding and
repair industry (Vol. 26): Office of Compliance, Office of Enforcement and Compliance Assurance, US Environmental Protection Agency.
US-EPA. (2006). Section 7.1: Organic liquid storage tanks AP - 42 (Vol. I). US-EPA. (2008a). Cruise Ship Discharge Assessment Report. US-EPA. (2008b). Section 5.2: Transportation and Marketing of Petroleum Liquids AP- 42 (Vol. I). Wallentin, B. (2012). The illusion of fuel savings - an urgent need for improved transparency in marine
hull coatings. Norway: JOTUN Coatings. Weisman, C., & Kearns, W. (2001). Welding handbook (9th ed. Vol. 2): American Welding Society. Wenzel, H., Hauschild, M. Z., & Alting, L. (2000). Environmental Assessment of Products: Volume 1:
Methodology, tools and case studies in product development (Vol. 1): Springer. Wijnolst, N., & Wergeland, T. (1997). Shipping: Delft University Press. Willems, M., Schouten, M., & Heidbuurt, P. (2003). Uitloging van anodemateriaal van zeeschepen: BECO
Groep BV. Willsher, J. (2008). The Effect of Biocide Free Foul Release Systems on Vessel Performance. London, UK:
International Paint Ltd. Wolterink, J. W. K., Hess, M., Schoof, A., & Wijnen, J. W. (2004). Activity B: Optimum solutions for
collection, treatment and disposal of relevant ship generated solid and liquid wastes, bilge waters and oily wastes in the Mediterranean. In REMPEC (Ed.), Port reception facilities for collecting ship-generated garbage, bilge waters and oily wastes. Hague, the Netherlands: Tebodin B.V.