The Conceptual Design of A Graving Dock FinalReport2.0 Thesis
The Conceptual Design of A Graving Dock FinalReport2.0 Thesis
5/1/2023
graving dock
Using Life Cycle Analysis to reduce the
carbon footprint of a graving dock for
Damen Harlingen
Johannes Idsinga
4719646
Image cover page: aquatech diving website, source: https://www.aquatech-diving.com/nieuwbouw-civiel/
i
The conceptual design of a graving
dock
Using Life Cycle Analysis to reduce the carbon footprint
of a graving dock for Damen Harlingen
by
Johannes Idsinga
Student number: 4719646
ii
Preface
This report present the final thesis for my Master Hydraulic Engineering at the Delft University
of Technology. The topic of this report is the conceptual design of a new graving dock for
Damen Shiprepair and Conversion B.V. in Harlingen, where the sustainability through
minimization of the carbon footprint of the dock is the main focus.
I would like to thank my thesis committee for guiding me throughout this project and providing
me with useful feedback and advice during the process of writing this report. A special thanks
to Jan Jacob Altenburg, Gert de Vries and the other colleagues at ADONIN B.V. for welcoming
me at their office where I could spend time on working on my thesis and have interesting
discussions which helped me out tremendously. I would like to thank Frank Seinen of Damen
Harlingen for allowing me to have such a broad and interesting thesis topic and his help
throughout the process. Finally I would like to thank my girlfriend, friends and family for
supporting me, with a special thanks to my dad for his endless support, ideas and enthusiasm
during the past months and beyond.
Johannes Idsinga
Bolsward, March 2023
iii
Summary
In order to meet the increasing demand for a larger capacity, Damen Shiprepair and
Conversion B.V. in Harlingen is considering to construct a graving dock near its current
shipyard. This thesis aims to facilitate this process by developing a conceptual design of the
graving dock, where the minimisation of the environmental impact of the whole life-cycle of the
dock is crucial due to the vicinity of Natura-2000 areas and the increased importance of
sustainability in the construction sector. The design should fit in the current surroundings while
all the specified requirements regarding the functionality and structural integrity of the dock are
still met.
Between the various types of dock it soon became clear that the dock in question should be a
graving dock with a hall on top. The scope of this report is determined in such a way that the
focus will not be on the design of the hall structure but the report substantiates this decision by
pointing out the benefits that a hall cover where the temperature can potentially be regulated
offers for the quality and efficiency of the ship maintenance works, but also for the structural
integrity of the dock floor. The location-specific challenges that are presented include the
sedimentation of the dock chamber that is expected to be caused by tugboats that stir up the
bottom of the harbour, leading to an inflow of sediment into the dock chamber because of the
pressure gradient that is created when a vessel enters or leaves the dock. Mitigation measures
to prevent this inflow of sediment are presented in order to prevent the cleaning and disposal
of the sediment. Another sustainable design alternative that is included focuses on the
optimization of the dock floor package by adding a structural function to the underwater
concrete floor (UCF) layer that traditionally only has a water retaining function by the addition
of fibres to the concrete mixture.
Firstly, a base design is made that includes this traditional floor build-up and does not include
any sedimentation reduction strategies. In this way, the dimensions of the base elements of
the dock such as the retaining walls, tension piles and dock gate are designed and the effects
of applying the sustainable design alternatives can be determined. The Terms of Reference
summarizes the main requirements for the graving dock, that needs to be able to dock a design
vessel with a length of 135 meters, width of 21.5 meters and a draught of 7 meters. Including
sufficient tolerances on either side of the vessel to perform the maintenance works, the total
dimensions of the dock chamber have been set at a length of 150 meters, width of 30 meters
and depth of 11.8 meters. The ship-carrying block height must be 1.8 meters and the
dimension of the covering hall are a minimum length of 155 meters, width of 50 meters and a
height of 40 meters. The requirement for the tension piles of the dock foundation is that they
must reach the Pleistocene soil layer in order to prevent pile settlement over time that has
been known to cause damage nearby.
Eventually, the dock chamber structure consists of a combi-wall profile with GEWI63.5 anchors
and tension piles that are installed at a grid of 2.5 by 2.4 meters. The UCF of the base design
has a thickness of 1 meter and structural floor with a thickness of 0.5 meters is placed above
it. The design gate has a width of 30 meters, a height of 12.8 meters and weighs 270 tons.
Using this base design, the environmental hotspots are determined using a life cycle analysis
(LCA) and clearly identify the amount of reinforced concrete and the sediment removal process
as the two ‘hotspots’ in the total carbon footprint, for which sustainable alternatives are
developed. Three design variants are initially created, which each differ on three design
aspects: the build-up of the UCF package, the sedimentation reduction strategy and the gate
operation type. Variant A has a 800 millimetre thick steel fibre reinforced UCF integrated with
a 330 millimetre thick structural floor, saving a considerable amount of concrete compared to
the base design. Variant B contains basalt fibres instead of steel fibres, the total thickness of
these two variants are equal. Variant C has a 850 millimetre thick reinforced UCF with
traditional rebar reinforcements. In the design of these floor variants, it has found that the
influence of the thermal load, caused by fluctuations in temperature throughout the year, will
lead to significant stresses in the floor and the temperature decrease during winter can lead to
cracks in the top fibre of the concrete floor, which further substantiates the idea of placing a
iv
hall on top of the dock where the temperature can be regulated using excess heat from the
waste incineration plant since this will increase the durability of the dock floor.
As far as the sediment reduction strategies are considered, variant A contains a gated inlet on
the opposite side of the chamber that aims to take away the pressure gradient during docking,
Variant B has a bubble screen that filters the sediment out of the inflowing water and Variant
C aims to make the use of sediment-stirring tugboats redundant by using winches to dock the
vessel. Variant A contains a caisson gate, Variant B combines the bubble screen with a sliding
gate that is powered by the same hydro jet installation and Variant C has a mitre gate system.
The reduction in carbon footprint is first assessed by performing the LCA, where the different
design aspects are also assessed separately. This eventually leads to another design Variant
D, which is a combination of the most sustainable aspects, namely a basalt fibre reinforced
UCF with the gated inlet and caisson gate. In this configuration the total carbon footprint of the
dock is reduced by nearly 50% compared to the base design, amounting to roughly 12.500
tons of CO2-equivalent emissions. The required cost of investment and expected revenues are
determined next, where investments of roughly 7.6 million euros and the return of investment
is 20 years are found on average. The results of the LCA and cost benefit analysis (CBA) for
each design variant are then used to perform a multi criteria analysis (MCA), that also take the
ease of operation, ease of maintenance and constructability of the design variants into account.
The conclusion of this MCA was that variant D gives the optimal combination of strengths of
the different design elements, since the basalt fibre reinforced UCF leads to an optimal
combination of construction costs and emission reduction benefits, the use of the gated inlet
allows for the most swift and easy operation of the dock and maintenance can easily be
performed to the caisson gate. A more general conclusion is that the LCA and Cost Benefit
Analysis also proved that sediment reduction strategies should always be a point of emphasis
in the design of a graving dock, since it targets one of the main contributors to the carbon
footprint, saves a considerable amount of operational costs each year and allows for a faster
docking procedure.
v
Table of contents
Preface .................................................................................................................................. iii
Summary ............................................................................................................................... iv
Chapter 1: Introduction .......................................................................................................... 1
Chapter 2: Literature review .................................................................................................. 3
Chapter 2.1: Dock types .................................................................................................... 3
Chapter 2.2: Location-specific effects ................................................................................ 5
Section 2.2.1: Siltation of the dock ................................................................................. 5
Chapter 2.2.2: Nitrogen emissions ................................................................................. 6
Chapter 2.3: Main dock components .................................................................................. 7
Section 2.3.1: Dock gates............................................................................................... 7
Section 2.3.2: Filling and Emptying system .................................................................... 8
Section 2.3.3: Dock head structure ................................................................................. 9
Section 2.3.4: Dock ship chamber structure ................................................................... 9
Chapter 3: Environmental impact of a graving dock ..............................................................11
Chapter 3.1: Introduction of sustainability .........................................................................11
Chapter 3.2: Life Cycle Assessment .................................................................................14
Chapter 3.3: Emission sources and sustainable alternatives.............................................14
Chapter 4: Functional Analysis & Programme of Requirements ...........................................16
Chapter 4.1: Wishes and requirements .............................................................................16
Section 4.1.1: Functional requirements .........................................................................16
Section 4.1.2: Specific requirements and wishes ...........................................................17
Section 4.1.3: List of Requirements and visualizations ..................................................17
Chapter 4.2: Boundary conditions .....................................................................................18
Section 4.2.1: Natural boundary conditions ...................................................................19
Section 4.2.2: Field boundary conditions and cables and pipelines ...............................23
Chapter 4.3: Stakeholder Analysis ....................................................................................24
Chapter 5: Development of design alternatives ....................................................................25
Chapter 5.1: Base design .................................................................................................25
Section 5.1.1: Retaining walls........................................................................................25
Section 5.1.2: UCF and Tension Piles ...........................................................................28
Section 5.1.3: Gate design ............................................................................................29
Chapter 5.2: Design alternatives .......................................................................................30
Section 5.2.1: Dock sedimentation reduction strategies.................................................30
Section 5.2.2: UCF alternatives .....................................................................................32
Section 5.2.3: Dock gate mode of operation ..................................................................36
Section 5.2.4: Inlet system ............................................................................................36
Section 5.2.5: Entrance ramp vs overhead cranes ........................................................36
Section 5.2.6: Design variants .......................................................................................37
vi
Chapter 5.3: Design alternatives specification...................................................................38
Section 5.3.1: UCF design finalization ...........................................................................38
Section 5.3.2: Sedimentation reduction measures .........................................................43
Chapter 6: Evaluation of alternatives and selection ..............................................................44
Chapter 6.1: Life Cycle Analysis .......................................................................................44
Section 6.1.1: Life Cycle Inventory ................................................................................45
Section 6.1.2: Impact assessment and results ...............................................................46
Section 6.1.3: End-of-life phase.....................................................................................49
Section 6.1.4: Conclusions ............................................................................................49
Chapter 6.2: Cost-Benefit Analysis ...................................................................................50
Section 6.2.1: Construction costs ..................................................................................50
Section 6.2.2: Operational & maintenance costs ...........................................................52
Section 6.2.3: Benefits – docking revenues ...................................................................54
Section 6.2.4: Cost-benefit development over time........................................................54
Chapter 6.3: Multi Criteria Analysis ...................................................................................56
Chapter 7: Design Optimization ............................................................................................58
Chapter 7.1: Remaining dock elements ............................................................................58
Chapter 7.2: Concluding sketches ....................................................................................60
Chapter 8: Conclusion, discussion and recommendations ....................................................62
Chapter 8.1: Conclusions ..................................................................................................62
Chapter 8.2: Discussion ....................................................................................................63
References ...........................................................................................................................66
Appendix A: Design vessels .................................................................................................71
Appendix B: Boundary conditions .........................................................................................74
Appendix C: Base design calculations ..................................................................................83
Appendix D: Development of design alternatives................................................................100
Appendix E: UCF loads ......................................................................................................105
Appendix F: UCF Design calculations ................................................................................110
Appendix G: Sediment reduction strategies ........................................................................139
Appendix H: LCA details.....................................................................................................142
Appendix I: Cost-Benefit analysis details ............................................................................156
Appendix J: Multi Criteria Analysis details ..........................................................................166
Appendix K: Dock operations at the current shipyard .........................................................169
Appendix L: Location-specific design aspects ....................................................................170
Appendix M: Dock components specification......................................................................173
Appendix N: Sustainable design alternative details ............................................................175
Appendix O: Dock design requirements..............................................................................178
Appendix P: Stakeholder analysis ......................................................................................180
Appendix Q: Dock head design ..........................................................................................183
vii
Chapter 1: Introduction
In view of the expected development in the maritime industry regarding vessel dimensions
and the increasing demand for capacity, Damen Shiprepair and Conversion B.V. in Harlingen
potentially wants to construct a graving dock. A graving dock is a permanent dock where
construction, repair and maintenance of large vessels can be carried out. After the vessel is
manoeuvred inside the dock and placed on blocks, the dock doors will close and the water is
pumped out of the dock, exposing the vessel and allowing the maintenance or repair works.
The new dock will be constructed on a large piece of unused land next to a waste incineration
plant, near the current Shipyard in Harlingen. The new dock is to be constructed on a large
piece of unused land next to a waste incineration plant, near the current Shipyard in Harlingen.
An overview of the potential construction site is shown in Figure 1 below, which also shows
the location of the shipyard and the waste incineration plant. It has to be noted that the red
square doesn’t represent the dimensions of the graving dock but merely indicates the global
position of the building location.
In order to reduce global warming, the Netherlands aims to cut down its CO2 emissions by
2030 by 49% as determined by the Paris Climate Agreement in 2015 (Rijksoverheid, 2022). In
combination with the current nitrogen problems in the Netherlands, where an excess of
nitrogen emission leads to deterioration of the flora and fauna, the need for an increased
sustainability and subsequently the reduction of environmental impact during construction and
operation has become essential for the civil engineering industry (PH bouwadvies, 2022).
Therefore, the main objective of this thesis is to develop a conceptual design of the graving
dock, where the minimization of the environmental impact of the whole life-cycle (construction,
operation and end-of-life) of the dock is key. The design should fit in the current surroundings
while all the specified requirements regarding the functionality and structural integrity of the
dock will still have to be met.
1
In order to do so, firstly a literature review will be conducted to investigate all the technical
aspects that play a role in the design, construction and operation of a graving dock. Location
specific factors such as the sedimentation rate of the dock are researched as well as the
structural components that make up a dock. Afterwards, the environmental impact of a graving
dock is discussed and the possible reduction strategies are treated. Conclusions from these
chapter will be taken into account when setting up the Terms of Reference that serves as a
basis for the design and includes all design wishes and requirements given by the client. The
boundary conditions such as soil characteristics, water levels and groundwater levels will be
determined so that a base design can be made. This base design is a conceptual design that
contains the main dock components such as the retaining walls, UCF floor and gate. The
sustainable design alternatives are then applied to the base design and design alternatives
are developed that focus on reducing the carbon footprint of the base design by looking at the
build-up of the floor package, the prevention of inflowing sediment and the different modes of
operation for the dock gate.
These design alternatives are evaluated based on their performance in a Life Cycle Analysis
that assess the carbon footprint of the base design and design variants to firstly identify and
justify the hotspots in the base design and then compare the effects of implementing the
different design alternatives. A multi criteria analysis eventually determines the optimal
combination of sustainable design options by also considering the required investments (by
performing a cost benefit analysis), feasibility and ease of operation.
All in all, this report is build up in the following way. Chapter 2 contains the literature review
focusing on the technical details that are involved with the design of a graving dock. Chapter
3 will dive into the environment footprint of a graving dock, Chapter 4 will describe all the
requirements as specified by the stakeholders in the Terms of Reference. Here, section 4.2
will describe all the relevant boundary conditions and Section 4.3 will conduct a stakeholder
analysis. Chapter 5 then starts the development of design alternatives by making the base
design in chapter 5.1. Next, chapter 5.2 starts the develop of design alternatives which are
further elaborated in chapter 5.3. In chapter 6 these design variants are evaluated based on
their environmental impact (chapter 6.1) and costs (chapter 6.2) and the results are then
cooperated in making a decision to determine the optimal design combination in chapter 6.3.
Finally chapter 7 shows a final overview of the optimal design combination and chapter 8 gives
the conclusions and recommendations.
2
Chapter 2: Literature review
The aim of this literature study is to introduce ship docks by looking at the types of ship docks
that exist, the main components that make up a dock and the desired function. Furthermore,
the desired location of the ship dock is analyzed with regards to the external factors that will
play a role in the design, the following sub questions are aimed to be answered by this chapter:
• What types of docks exist and which type will form the basis for the design?
• Which location-specific factors will play a role in design of the dock?
• What are the main structural components of a dock?
Answering these questions enables a good preparation for the next phases of the project.
Figure 2: The floating dock before (right) and after (left) the vessel has entered the dock
3
The other main dock type is a graving dock or permanent dry dock. It is constructed on land
and consists of a concrete and steel structure that includes a (re)movable gate, floor and walls.
During operation, the vessel is maneuvered into the filled dock and the gates are closed as
soon as the vessel is maneuvered on the blocks in the correct position. Pumps are then used
to empty the dock and expose the vessel as a result, allowing the maintenance of construction
works to take place. Graving docks can also be used for the construction of vessels, as is the
case with the graving dock at Icon Yachts in Harlingen. When the gate of the Icon Yachts
needs to be opened, it is completely removed by using the cranes that are present at the dock.
The pictures below show the dock gate and the rails at the top of the gate that stick out so that
the cranes can be moved in the adequate position to remove the gate. The gates of the new
dock are expected to be opened more frequently and other solutions for opening the gate will
also need to be explored but the option of removing the gate is interesting nonetheless.
Furthermore, the graving dock can be split into section by the placement of the intermediate
gate, which can also be found on the picture below.
Figure 3: Icon yacht’s graving dock, including the intermediate caisson gate (left), the crane rails sticking out of the
dock to facilitate removal of the gate (top) and the dock gate seen from inside the dock (bottom)
Advantages of a graving dock include it’s long life span and larger capacity. The supply of
equipment, personnel and machinery is a lot easier due to the location of the dock near the
land and the possibility of placing one or more ramps facilitates this.
On the contrary, the initial construction costs are high and the maintenance costs of the graving
dock can become high as the lifetime of the dock increases. Typical docks can have lifetimes
of up to 100 to even 200 years and to stay functional, maintenance costs might increase
significantly. Furthermore, the presence of a movable gate brings the risk of jamming,
mechanical issues or other problems which would make the complete dock non-operational
(Marine Insight, 2021).
It is possible to construct a large hall that covers the graving dock which can ensure a controlled
environment for performing the maintenance works. The emissions that might occur during the
works, such as dust, sand or paint, is trapped inside the dock and prevents it from directly
being exposed into the environment which reduces the climate impact. The presence of a hall
also adds the possibilities of installing workshops near the dock which prevents the transport
of products to and from the vessel and it opens up the possibilities for placement of overhead
travelling cranes, as is the case at the Icon Yachts dock (British Standard, 1988). On the
contrary, the construction of a hall requires a significant amount of steel which increases the
costs of construction and considering the typically large dimensions of such a hall, the
magnitude of the wind force will subsequently be significantly large as well.
Other alternatives include a shipping lift, which is currently the way of docking vessels at
Damen Harlingen, a closer description of this docking procedure can be found in Appendix K.
This lift can only handle ships with a limited dimension and uses large blocks which can move
over rails that are installed in the floor of the shipyard to position the vessel on the desired
location for the construction works.
4
Since the client’s main wish is to increase the capacity of the shipyard, a graving dock is
preferred over a floating dock which is mainly due to the higher future-proofness of a graving
dock. Not only can larger vessels be handled in a graving dock, which is desirable considering
the development of vessel dimensions in the shipping industry over the last decades (Tran,
2015), but the possibility for placement of an intermediate door(s) allows multiple vessels to
be handled simultaneously. Furthermore, the benefits of a hall on top of the dock with regards
to the reduction of emissions and consequently the environmental footprint during operation of
the dock as well as the opportunities for conducting repair works in a controlled environment
leads to the final decision: the dock that will be designed will be a graving dock with a hall on
top that covers the dock. This decision will be taken into the Terms of Reference in the next
chapter as wish of the client.
Local processes will play a more prominent role in the sedimentation of a dock, for example
the maneuvering of the vessel in the turning circle and entering the dock which will locally
increase the inflow of silt caused by exchange due to turbulence and the resulting velocity
difference. In fact, the majority of the sediment is expected to enter the dock during the
entrance of the ship into the dock chamber and (especially) during the exit of the ship from the
chamber. When the vessel is maneuvered into the chamber with help of a tugboat, the
propeller jet of the tugboat will stir up the layer of sediment at the bottom of the harbor (port
authorities expect a layer of fine sediment or sludge of 0.5 to 1.5 meter at the bottom of the
harbor), increasing the turbulence near the bed and sediment concentration of the water near
the bed. At this moment in time, the dock chamber is completely filled up with water as the
dock gate is opened and the ship is ready to sail into the chamber. As the ship slowly moves
into the port, a body of water with the same volume as the displacement of the ship needs to
leave the chamber, which creates a return flow out of the chamber along both sides and
underneath the vessel that will (partly) wash away the sediment flow but can also leads to the
entrance of sediment together with the vessel. This process is illustrated in figure 4, where a
side view and top view is drawn as well as an indication of the flow lines of the return flow and
the expected location where the increase in turbulence will be.
5
Figure 4: Sedimentation process during docking
When the vessel exits the dock chamber, a flow of water which is equal to the displacement of
the vessel will flow into the chamber, through a kind of suction effect. In this situation the flow
will take the sediment that has been stirred up by the propeller of the tugboat, that pulls the
vessel out, into the dock chamber. This process is illustrated in figure 5 and most of the
sediment inflow is expected to occur in this process.
Possible mitigation measures for the sedimentation of the dock chamber are developed in
Chapter 5.2.
Chapter 2.2.2: Nitrogen emissions
Figure 6: Relative location of Damen shipyard (red dot) with Natura2000 area,
source: (Natura2000, 2022)
As visible on Figure 6, the Damen Harlingen Shipyard, highlighted with the red dot, is located
at close proximity to the Wadden Sea, which is identified as a Natura 2000 area and is listed
under both the Birds Directive and Habitat Directive. This means that active protection
guidelines exist to preserve the Wadden Sea habitat and the species that rely on it. In fact, the
Wadden Sea is mentioned as the “biggest and - from an international perspective – most
important Natura 2000-area in the Netherlands” (Natura2000, 2022).
6
Active protection guidelines are mainly focused on the reduction of emissions, especially CO2
and nitrogen. Especially the latter is a large cause of concern for the Netherlands and in order
to reduce the emissions of these gasses and improve the wellbeing of nature areas in the
Netherlands, the government created the ‘nitrogen law’, called ‘Wetsvoorstel stikstofreductie
en natuurverbetering’, which aims to have the nitrogen level in at least half of the Natura 2000
areas in the country at a safe level by the year 2030 (Rijksoverheid, 2020). This law became
active starting from the 1st of July 2021 and contains an exemption for the building sector, the
so-called ‘Bouwvrijstelling’. This states that the nitrogen emissions during the construction
phase for building projects will not count towards determining the effect of the project on the
Natura 2000 area. This relies on the temporary nature of the construction works and the
relatively small impact of the emissions in the greater picture of the nitrogen problems
(Rijksoverheid, 2021). This statement is supported by Figure 7 where the contribution of each
sector to the total nitrogen emission can be seen and the contribution of the construction sector
is negligibly small (RIVM, 2021).
Several options exist for minimization of the emissions during the construction phase of the
dock but the main focus should be on the operational phase of the dock. The current laws do
not allow for any nitrogen disposal, so 0,00 mol/ha/year during the use phase of the dock
(Bij12, 2022). Since the main source of nitrogen dioxide emissions is the combustion of fuels,
other zero-emission energy sources need to be applied during the use phase of the dock,
appendix L gives more background information regarding this issue.
7
Various dock gate types exist, varying in gate movement, which is often either a translational
or a rotational movement. The most frequently applied gate types are described in Appendix
M, though it has to be noted that these are the most commonly used lock gates, the
requirements for lock gates regarding the water tightness of the door differ from that of a dock
door. The most commonly applied lock gate types are mitre gates that move by a rotation
around the vertical axis, single leaf gate that operate in a similar way as the mitre gates, sector
gates that rotate around either the vertical or horizontal axis, lift gates that use a vertical
translation movement to open, rolling- or sliding gates that uses a horizontal translational
movement during operation of the dock and finally a caisson gate that is completely removed
by cranes or other machinery. Other types exist as well but are less commonly applied in
practice, caused by various reasons such as construction costs or complexity of operation.
Further details such as the advantages and disadvantages of the gate types are treated in
Appendix M, figure 8 shows the operation of a caisson gate for reference.
Figure 8: Operation of a caisson dock, source: (Ravestein Shipyard & Construction Company, 2017)
All in all, the main challenges in design of the lock gate is ensuring a relatively fast and efficient
operation procedure while keeping the watertightness of the gate intact. Even though a certain
leakage discharge is inevitable, this needs to be minimized. Furthermore, the operation of the
opening and closure of the gate must be sufficiently robust, so minimization of the probability
of jamming or mechanical issues is crucial.
Figure 9: Head filling system (left) with valves and energy-breaking structure circled red and the longitudinal system
(right)
8
along the length of the dock, in this way no turbulent zones will be created in the dock. The
large increase in cost of construction is the major downside of this filling and emptying system
(Molenaar, Locks, 2020). Figure 9 shows two types of filling and emptying systems.
For the design of the graving dock, the head filling seems to be the most desirable option
considering the low costs that are involved with this type of filling system that seem to outweigh
the benefits of installing a system with longitudinal culverts. This decision is further supported
by the fact that longitudinal culverts are most often applied in cases of extreme head
differences which will not be the case in Harlingen.
Figure 10: Standard dock head U-shaped profile, source: (Molenaar, Locks, 2020)
9
The floor of the dock chamber needs to transfer the forces of the structure resting on top of it
towards the soil. Another important requirement for the floor is the fact that it needs to be
impermeable so that there is no risk of water flowing into the chamber during the repair works.
A commonly applied floor type is an underwater concrete floor in combination with tension piles
that prevent the uplift of the floor.
The selection of the right combination of walls and floor depends on a number of factors such
as the required dimensions of the chamber, soil and water level conditions, construction
requirements such as the dimensions of the available construction site and the availability of
materials and equipment. Naturally, the costs and environmental impact that is involved in the
construction of the dock chamber play an important role as well.
An interesting option is the use of a foil construction, or foil polder. This method is often applied
in underground infrastructure such as viaducts or tunnels. In principle, a large building pit is
dug with a slope of approximately 1:3. On the bottom of this building pit, a watertight plastic
foil (PVC or HDPE) of a few millimeters thickness is placed. The building pit is filled with ballast
soil and the remaining water is pumped out. In the middle of the building pit, the graving dock
can be constructed. The foil layer is impermeable and the ballast soil ensures vertical
equilibrium of forces. The main advantage of this type of construction is the fact that a large
thick underwater concrete floor is no longer necessary, reducing the costs and climate impact
of the design considerably. However, sufficient space has to be available on both side of the
graving dock, since the slopes need to be constructed initially.
10
Chapter 3: Environmental impact of a
graving dock
In this chapter, the concept of sustainability and the way in which it can be applied in
engineering projects is treated. First of all, the definition of sustainability in design and the
relevant philosophies are introduced in chapter 3.1. Afterwards, the way in which the
sustainability of a design is determined in practice, namely by performing a Life Cycle
Assessment, is introduced in chapter 3.2. Finally, the main sources of emissions during the life
time of a graving dock and potential sustainable alternatives are treated in chapter 3.3.
‘Sustainable development ensures that it meets the needs of the present without compromising
the ability of future generations to meet their own needs’ (Brundtland, 1987)
The term ‘needs’ here describes both the availability of resources which might otherwise
become depleted as well as the presence of a ‘clean environment’ and ‘healthy and prosperous
living conditions’. The use of substances which are harmful to the environment and those that
cannot be replenished or easily removed from the atmosphere should therefore be prevented.
The three main subjects that were addressed by The Brundtland Commission in relation to
sustainable development were: 1. Environmental aspects including limited use of finite
resources and no release of harmful emissions, 2. Social aspects such as equality and
education and 3. Economic growth (Jonkers, 2018).
In this definition, a link can be made to the 17 sustainable development goals which were
defined by the United Nations in 2015 as a ‘call to action to end poverty, protect the planet and
ensure that by 2030 all people enjoy peace and prosperity’ (UNDP, 2022). Similar to the
subjects that were defined by the Brundtland Commission, these SDGs are focused on ‘social,
economic and environmental sustainability’. Figure 11 shows the 17 defined sustainable
development goals (Global Compact Network Netherlands, 2022).
11
Relating these development goals and the subjects that were identified by the Brundtland
Commission, it can be concluded that the realization of a new graving dock would mostly create
opportunities for economic growth. In case Damen would decide not to follow the
developments in the shipping industries by increasing their capacity, clients will choose for a
different shipyard instead, which would not only have consequences for the economic
development of Damen and its employees, but also for the economic development of the
Harlingen municipality and surroundings. Where opportunities are created for economic
growth, the main challenge for a construction project is to follow the sustainable developments
goals that are related to environmental aspects. In civil engineering practices, this sustainable
development means the transition from a linear to a circular building process, meaning the
limited use of finite resources and no release of harmful emissions. Figure 12 shows this
building cycle (Jonkers, 2018).
In an ideal future scenario, this building cycle is completely closed as the input of new raw
materials is barely necessary due to the recycling and reuse of materials and elements of
structures which are at the end of their lifetime. No emissions of harmful compounds occur
anymore and waste is no longer considered waste but serves as a useful resource for new
projects. This scenario can be seen in figure 13 (Jonkers, 2018).
In order to contribute to this circular building cycle, the 9R framework has been developed by
Potting et al in 2017, which states the possible strategies that can be implemented in a
12
construction project. Figure 14 shows the 9R framework (Potting, Hekkert, Worrell, &
Hanemaaijer, 2017).
As can also be read from figure 14, the shorter the loop in the R-framework, meaner the lower
the ‘R’ strategy, the bigger the contribution to an increase in circularity is. For a construction
project, this often means that the earlier these strategies are taken into account during the
decision-making process, the more possibilities there are and the larger the level of impact of
these strategies is. With level of impact, the project scale which is influenced by the strategy
is meant. Looking at the figure 15, we can say that shorter strategy loops correspond to
strategies which influence the structure on a macro level. The complete design of the structure
with all its systems is done while keeping the strategy in mind. On the other hand, the largest
strategy loops might only influence the structure on micro level, meaning the individual
elements of the structure (Geng & Doberstein, 2008).
The shortest loops in the 9R framework are Refuse (R0), Rethink (R1) and Reduce (R2). These
strategies aim to use smart manufacturing, and smarter product use to eliminate the waste at
the design stage itself. In Appendix N, it is described how these strategies could be
implemented in the design of a graving dock.
The medium loops of the 9R framework (Reuse (R3), Repair (R4), Refurbish (R5),
Remanufacture (R6), Repurpose (R7)) aim to extend the lifespan of materials in a structure.
These strategies are closely related to the ease of disassembly of the design and depend on
the amount of additional energy that is needed to make the element compatible with the current
technology and market standards.
Recycle and Recovery (R8 and R9) are the longest loops in the framework and are typically
applied to elements which are considered to be ‘waste’ which would require a large amount of
energy input to create a new purpose. In a project, one should aim for the shortest loop
13
possible. For this project this would mean to not construct a graving dock at all and try to
increase Damen’s capacity in a different way, for example by purchasing an already existing
floating dock which has the same dimensions. However, the decision of constructing a new
graving dock has already been made at an earlier stage and at a ‘higher level’ of decision
making. For the purpose of this thesis, the designer therefore has no say in this and this
strategy (R0) will therefore not be considered. This can also be said for the Rethink (R1)
strategy since structures such as a graving dock simply do not allow for a lot of multi-purpose
functions except for perhaps the generation of green energy by placing solar panels.
The focus will instead be on the Reduce (R2) strategy, so to increase the efficiency in product
manufacture or use by minimizing the amount of natural resources and materials since this
allows for the implementation of more innovative design techniques.
14
C2 – Transport from
construction site to waste
plant
C3 – Waste processing
C4 – Disposal
Some of these methods require some explanation, which is done in the following paragraph.
As far as the A3 stage is concerned, a possible alternative can be to add fibres to the
underwater concrete mixture. Most building pits consists of two concrete floor, the first of which
is the underwater concrete floor that serves as a watertight barrier to allow the dry pumping of
the construction pit and it typically has a temporary function. The second floor is then placed
inside the dry building pit and it is a normal reinforced concrete floor which is calculated as if
there is no underwater concrete floor underneath. Due to the development of steel fibre
reinforced concrete floor, it has become possible to integrate the underwater concrete floor
into the structural floor and use this combination for both the water retaining layer as well as
the structural floor for the dock. Some example where this Steel Fibre Reinforced Underwater
Concrete Floors (SFRUCF) has been applied is the construction of the Albert Cuypgarage in
Amsterdam and the ‘Royal Van Lent’ dry dock. When applying such a integrated SFRUCF, not
only is the total volume of concrete needed in the design reduced, the redistribution of forces
within the concrete floor is enhanced as well, leading to less leakages and a reduced risk of
shrinkage cracks (Apon, 2019). However, the design of such a SFRUCF is not included in the
current CUR77 guidelines, but a combination of CUR 77 and CUR 111 is used. Often, the
SFRUCF is designed using a linear elastic beam model with a fictive E-modulus of 10.000
MPa (Apon, 2019). Other, more sustainable fibre types could be applicable as well.
One of the main sources of nitrogen dioxide (NO2) and carbon dioxide (CO2) is the combustion
of fossil fuels, during the transportation and construction phases of the life cycle. In order to
reduce this impact it is possible to use emission-free construction machines and equipment
during the installation process (A5).
Considering the energy use of the dock, an interesting opportunity arises. The intended
location of the graving dock is next-door to the waste incineration energy plant (Reststoffen
Energie Centrale) owned by Omrin. This factory uses combustible non-hazardous residues
from households and companies to generate energy which is in turn transformed into
electricity. The residual heat that is created during this process gets transported to neighboring
industrial companies such as the salt production company Frisia, where part of the residual
heat is extracted and the remaining heat-carrying water is transported back to the REC. There
are opportunities to expand this pipeline system and connect more places to use the residual
heat for varying purposes. Also for a graving dock, this serves as an interesting possibility
since some repair works require a constant working environment, for instance when specific
types of vessel-coatings and/or paints have to be applied. Connecting the dock to the REC by
expanding the existing pipeline infrastructure could serve as a sustainable energy source for
the dock.
After the base design is made, the implementation of these a few of these design alternatives
will take place.
15
Chapter 4: Functional Analysis &
Programme of Requirements
This chapter aims to describe the basis of the design of the new dock. First of all, the wishes
and requirements of the client regarding the facilities, dimensions and general layout of the
port is described in section 4.1. Afterwards the various boundary conditions, including the
natural, hydraulic and field boundary conditions and in the final section of this chapter, a stake
holder analysis is performed.
16
Section 4.1.2: Specific requirements and wishes
Now that the functional requirements of the dock are known, the specific requirements and
wishes from the client are treated. Appendix O elaborates on these requirements and can be
identified into a so-called object tree, as visible in Figure 17. All of these element make up the
total dock and need to be optimized in order to design the dock that meets all the requirements
mentioned above.
- Entrance ramp must enable entrance of aerial lift, a maximum slope of 20%.
OR
- 2 overhead rail cranes with 2 trolleys each, capacity of 20 t per trolley
For reference, the dimensions of the dock and the dock hall are visualized in the following
sketches.
Figure 18 shows a front view of the dock, where the design vessel MV Endeavour is drawn as
well.
17
Figure 18: Visualization sketch front view
18
Section 4.2.1: Natural boundary conditions
Natural boundary conditions consist of soil conditions and various hydraulic boundary
conditions such as water levels and groundwater levels and the governing wind directions and
forces.
For the soil conditions, three CPT tests in the vicinity of the intended location of the new dock
are used from past projects at the LEVVEL factory, which originate from the database of
ADONIN B.V. Another CPT test at the location of the current Damen shipyard that stems from
1992 and has been acquired from the Harlingen municipality database. The exact locations
and CPT test results can be found in Appendix B. Figure 20 below shows the soil profile which
is interpreted from overlapping the three combined CPT test at the intended location of the
new graving dock.
The values for the soil parameters where determined using table 2b from NEN9997. In all CPT
tests that were performed, the soil profile is characterized by a man-made top layer consisting
of clean sand, below which a number of soft layers can be found with a high concentration of
organic material, including a 1 meter thick peat layer and consequently a low bearing capacity.
Continuing downwards, a number of alternating sand and clay layers are found with a thickness
of approximately 3 meters each, with another peat layer at between -11 m NAP and -12 m
NAP. A thicker, sandy, clay layer can be found at a depth of -21 m NAP until -25 m NAP. From
this point on, the soil consists of dense, strong sand with a large bearing capacity. This strong
bottom sand layer serves as a suitable layer to install the retaining walls for the dock as it can
transfer the loads from the foundation structure. The deepest clay layer could potentially play
a role in preventing the uplift of the dock due to its water retaining properties.
When looking at the lithostratigraphy of the soil composition we can learn more about the
properties of the different soil layers. Figure 21 from Dinoloket shows the lithostratigraphic
classification of the different soil layers, describing from which historical period the soil
originates and consequently what characteristics this soil structure might have (DINOLoket,
2022).
19
Figure 21: Lithostratigraphic soil profile, source: (DINOLoket, 2022)
The first thing we observe is that the deep sand layer which starts at around -24 m NAP is from
the ‘Formatie van Urk’ which is formed during the Pleistocene-era, the first period of the
Quaternary during which several glacial and interglacial period took place. The layer on top of
this, the ‘Formatie van Drente’ originates from this cold period of time as well and has been
shaped during the Saalian era (370.000 – 130.000 year ago) during which a thick layer of ice
covered the soil. This specific layer package is from the ‘Gieten’ layer package, consisting of
clay and/or loam so it can be concluded that the clay layer that has been found from the CPT
tests between -20 and -24 m NAP also originates from the Pleistocene era (Vereniging
Ondernemers Technisch Bodemonderzoek, 2006). When looking at the effects of this historical
period on the behaviour of the soil, we can conclude that the OCR (over consolidation ratio) of
this clay layer from the ‘Formatie van Drente’ and the sand layer from the ‘Formatie van Urk’
can be assumed to be high, meaning that the current stress state of the soil layer is low
compared to the highest stress that it has been subjected to in the past and so additional
settlements due to consolidation of these particular soil layers can be considered to be unlikely.
This is different however for the higher laying clay layers, which all originate from the Holocene
era, which is a warm, interglacial period that makes up the past 10.000 years until present
time. The deepest Holocene formation layer that we can identify from figure 20 is from the
‘Eem Formatie’ which is characterized by shell-containing sand and clay layers. The sand layer
between -17 m NAP and -20 m NAP can be expected to be from this era. The sand layers
consist of moderately fine to very coarse grain sizes. Above this, a soil layer package from the
‘Formatie van Boxtel’, mainly consisting of sand layers which can be interrupted by peat, or
organically rich soil layers. This formation makes up the majority of the soil profile from figure
20, approximately from -17 m NAP until -5 m NAP. The sand layers are characterized by well-
graded, fine grains which make it suitable to use for different purposes after excavation such
as fill sand to be used for construction works. The final formation that can be identified is from
the ‘Formatie van Naaldwijk’, consisting of alternating sand and clay layers, in the soil profile
this formation is expected to be found between -5 m NAP and +2.5 m NAP. The OCR of this
Holocene soil package is expected to be low, meaning that increased stresses may lead to
settlements. Since the top meters of the soil profile consists of artificially placed sand, this layer
package may currently still be in the process of consolidation and therefore it is recommended
that the foundation of the graving dock reaches the deep sand layer that originates from the
Pleistocene era.
20
The ground water levels are again retrieved from the DINOloket website, and is based on data
from a period between 1996 and 2015, where the ground water table was monitored at a
location near the Damen Shipyard. The full results can again be found in Appendix B.
The average ground water level can be estimated to be at +3.5 m NAP during these 20 years
of monitoring, with a maximum value of +3.66 m NAP. The design ground water level will play
a large role in the vertical stability of the graving dock and potential soil settlements.
The wind setup is computed using wave information as described later in this chapter and
using equation (4.3) from the Flood Defences reader (Jonkman, Flood Defences, 2021). The
governing combination of wind velocity, using a design velocity of 1/100 years at Texel from
the Rijkoort-Weibull model of the KNMI of vb,0=33.4 m/s and a reduction factor of 0.94 (Smits,
2001), so a design wind speed of vd=31.4, combining with the fetch from West direction results
in a wind setup of 0.5 metres. This wind setup is already included in the chosen water level
from the Toetspeil, the governing wind direction and fetch within the port itself that would
generate the significant wind waves is from a NNE direction together with a fetch of 370 meters.
For determination of the significant wave height and period, the Young & Verhagen equations
have been used, leading to Hs=0.2 m and Ts=1.7 s for the governing wind direction and velocity
inside the port. A Rayleigh distribution is then assumed for small wave heights. The
exceedance probability is taken as 1%, since the scenario of a 1/100 year wave and a 1/100
year storm simultaneously is considered to be sufficiently conservative. The corresponding
design wave height is Hd=0.52 m. The governing wave height is the maximum wave height
between the significant wind wave height and the significant ship wave height, after consulting
multiple employees that work in the port environment, the maximum value for the significant
ship wave height is determined to be Hship=0.5 meters. Therefore, required freeboard will be
calculated based on the maximum ship wave.
The freeboard depends on the maximum allowable wave overtopping discharge. This value is
taken as qmax=1 l/s/m, corresponding to the maximum allowable value for building structure
elements, as determined in the European Overtopping Manual 2007 (Voorendt & Molenaar,
2020). Using Franco’s method for structures with a vertical wall such as the dock door, a
freeboard of 0.62 meters is maintained.
For the long term effects a value 0.5 meters is taken, which includes sea level rise, where a
relative sea level rise of the North Sea of at least 40 centimeters per century can be expected
according to research in 1990 (Voorendt & Molenaar, 2020). An additional 10 centimeters
accounts for land subsidence, which is in accordance with a prognosis as made by De Lange
& Gunnik from 2011 (Voorendt & Molenaar, 2020). It has to be mentioned that this corresponds
to the global, tectonic land subsidence as opposed to a local subsidence which is caused by
soil settlements. This value for the long term effects is also in accordance with the RCP4.5
from the IPCC reports, meaning a medium-pessimistic prediction scenario (Church, 2013).
21
In total, this means that the top of the dock door needs to be placed at +6.10 m NAP.
This is the extreme situation for which the dock door needs to be designed, as both the
governing loads for the strength of the door as the required height for the water retaining
function of the door are determined by this load case. Furthermore, the average tidal variation
changes between -0.99 meters NAP (Mean Low Water (MLW)) and +0.95 meters NAP (Mean
High Water (MHW)). After consulting with the client, it is decided to put the bottom level of the
dock at the same level as the bottom of the port. In this way, the capacity of the dock with
regards to the available depth is maximized while making sure that the amount of inflowing
sediment is not enhanced which would be the case if the bottom level of the dock would exceed
the bottom level of the port.
At Damen, the required minimal water level for docking is at 0.00 meters NAP, with a rising
tidal water level. The same requirement is applied for the new dock which means that the
maximum unladen draught of the vessel that can be docked is at 5.7 meters when taking into
account the block height of 1.80 meters. In extreme cases, a tidal window can be applied for
vessels with a larger draught, but in the current configuration, the design vessel is capable of
docking. All in all, this means that the dock gate runs from +6.10 m NAP till -7.5 m NAP, so a
total depth of 13.6 meters. These water levels can be found in Figure 22, where the height of
the quay, based on the height of the quay at Damen, is shown as well.
As far as the governing wind direction and forces are concerned, incidents in the past have
proven that wind forces play an important role in the design of a sufficiently stable dock hall for
this location, due to its coastal orientation and lack of coverage from surrounding buildings, the
governing wind forces can take extreme values.
A detailed analysis that has been carried out by Ilja Smits of the Dutch meteorological institute
KNMI has developed a ‘Rijkoort-Weibull model’ that can be used to determine extreme wind
values that have not even been measured yet. For this case, the wind velocity with a return
period of 100 years for the measuring station at Texel has been chosen, since this is the
nearest measuring station that has a coastal situation, being the most similar to Harlingen. The
design windspeed, with a corresponding wind direction from the West, is equal to 33.4 m/s
(Smits, 2001). According to Eurocode, the value for the velocity pressure qp depends on the
region in the Netherlands where the structure is located, the type of surroundings that the
structure experiences (rural, urban or coastal) and the reference height ze. For our dimensions,
this reference height is equal to the minimum value between the design width of the dock in
cross wind direction and two time the height of the structure, so ze=min(b,2h). For the chosen
dock orientation, b=155 and 2h=80, so the value of ze equals to 80 meters. For the wind region
and surroundings, it is determined that the dock is located in Area I with a coastal surrounding,
which gives a value for qp of 2.30 kN/m2.
22
Figure 23: Wind-area map for the Netherlands with Harlingen
highlighted by the red dot
Figure 24 shows the field boundaries for the Southern plot, owned by REC.
23
Figure 25: Field boundaries Southern plot, source: (Kadasterdata, 2022)
After investigating the cable and pipeline situation near the intended construction site, it can
be concluded that all cables and pipelines are situation along the access road, Lange Lijnbaan,
and crosses the Port in line with the road. These include data and electricity cables as well as
some major water transport pipelines owned by North Water, which is probably part of the
network that supplies industrial water. Altering of these water pipelines would lead to
substantial additional costs and should therefore be avoided. Appendix B shows the full
pipelines and cable situation, as well as a cross section of the area where these cables and
pipelines travel underneath the water. It is visible that one single data cable can potentially
intersect the construction site of the dock, this cable would have to be altered or removed.
24
Chapter 5: Development of design
alternatives
In this chapter, the first steps in the design of the dock are made by developing a base design
in chapter 5.1, where a conceptual design for the basic elements of the dock are made. These
consist of the construction pit for the dock chamber and the dock gate. In chapter 5.2, the
sustainable design alternatives are implemented in different design alternatives where also
other aspects such as the gate operating systems, pumping system and overall layout of the
dock can be varied to develop these design alternatives.
Chapter 5.1: Base design
This chapter treats the design of the retaining walls, underwater concrete floor and tension
piles and dock gate that forms the core of the base design of the dock. In the design of the
floor, the traditional layout is maintained which consists of a layer of water-retaining UCF,
followed by a layer of sand-fill above it, followed by a structural floor that carries the ship load.
Section 5.1.1: Retaining walls
The main types of retaining walls that are commonly applied in deep excavation pits are the
diaphragm wall and the sheet pile wall. In our case, it would be possible to select one of these
wall types for the base design and come up with design alternative that contain different wall
types. However, since we aim to minimize the environmental impact of our design it is more
time-efficient to select the optimal wall type up front based on available literature. An LCA
carried out by ArcelorMittal that compares two alternatives for the execution of a quay wall of
a cruise ship terminal concluded that the sheet pile design saved 44% of the amount of CO 2-
eq that is emitted compared to the D-wall design (ArcelorMittal Commercial RPS, 2021).
Furthermore, a thesis written by L. Van Holst from 2019, which uses a parametric design model
for the design of ship locks in the Netherlands, also found that variants which consist of a sheet
pile wall shipping chamber had a lower ECI-value and Global Warming Potential than the D-
wall alternatives (Van Olst, 2019). The ECI-value for one squared meter of sheet pile is
approximately equal to €35 while for a D-wall with a thickness of 1.2 meters without
reinforcement this value is €47 (DuboCalc, sd). A combi-wall, consisting of tubular piles that
are connected by two Z-profile sheet pile sheets are chosen due to their relatively large section
modulus compared to I or H-profiles. In this configuration, the piles are responsible for
supplying the required bending stiffness to withstand the earth and water pressures while the
sheet piles supply the water-retaining function and support the anchors. The profile of the piles
will be 1220x12,5 with a AZ-18 700 sheet pile, both in steel quality S430. Figure 27 shows a
schematization of this wall type.
25
D-sheetpiling is used for the calculation of this combi-wall, the complete calculations can be
found in Appendix C. The piles will need to have a minimal length of 18 meters as determined
by the D-sheetpiling program. In order to install the bottom of the piles in a sand layer it is
chosen to have the bottom of the piles at a level of -18 m NAP. Buckling of the tubular piles
can be prevented by making sure that these piles filled with or installed in over consolidated
sand. The bottom of the sheet piles are installed at a level of -13 m NAP. GEWI 63.5 anchors
are installed at +2 m NAP under an angle of 30 degrees and they have a total length of 23
meters. The grout body is 7 meters long and this configurations ensures that the grout body is
placed within a sand layer with a qc of at least 5 MPa and at a depth of at least 5 meters and
at least 1 meter of sand above the toe of the grout body, which is demanded according to the
CUR166 (CIE4363 reader, 2018). The centre-to-centre distance of the anchors is equal to the
total width of one combi-wall sections, 2.68 meters. Each anchor is pretensioned with a value
of 50 kN/m’, resulting in a prestressing force of 124 kN per anchor. A waling of 2 UNP260
profiles ties the anchors together. The D-sheetpiling input can be seen in Figure 28, where
the water levels and soil properties have been inserted corresponding to the findings in chapter
4. A surcharge load of 20 kPa is applied as is prescribed for RC2 structures according to the
Eurocode, while it is reasonable to think that this value might be higher in practice considering
the heavy equipment that will be required.
The construction stages of the retaining wall have to be defined next. After installation of the
combi wall, an initial excavation is performed to a level of +1.5 m NAP for the installation of the
anchors. Next, the building pit is fully excavated to a level of -8.5 m NAP in order to install the
tension piles and a 1 meter thick underwater concrete floor. The final stage is the dewatering
of the building pit. An overview of these stages can be seen in Figure 29 below.
Following the step-by-step plan described in CUR166, the maximum bending moments, shear
forces, anchor forces and displacement are determined during each of these stages and for 5
different scenarios which vary in the combination of high or low values for the soil stiffness and
ground and water levels. This variation is determined by the safety factors which are already
26
present in the D-sheet piling software, for a more detailed description of these scenarios
Appendix C can be consulted. These maximum values overall can be found in Table 2.
Value Stage
Bending moment Med [kNm] 2257 Stage 6 – Step 6.5 * 1.2
Shear Force Ved [kN] 1600 Stage 6 – Step 6.5 * 1.2
Deflection [mm] 42 Stage 6 – Step 6.5
Anchor Force Fanchor;d [kN] 947 Stage 6 – Step 6.5 * 1.2
The checks for the strength and stiffness of the combiwall using the design loads from Table
2 are found to be sufficient, as can be seen in Appendix C. For the check of the anchors, the
maximum force on the grout body and the maximum force on the anchor rod have to be
checked. While in reality anchor tests are supposed to be performed to determine the capacity
of the grout body, the CUR166 provides an estimation based on the length of the grout body
and the type of sand. This estimation is used to determine the required grout body length of 7
meters. Since the c.t.c. distance of the anchors is 2.68 meters, no mutual influence among
anchors is expected. Finally, the critical macro-stability of the system is checked, where the
stability factor for Bishop’s method is still equal to 2.27, as displayed in Figure 30. It has to be
mentioned that the length of the grout body could be increased to a length of 10 meters to use
its complete capacity but the current anchor configuration already meets all requirements.
It can be concluded that the current designed combi wall consisting of tubular piles with a
1220x12.5 profile and AZ-18 700 sheet piles together with the GEWI63.5 anchors meets the
design requirements. The total amount of anchors is equal to 124 and the combi wall consists
of 125 piles and 248 section of sheet pile. While both the bending moment capacity and shear
force capacity of the wall lead to low unity checks that might hint towards an overdesigned
retaining wall system, the anchor force and displacement prove to be governing in this case.
27
Section 5.1.2: UCF and Tension Piles
The design of the underwater concrete floor (UCF) and the tension piles are closely related
since they fulfil a combined function to ensure a vertical equilibrium of forces and to prevent
water from entering the excavation pit from underneath. The UCF will also function as a strut
that locally contributes to the horizontal equilibrium of forces. For the preliminary design of the
UCF, an average thickness of 1 meter (1000 millimetres) is assumed with a C20/25 concrete
strength class. The center-to-center distance of the tension piles in both the long and short
direction influence the strength of the UCF and so the first check is performed in long direction,
which is labelled as ‘check A’ in the CUR77 (CUR-Aanbeveling 77, 2014). These guidelines
are for the design unreinforced underwater concrete floors, from which it follows that the
maximum c.t.c. distance in long direction is 3 meters, a convenient value of Ly=2.5 meters is
chosen since a large c.t.c distance would reduce the amount of tension piles that are needed
and consequently reduce the design costs and environmental impact. In short direction the
c.t.c. distance should be chosen to be less than Ly to ensure ULS failure is not critical in the
long direction, Lx=2.4 meters and Lx,edge=1.8 meters which is the distance between the sheet
pile wall and the first row of tension piles, are chosen. This configuration leads to 60 rows of
12 tension piles, so 720 piles in total. A schematization of the pile plan can be seen in Figure
31 below, where we see a top view of the combi-wall and a more detailed schematization which
annotates the chosen c.t.c. distance of the tension piles.
For the UCF, checks B2 and G from the CUR77 have been performed which are deemed as
the most critical failure mechanisms. By assuming that bending cracks will occur in short
direction and the arching phenomenon will be governing, check B2 can be performed without
the need for a beam model which simplifies this preliminary design. Check B2 concerns the
equilibrium of forces in each arch, using the normal force in the concrete that has been derived
from the D-sheetpiling model by modelling the concrete floor as a strut. Check G concerns the
punching shear for each tension pile. The full calculation results as well as the strut-
schematization of the concrete floor can be found in Appendix C, but it can be concluded that
the chosen thickness and concrete strength class together with the chosen c.t.c. distance of
the tension piles leads to a sufficiently strong underwater concrete floor.
The chosen tension pile type is a GEWI63.5 anchor with a grouted diameter of 500 mm. As
has been determined in chapter 4, the piles need to be installed into the deep dense sand
layer which start at a level of approximately -24 m NAP in order to prevent problems due to
consolidation. The piles have been checked for the pull-out of a pile group, pull-out of a
single pile and the clump criterion. Since the centre-to-centre distance among the piles are
28
relatively large, the group-effect of the piles is negligible small which can also be concluded
from the calculation results in Appendix C. The pull out of a single pile is found to be the
governing failure mechanism and a final installation depth of -33 m NAP is necessary to
ensure sufficient strength. A cross-section of the final design of the construction pit can be
seen in Figure 32, including the anchors, retaining walls, underwater concrete floor and the
tension piles.
Assuming the gate as a simply supported beam with a width of 13.6 metres and a length of 30
metres, the schematization of the gate in the long direction is displayed in Figure 33.
In the short direction, the schematization of the load and the resulting maximum bending
moment is less trivial and can be seen in Figure 34 below, consisting of both the hydrostatic
and wave loads.
29
Figure 34: gate schematization short direction
The skin plate of the gate is the first element that has to transfer the loads to the supports and
so it also has to be designed to have sufficient strength in the ULS condition and the
deformation should be limited during the SLS condition. A thickness of 20 mm is chosen to
meet the strength requirements and it is supported by primary load carrying elements in both
vertical and horizontal direction in the forms of 7 horizontal girders and 5 vertical posts. In order
to meet the SLS requirements with regards to the deformation of the skin plate, 18 additional
stiffening elements have been added in vertical direction that are connected to the girders. At
the back of the gate, flanges with a thickness of 40 mm with varying widths in both directions
are placed to meet the SLS requirements for the total gate in the long direction (with a span of
30 meters). The gate will thus have a total thickness of 1.96 meters, sketches of the side view,
top view and total view can be seen in Figure 35 below.
The centre-to-centre distance of the girders is reduced near the bottom of the gate in order to
follow the shape of the hydrostatic load. In the long direction, the centre-to-centre distance is
kept constant at 1.5 meters, except for the final stiffeners that have been added in the centre
of the last span in order to meet the requirements for the maximum deflection of the skin plate.
It has to be mentioned that the load has be schematized in both direction while in reality the
short direction will be governing in carrying the load and so the total weight of the gate can be
optimized a bit. However, in reality this margin is considered to be negligible for a preliminary
design stage. The results of the design calculations can be found in Appendix C.
30
- Creating a water inlet at the opposite short side of the chamber that will enable an
inflow of water, either through gravity or by installing pumps, when the ships exits the
dock chamber, reducing the pressure difference at both sides of the ships and
consequently reducing the inflowing, sediment containing, current. In case a small
amount of sediment is still inside the dock chamber, the inlet can be used to force the
sediment towards the gate of the chamber and away from the ship, which would greatly
reduce the time and money that is required for cleaning the dock chamber. A number
of inlets should be installed that evenly spread the inflowing water over the full width of
the dock, to reduce the turbulence of this inflow, leading to the three evenly placed
inlets in the figure below. In order to tackle the inflow of sediment during entry of the
vessel into the dock, a system of winches can be installed so that tugboats are no
longer necessary for this process or at the very most merely to keep the vessel in the
correct position for docking. See figure 36 for an illustration of this solution.
- As mentioned previously, a system of winches to guide the vessel into the dock makes
the use of a tugboat unnecessary, but the addition of a system of winches to pull the
vessel back out the dock after the maintenance works have been performed would
make tugboats unnecessary for the complete docking and undocking procedure.
This can be accomplished by constructing a jetty that guides a winch system outside
of the dock entrance. The main downside of this strategy is the fact that a large jetty
with a length equal to the length of the design vessel would stick out into the harbor
which would form an obstacle for other marine traffic and operations.
31
flow velocity will be minimal here. The aim of this mitigation method is to form a
barrier for the inflowing sediment which forces it to settle earlier and be collected near
the entrance of the dock. The efficiency of this method is questionable however due
to the expected turbulence of the inflowing sediment flows, which are likely to not
follow laminar flow patterns around the small structures.
- A bubble screen can be applied at the entrance of the gate that aims to
reduce/prevent the inflow of sediment by forming a barrier for the stirred up sediment.
This can potentially be combined with a sliding dock gate that uses hydro jets to
move the gate by making these hydro jets multifunctional and supply the required
bubble-like barrier. The main downside of this mitigation method is the additional
required energy that is necessary.
- Excavation of a sediment retaining barrier in front of the entrance of the dock, where
the vertical distance between the bottom of the dock floor and the bottom of the
harbor is increased which creates a step that forms a barrier for the sediment near
the bottom to enter. This can be elaborated on by placement of a bottom protection.
This method aims to minimize / take away the inflow of sediment into the dock
chamber. Creation of this sediment trap in front of the dock chamber will locally
require additional dredging works and over time, this process will have to be repeated
due to unavoidable sedimentation of this trap. The effectiveness of this mitigation
measure is therefore questionable also due to the role that turbulence plays and the
relatively short duration of the docking/undocking procedure which likely will not be
long enough for sediments to settle.
Not all of these mitigation measures are equally appropriate to tackle the sediment issue of the
dock chamber and in order to determine which methods are most suitable and should be taken
into account when developing the design variants for the dock, a short preliminary multi criteria
analysis is performed to be able to discard those solutions which will prove to be either
unfeasible or unrealistic. Further details about this MCA can be found in appendix D, which
takes the constructability, effectiveness, costs, sustainability and amount of maintenance into
account.
From this MCA, two mitigation measures are discarded which are the sediment obstacle and
sedimentation trap. Their poor performance on the effectiveness criteria is the main reason for
this. The jetty is replaced by a normal system of winches since the construction of a long jetty
sticking out into the harbor will be in conflict with harbor regulations.
32
serves as a watertight barrier to allow the dry pumping of the construction pit and a reinforced
structural floor which is designed as if there is no underwater concrete floor underneath. For
the design of this graving dock, sustainable alternatives are considered that aim to reduce the
total volume of concrete that is necessary in the design of the construction pit. The alternatives
that will be assessed are a steel fibre reinforced underwater concrete floor (SFRUCF), a basalt
fibre reinforced underwater concrete floor (BFRUCF), a foil construction and a fully reinforced
underwater concrete floor.
A steel fibre reinforced UCF is about adding small steel fibres in the concrete mixture to
increase the compressive strength, flexural strength, tensile splitting strength and stiffness of
the concrete which allows the UCF to also have a structural function or reduce the amount of
concrete required. In practice, SFRUCF has been applied in a number of projects such as the
Albert Cuyp garage in Amsterdam and the ‘Royal Van Lent’ dry dock. In these projects, the
SFRUCF has been integrated into the structural floor. Not only is the total volume of concrete
needed reduced in this way, also the redistribution of forces inside of the floor is improved with
regards to a ‘traditionally’ reinforced floor, leading to less leakages and shrinkage cracks.
The main downside of adding steel fibres is the reduction of the concrete workability and the
corrosion of the material. Other fibres such as glass, carbon, polypropylene and basalt have
been included in concrete mixes to enhance the mechanical properties of concrete, each of
these materials having their own pros and cons, but the use of basalt fibres has risen in recent
times due to its high strength and economic efficiency compared to other fibres. Basalt is a
volcanic rock that is created by the eruption of igneous rock and typically originates from East
Asia. While carbon has similar strength properties as basalt, the main drawback of carbon
fibres is its high costs and the production of basalt fibres doesn’t need additives and requires
less energy than the production of carbon and glass fibres which makes it considerably
cheaper. Furthermore, basalt has good mechanical and physical characteristics since it has
high corrosion, thermal and freeze-thaw resistance (Bheel, 2020).
When comparing steel and basalt fibres, the influence on the concrete mixture is considered
first. The mechanical properties of basalt and steel can be seen in Table 3 below, stemming
from a research performed on the characteristics of basalt fibre reinforced concrete (Bheel,
2020).
Table 3: Fibre properties
Types of fibre Tensile strength [MPa] Elastic modulus [GPa] Strain at failure
Basalt fibre 3800-4840 79.3-93.1 3.1-6
Steel fibre 1700-2200 190-210 5-35
As described before, basalt has a higher tensile strength but a lower stiffness than steel. The
increased strength properties of basalt and steel fibre reinforced concrete has been subject to
research and the results can be compared to further assess the applicability of basalt as an
environmentally friendly alternative to steel in the construction industry. Ramesh and Eswari
(2021) tested the effects of basalt fibres on the strength properties of concrete by performing
failure tests on concrete samples that contained 0%, 0.5%, 1%, 1.5% and 2% basalt fibre
volume. The results showed that the highest enhancement in concrete strength occurred at a
fibre volume of 1.5% (Ramesh & Eswari, 2021).
Zheng et al (2018) performed the same experiments on steel fibre-reinforced concrete samples
that also contained 0%, 0.5%, 1%, 1.5% and 2% fibre volume. These results showed that the
strength properties always increased with increased amount of fibre and no ‘optimal amount’
was found, in contrast to the basalt fibre experiments (Zheng, et al., 2018). The results
regarding the increase in strength properties from both experiments for a concrete element
that contains 1.5% fibre volume can be found in Table 4 below.
33
Basalt fibre reinforced concrete (Ramesh Steel fibre reinforced concrete (Zheng, et al.,
& Eswari, 2021) 2018)
Compressive strength increase 4.5 18.2
[%]
Tensile splitting strength 57 58.5
increase [%]
Flexural strength increase [%] 22.6 25.2
Table 4: Fibre influence on concrete strength comparison
As table 4 shows, the main advantage of applying steel fibres is the significant increase in
compressive strength of the concrete compared to the basalt fibres, the other strength
increases are of the same order. It has to be mentioned here these conclusions are based on
merely one report for each fibre-type and additional research is recommended.
The implementation of these steel and basalt fibres in the base design is performed by altering
the material properties of the concrete based on the results portrayed in table above and
adding a term to the punching shear capacity of the concrete that is described by the following
relation, from CUR111:
𝑓𝑒𝑞𝑘,3
𝑣𝑅𝑓𝑑 = 0.18 ∗ 1.4∗𝛾 , where 𝑓𝑒𝑞𝑘,3 is the flexural stress corresponding to a crack width of 2.5
𝑓𝑡
mm when performing a three-point flexural test and 𝛾𝑓𝑡 = 1.25 according to CUR111 (CUR,
2018).
Since the required information to determine the value of 𝑓𝑒𝑞𝑘,3 is not present and can only be
retrieved by performing the experiments, the increased value for the characteristic flexural
strength 𝑓𝑐𝑡𝑘,𝑓𝑖𝑏𝑟𝑒 that takes into account the contribution of the fibres is used to compute the
value of 𝑣𝑅𝑓𝑑 . In reality the value for this strength property will be higher so it is a conservative
choice to use the flexural strength of the concrete, it is recommended to perform the test
beforehand since this would lead to a better (economic) optimization of the design.
The presence of the fibres increases the strength properties of the concrete as follows:
Table 5: Increased concrete characteristics by implementing fibres
Basalt fibre reinforced concrete (Ramesh & Eswari, 2021) Steel fibre reinforced concrete (Zheng, et al.,
2018)
fctk fctk,bf= fctk*1.226=1.9 N/mm2 fctk,sf= fctk*1.252=1.94 N/mm2
fctd,pl fctd,pl,bf= fctk*1.226=1.011 N/mm2 fctd,pl,sf= fctk*1.252=1.03 N/mm2
fck fck,bf = fck * 1.045 = 20.9 N/mm2 fck,sf = fck * 1.182 = 23.6 N/mm2
fcd,pl fcd,pl,bf= fck,pl * 1.045 = 11.18 N/mm2 fcd,pl,sf= fck,pl * 1.182 = 12.63 N/mm2
These altered properties will lead to a reduction of the thickness of the UCF and consequently
a large volume of concrete will be saved. The steel fibre reinforced UCF can be reduced to an
average thickness of 800 millimeters when maintaining the installation depth p of the anchor
head to its minimum allowed value of 175 mm, this would save 900 m3 of concrete compared
to the base design. The thickness of the basalt fibre variant can be reduced to 850 millimeters,
saving 675 m3 of concrete compared to the base design. When maintaining a fibre content of
1.5% by volume which leads to optimal performance of the basalt fibres as determined by
(Ramesh & Eswari, 2021), 54 m3 of steel fibres is required with corresponds to a total weight
of roughly 424000 kg of steel fibres. Comparatively, the basalt fibre variant would require
roughly 58 m3 of fibres which correspond to 12300 kg.
The full calculation results of the two altered UCF designs can be seen in Appendix D.
In conclusion, steel-fibre reinforced concrete has higher strength properties that would lead to
an increase in concrete savings compared to basalt fibre reinforced concrete, however the
environmental impact that is related to the fibres themselves is beneficial for the basalt fibre
alternative. Steel fibres are cheaper to produce but comprise of more emissions during the
handling and production processes, whereas basalt fibres are lighter and have better
endurance towards corrosion and other external factors. Furthermore, three point bending
experiments should actually be performed to determine the influence of the fibres on the
strength properties of the concrete but due to the lack of resources and the scope of this report
this has been omitted.
34
Foil polder
As mentioned in Section 2.3.4, the use of a foil polder construction could potentially make the
need for UCF redundant, as the plastic foil would ensure the water retaining function and the
ballast sand on top of the foil ensures the vertical equilibrium. The main downside of this
method is the required amount of space since the slopes of the building pit need to be at
approximately a 1:3 angle. Since available space is indeed scarce at the desired location of
the graving dock due to the presence of the waste incineration plant on the westside and the
harbor on the east, it is first established whether the construction of a foil polder is a viable
option. For this, an average water level of +0.00 m NAP and a required level of the bottom of
the graving dock at -8 m NAP is taken, which a volumetric weight of the ballast sand of γsand=
16 kN/m3 and for water γwater=10 kN/m3. The required thickness for the ballast sand layer can
8∗10
be found to be: (16−10) = 13.33 𝑚, meaning that the bottom of the foil is to be placed at a depth
of -21.33 m NAP and since the ground level is at +4 m NAP, maintaining a slope of 1:3 and a
dock hall width of 50 meters, the total width of the construction pit will amount to approximately
200 meters. Figure 41 shows that this amount of space is simply not available at the desired
location due to the overlap with the REC.
This design alternative is due to its unfeasible amount of required space discarded as a
reasonable option and it will not be included in the development of the design variants.
35
Section 5.2.3: Dock gate mode of operation
Section 2.3.2 mentioned the different types of dock gates that exist as well as the pros and
cons for each gate type. Using this information, a selection can be made of the most suitable
gate types that should be included in the development of the design variants. The main criteria
for the dock gate based on which the decision will be made are the ease & speed of operation
and removal, water tightness, low maintenance and cost. It has been chosen to include the
mitre gate, sliding gate and caisson gate in the development of the design alternative. For a
description of these gate types, the reader is referred to Appendix M.
We can conclude from this graph that the minimum valve area should be 5 m2 as the filling
time exponentially increases for smaller values. Taking this minimal valve area, the graphs
showing the development of the inflowing discharge Q and the water level difference between
the harbor and the dock chamber can be plotted, shown in Figure 43 and Figure 44.
The time required for filling up the dock is around 1400 seconds, which is roughly 25 minutes.
The required valve area can be achieved by implementing 2 circular openings with a diameter
of 1.8 m2.
Figure 44: Water level difference over time Figure 43: Inflowing discharge over time
36
foundation. It can be concluded that the costs and constructability of the cranes is significantly
more beneficial than the entrance ramp.
Choosing the overhead cranes would require a vertical clearance above the vessels and
consequently the height of the hall would have to be increased. On the other hand, the cranes
could also be valuable in the removal of the caisson gate in case the overhead rails are
extended beyond the gate of the dock. All in all, the construction of the entrance ramp requires
a significant amount of construction costs and together with the negative effects that it would
have on the structural integrity of the graving dock chamber leads to the choice of installing
overhead cranes that are multifunctional due to their potential role in the removal, placement
and maintenance of the gate. The requirements for the cranes as mentioned in the Terms of
Reference are a minimum of 2 cranes, each consisting of 2 trolleys with a minimum capacity
of 20 t per trolley.
Design variant A consists of the additional inlets on the opposite side of the dock chamber in
combination with the winch system to reduce sediment inflow into the dock chamber during
docking procedures, the traditional UCF is modified by implementing steel fibres into the
concrete mix and the structural floor will be integrated into the UCF to reduce the volume of
concrete necessary for the design. The mode of operation of the gate is a caisson gate that
can be removed with help of the overhead cranes that stick out of the dock hall.
Design variant B uses a bubble screen to filter out the sediment from the inflowing water during
docking and a sliding dock is applied here to be able to combine these two objects in the form
of hydro jets that create the bubble curtain as well as enable the sliding of the gate. A basalt
fibre reinforced underwater concrete floor is applied here, again the structural floor is integrated
into the UCF.
Finally, design variant C combines winch system with the fully reinforced UCF to develop a
more unorthodox design variant. The mode of operation of the gate is a single leaf miter gate
that swings open.
Now that the design variants are developed, the next chapter will elaborate the designs of
these design variant.
37
Chapter 5.3: Design alternatives specification
While the three design variants have been globally introduced in the previous section, a
number of design steps need to be performed in order to know the required amount of raw
materials, transport and installation works for each variant that can be used in Chapter 6 for
the evaluation of the three designs.
Section 5.3.1: UCF design finalization
The design of the UCF needs to be finalized by including the structural function, which will set
additional requirements for the thickness of the different floor types. First of all, the various
types of loads acting on the floors need to be specified. These are:
- Upward water pressure: A permanent load acting on the bottom of the UCF. The
magnitude of the nett upward water pressure depends on the ground water table, the
thickness of the floor and the installation depth of the floor.
o Ground water table: two extreme levels for the ground water table will
determine the representative load conditions, namely zmin=+0.00 m NAP and
zmax=+3.66 m NAP
o The installation depth of the floor stems from the requirement that the top of
the floor needs to be at -7.5m NAP. The installation depth hinstall=-7.5 - hfloor
o The thickness of the floor determines the magnitude of the self-weight of the
floor, using the volumetric weight of concrete: γconcrete=23 kN/m3.
This leads to an upward water pressure of:
𝛾𝑐𝑜𝑛𝑐𝑟𝑒𝑡𝑒
𝑞𝑢𝑝 = (ℎ𝑖𝑛𝑠𝑡𝑎𝑙𝑙𝑎𝑡𝑖𝑜𝑛 + 𝑧𝐺𝑊𝑇 ) ∗ 𝛾𝑤𝑎𝑡𝑒𝑟 − (ℎ𝑓𝑙𝑜𝑜𝑟 ∗ ) , with 𝛾𝑤𝑎𝑡𝑒𝑟 = 10 𝑘𝑁/𝑚3 , 𝛾𝑚 = 1.1.
𝛾𝑚
In the design, the value of qrep is used for SLS calculations and qEd is used for ULS calculations.
For the bottom fibre check under ship load, the upward water pressure works as a favourable
load and so it is corrected by applying a safety factor of 0.9.
𝑞𝑟𝑒𝑝 = 𝑞𝑢𝑝 ∗ 𝛾𝐺,𝑟𝑒𝑝 = 𝑞𝑢𝑝 ∗ 1.0
- Ship load: The ship load has been determined for the governing design vessel
Kommandor Susan which is a British supply vessel that would lead to the largest
concentrated load on the floor of the dock. The vessel’s weight is carried by three
rows of blocks over the length of the vessel, where the middle row of block is
responsible for the majority of the weight. For exact calculations, the reader is
referred to appendix E. Figure 46 below show the final weight distribution over a unit
width of the floor.
Again, safety factor γrep =1.0 is used for SLS calculations and γEd =1.5 is used for ULS
calculations.
38
- Shrinkage
Shrinkage of the concrete can lead to significant strains in the UCF and since the
deformations are restrained by the combi walls the resulting tensile stresses can lead
to cracks in the floor. Examples of this phenomenon causing cracks which can
eventually leads to a loss of water tightness of the floor or a deterioration of the concrete
quality are known from neighbouring docks and so it is crucial that sufficient attention
is paid to this time dependant effect.
In order to check the shrinkage of the concrete, the shrinkage strain εcs must be
determined, which is the sum of the autogenous shrinkage εca which develops in the
first 28 days after the concrete is casted during the hardening of the concrete. The
second term is the drying shrinkage strain εcd which is a slow process of tens of years
since it is a function of the migration of water through the hardened concrete. The
values of both terms have been determined conform the methods described in NEN-
EN1992-1-1, the complete computations can be found in Appendix E and the resulting
strains have been found to be:
𝜀𝑐𝑠 = 𝜀𝑐𝑎 + 𝜀𝑐𝑑 = 0.25 ∗ 10−4 + 0.286 ∗ 10−3 = 0.311 ∗ 10−3
CUR recommendation 111 prescribes that the loading due to shrinkage is to be applied
on the floor through a combination of a tensile force and a bending moment caused by
a distribution of the shrinkage strain of 0.9 εcs at the top of the floor and 0.6εcs at the
bottom of the floor.
- Thermal load
Annual and daily temperature fluctuations will cause additional strains in the floor. The
temperature underneath the floor, which is in constant contact with the groundwater,
will stay relatively constant. The temperature inside of the dock hall can take extreme
values during winter or summer which causes this the concrete to either expand (during
summer) or contract (during winter). These strain values have been determined
conform NEN-EN 1991-1-5, using the methods that are given for the thermal loads on
buildings and a thermal expansion coefficient of 10 *10-6 / ֯C for reinforced concrete.
For the design of the UCF, the winter conditions will lead to an additional tensile load
and bending moment leading to tension in the top fibre of the floor. The compression
caused by expansion of the floor during summer actually has a beneficial effect on the
strength of the floor. This means that being able to control the temperature inside of
the dock hall has a beneficial side effect for the bearing capacity of the UCF.
Fatigue
Another point of attention for the UCF is the influence of fatigue since the cyclic behavior of
the ship load acting on the floor can lead to fatigue of both the concrete and steel in the floor
over time. Fatigue can be described as the weaking of a material by repeated loads and checks
need to be performed for the concrete and steel separately (Khatri, 2016). The way in which
this is incorporated in the design of the UCF is by performing checks conform Section 6.8 of
the NEN-EN1992-1-1 norm (NEN, 2005). The checks can be seen in appendix F, from which
it is concluded that the check for the compression of the concrete and shear check can cause
problems long-term. It may therefore be advised to increase the concrete strength class to a
C30/C37. However, the checks described in the norm are more applicable to short duration
cycles such as traffic loads over a bridge deck or railways, the design amount of cycles that is
prescribed is in the order of 106. For the UCF, with an expected lifetime of 100 years and an
average of 20 load cycles per year, this amount of cycles is in the order of 2000, so a factor
500 less. The effect of fatigue can therefore also be expected to be less severe, it is assumed
that the current concrete strength class C20/25 is sufficient.
Beam model
A beam model is used to perform more detailed design of the UCF. For this, the floor is
regarded as a spring-supported beam with unit width, where the springs in the span of the
beam represent the tension piles and at the edges the spring supports represent the
39
connection between the UCF and the retaining wall. The strut force from the retaining wall acts
on the UCF as a normal force and a resulting moment is included at the sides of the beam.
The figure below shows an overview of this beam model.
The representative spring stiffnesses of the tension piles and retaining walls can be determined
using geotechnical information, the calculations can be found in Appendix E
CUR77 describes that MEd due to the eccentricity of the connection between the floor and the
wall is equal to:
ℎ𝑓𝑙𝑜𝑜𝑟
𝑀𝐸𝑑 = 𝑁𝐸𝑑 ∗
4
In the design of the integrated dock floor, several load cases can be identified that determine
the design of the floor:
• Initial UCF placement
• Placement of (reinforced) structural floor
• Ship load acting on the floor
In the following paragraphs, these load cases are introduced as well as the boundary
conditions that form the basis of the design for the 3 UCF design variants.
The load scheme can be seen in figure 48 and this is the similar case for which the base design
of the UCF is performed.
40
The checks that will be performed on this load case are a bending moment check (ULS) in the
top fibre of the floor, for which is assumed that cracking of the floor is allowed. A punching
shear check at the location of the tension piles and a shear force check (both ULS) complete
the design. Finally, shrinkage effects require additional attentional attention during the
complete design process of the floor.
Phase 2: Placement of (reinforced) structural floor
After the initial UCF has fully hardened, the rest of the floor is placed. This layer is traditionally
reinforced with rebars as opposed to other steel fibre option since this allows for a more
efficient placement of the reinforcement and consequently a reduction in the total amount of
steel. This top floor layer aims to make the chamber floor completely horizontal, so the even
out the tolerances in the initial UCF and it has a structural function to increase the total
thickness and bearing capacity of the floor during operation. Figure 49 shows a schematization
of this loading phase
The points of attention of this load scheme are a proper bond between the initial UCF and this
top layer, for which two layers of anchor disk plates from the tension piles are necessary and
dowels should be installed at a c.t.c. distance of approximately 500 millimetres. Furthermore,
the addition of the top layer influences the stress distribution in the UCF through the addition
of its self-weight.
Phase 3: Ship load acting on the floor
During operation, the ship load together with the upward water pressure is the governing load
case for the complete floor. For the determination of the ship load acting on the floor, it is
assumed that this ship is to be placed in the middle of the beam-model (in the short direction
41
of the floor) and no distinction of placement is made for the long direction of the floor is made,
meaning that the complete floor is designed to resist the governing ship load. The load case
can again be seen in the figure below. The load on the middle block is modelled as a distributed
load over a width of 1.5 meters and the load on the two outer blocks are modelled as a point
load due to the typically smaller width of these blocks.
The points of attention for this phase of the floor design is the maximum sagging moment at
the location of the ship load locally leading to cracks at the bottom fibre of the floor, where the
different fibre material can influence the maximum allowed crack width at the bottom since
corrosion will not play a factor in case of basalt fibres. Furthermore, the presence of the ship
will lead to increased hogging moment near the tension piles and consequently cracks in the
top fibres here. As a requirement, the client has determined that the floor should be absolutely
water tight due to the fact that oil and other chemicals are common on a dock floor and this
might lead to deterioration of the concrete floor if the crack width is too large. According to
NEN-EN 1992-3, this means that the UCF is to be allocated to Tightness class 3 (no leakage
permitted) and the corresponding requirements for the force distribution inside the floor is that
a minimum compressional zone height of 50 mm is required at all times. In other words, taking
the shrinkage strains and ship load acting on the floor into account, the resulting cracks in the
bottom of the UCF should not pass through the complete floor height (NEN-EN, 2011).
Finally, the foundation piles that are originally designed as tension piles will also serve as
compression piles during this loading phase and the capacity of the piles have to be checked
accordingly.
The first load phase has already been considered in chapter 5.2, the reader is referred to
Appendix F for the calculations behind load phase 2 and load phase 3 and the consequences
this has for the three design variants.
The required dimensions of the three floor variants are found to be as follows:
• The initial variant containing the integrated SFRUCF required a 800 mm thick steel-
fibre reinforced concrete layer with a 330 mm thick top layer of traditionally reinforced
concrete meaning a total floor thickness of 1130 mm.
• The variant containing the integrated BFRUCF required a 850 mm thick steel-fibre
reinforced concrete layer with a 280 mm thick top layer of traditionally reinforced
concrete meaning a total floor thickness of 1130 mm. While the crack width during the
use phase of the floor will also reach a significant magnitude, the favourable
characteristics of basalt with regards to its resistance against external conditions
means that the exposure of the basalt fibre should not affect the performance of the
concrete floor and therefore it can be accepted as long as the requirement regarding
the water tightness of the floor is met.
• A UCF with traditional rebars would lead to the smallest thickness, since only a floor
with a thickness of 850 mm is sufficient to perform both the water-retaining UCF
function and the structural function to carry the ship load. This is due to the fact that
the steel can be placed at an effective location, namely near the location of the outer
fibre, and the performance of the reinforcement bars is more effective than the
randomly spread fibres of the other two variant.
These conclusion may lead to another design alternative, namely a design that combines a
layer of basalt fibre reinforced concrete at the bottom for durability considerations in
combination with a top layer of steel fibre reinforced concrete for structural reasons topped off
by a structural floor with traditional reinforcement since it optimizes the beneficial
characteristics of each fibre type. For the current designs, the total volume of concrete for the
two fibre-reinforced variants is equal, meaning that the environmental impact score and costs
of the fibre material will determine which of these will score better in the evaluation of the
design alternatives.
Furthermore, the influence of the thermal load, caused by fluctuations in temperature
throughout the year, will lead to significant stresses in the floor and the temperature decrease
42
during winter can lead to cracks in the top fibre of the concrete floor. Actually when the
temperature inside the hall is below -6 ͦ C for an extended period of time, the concrete tensile
stress in the top fibre will exceed the cracking limits. This further substantiates the idea of
placing a hall on top of the dock where the temperature can be regulated using excess
heat from the waste incineration plant since this will increase the durability of the dock
floor.
Section 5.3.2: Sedimentation reduction measures
The variants each require the installation of different sedimentation reduction measures
contribution to the environmental impact and required costs for the designs.
Variant A: inlet at the other end of the dock
In order to reduce the inflow of sediment, the flow rate entering the dock through the water
inlets needs to equal the displacement of the design vessel. Taking friction losses along the
pipe and minor losses from the inlet and bend in the pipe, Appendix G determines that a pipe
with a diameter of 2 meters, with a length of 80 meters needs to be installed at a depth of -4
m NAP at the port side and at -6 m NAP at the dock. The installed pumps need to have
sufficient capacity to add 2.1 meters of head to a flow rate of 7 m 3/s. Schematization can be
seen in Figure 51 and Figure 52 below.
The manufacturing, transport and placement of the pipe will contribute to the environmental
impact as well as the manufacturing, transport, operation and maintenance of the pumping
system and gates.
For the other two variants, the manufacturing, transport and placement of respectively the
winches and bubble screen as well as the energy consumption during operation and
maintenance should be included in the environmental impact analysis which will be conducted
in chapter 6.1.
43
Chapter 6: Evaluation of alternatives and
selection
In this chapter, the design alternatives that have been developed in Chapter 5 are evaluated
based on their environmental impact and cost of construction, through performing a LCA and
Cost and Benefit Analysis respectively. In section 6.3, these results are processed into a Multi
Criteria Analysis which also take other aspects into account.
The first step is to determine the goal and scope of the LCA. In this case, the goal is to identify
the hotspots in the base design of the graving dock so that sustainable design alternatives can
be developed in an efficient and targeted way. On the other side, the emissions that are
involved with the different design variants are compared among each other and to the base
design, so that the effects of these sustainable design strategies can be quantified and
eventually an optimal design for the graving dock can be made. This optimal design can also
be a combination of design elements of the three design variants, which means that the design
solutions should also be compared separately. The audience for which the LCA is performed
are the clients which in this case are the members of Damen BV.
The functional unit for this LCA can be described as the amount of emissions of the graving
dock during a period of 100 years, with an average of 20 ships using the dock per year. The
phases that are included are the material production phase (A1-A3), transportation from the
production site to the graving dock construction site in Harlingen, installation/construction of
the dock, operational phase and maintenance phase of the graving dock. The residual value
of the design variants will only be included qualitatively and the transport on-site is neglected
due to the limited size of the construction site. Some other aspects that have been left out of
this LCA are the following:
- Factors associated with the hall structure, including the hall framework and walls,
workspaces, offices and the associated furniture and lightings, etc.
- Fuel used during the transport of the product is assumed to be included in the emission
factors given in the databases and are therefore not added
- Equipment used during the maintenance of ships inside of the dock chamber and in
the dock hall are not included since they are not part of the dock structure itself. This
includes paints, equipment and machinery.
44
- The contribution from wooden formwork for the construction of concrete elements is
considered to be negligibly small.
- The emissions that are associated with the manufacturing of the equipment that is used
is neglected due to the lack of available data of this topic and the fact that machinery
will be used for numerous building projects which mean the contribution to our total
emissions will be minimal nonetheless.
- The assembly and emissions of the concrete production plant. While it is most practical
to have a concrete production plant on site, the emissions that are associated with this
are neglected due to the lack of available data. It is recommended however to
investigate the possibility of using a floating, mobile concrete production plant since
this would be beneficial for the transport of the plant itself and the space use on site
(Bonton, 2022).
- Other elements that account for less than 5% of the total mass of the structure are
neglected.
The LCA aims to determine the Carbon Footprint of each design variant, meaning that only the
Global Warming Potential is considered and results are expressed in terms of kg CO 2-
equivalents. The reasoning behind this is the larger availability of data for CO2-emissions for
elements that are not present in the DuboCalc database. It is furthermore assumed that, in
general, a correlation exists between the GWP and the other environmental impact categories
and thus the variant that leads to the least amount of CO2 emissions can also be expected to
have the lowest value for the total Environmental Cost Indicator. For a first estimation of the
environmental impact of the design variants, this choice is therefore expected to be sufficiently
accurate.
The data has been retrieved from open databases such as DuboCalc, which contains a lot of
information for specific civil engineering products, and the Idemat app which has a lot of
information regarding the emissions related to raw materials, industrial processes and
transport. For some materials, additional research or estimations had to be made to determine
the CO2-emissions associated with that product. The data for the installation phase has been
estimated based on indicators for project cost estimations provided by De Boer & De Groot.
It has to be noted that the quality of the LCA largely depends on the quality of the data available
in the databases. For DuboCalc, the information is only available for a few elements that may
not completely match the profiles in the design and therefore need to be converted which can
lead to inaccuracies. Furthermore, information stemming from separate sources may lead to
inaccuracies since different producers could use different methods to determine the emissions
depending on the type of EPD (Environmental Product Declaration) that is maintained, the
standards that are used in different countries and the recency of the data. These factors may
cause the need for a critical review of the LCA study.
45
an Italian distributor is used and so the reinforcement steel is assumed to originate from this
factory.
The installation phase also requires some assumptions for equipment running on fuel. Here,
the emission factor for diesel is used for simplicity while in practice, some equipment could
potentially also run on more sustainable fuel alternatives such as HVO.
For the emissions during the operational phase, a short describing of the mode of operation of
the design variants is required. Each operational phase starts with the opening of the gate for
which Variant A (caisson gate) uses an external telescopic crane and the emissions stem from
its electricity consumption during the 30 minute period that is estimated for the duration of this
process, where the crane works at 50% of its maximum capacity. Energy consumption of the
sliding gate, which uses the hydro jets and winch system that reels the gate to the side
determine the emissions during the 30 minute gate operation for Variant B. Variant C only
needs the hydraulic power unit to operate the mitre gate which is expected to take around 30
minutes.
All three variants as well as the base design include the operation of the pumping unit to empty
the dock chamber within 3 hours.
The base design does not include any sedimentation mitigation measure and so the dock floor
needs to be cleaned by high pressure washers and a sediment removal machine. The
sediment then needs to be transported over a distance of 60 km to a special waste plant in
Middenmeer. The transport of a 5 centimeter thick sediment layer covering the dock floor by
truck is included in the computation of CO2-emissions of the base design. Furthermore, the
fuel consumption of two tugboats running at 50% of their engine capacity is included for the
docking duration of 2 hour.
Variant A includes the energy consumption of the pumping unit at the piped inlet during the
docking procedure and it is assumed that 1 tugboat is needed for support during docking
procedures, where this tugboat runs on 20% of its maximum power.
For Variant B the energy consumption of the hydrojet pumps that create the bubble screen
during docking procedures needs to be included as well as the single tugboat guiding the
vessel.
Variant C does not include the emissions of tugboats as they are no longer needed with the
presence of the winches and the emissions stem from the electricity used by these winches.
The maintenance strategies from which the emissions during the maintenance stage originate
are described in Appendix H.
Section 6.1.2: Impact assessment and results
The calculations to obtain the emissions during construction, operation and maintenance
phase of the dock base variant can now be determined, for the material phase this requires
the total amount of each material element and multiplying this by the emission factor. The
emissions for the transport stage also requires these amounts per element. For the installation
phase, the productivity of the equipment in combination with the amounts of each element
determines the fuel consumption and eventually the total emissions. The emissions during
operations are based on 20 ships entering and leaving the dock on average per year and finally
the emissions for the maintenance phase can be determine based on the maintenance
strategies.
In Appendix H, the calculations behind these phases can be found for each of the 4 design.
The first goal of the LCA was to gain insight into the so-called ‘hotspots’, the aspects that
contribute most to the total carbon footprint of the base design, which allows for an effective
development of sustainable design alternatives. Figure 54 below illustrates the total emissions
of the base design and the relative contribution of each lifetime phase to this total.
46
Emissions [ton CO2-eq] Contribution
Phase
Material 17790 71%
production
Transport to site 74 <1%
Construction 228 1%
Operation 7015 28%
Maintenance 80 <1%
Total 25385 100%
The material production phase is the predominant contributor to the total emissions of the
graving dock, which is mainly caused by the reinforced concrete which accounts for nearly
70% of the material production phase. The operational phase, and then mainly the cleaning
and disposal of the sediment, is the 2nd largest contributing phase. The contribution of the other
phases together is merely 2% of the total. It can be concluded that the sustainable design
aspects that have been developed, namely the reduction of the required amount of concrete
through an alternative floor package and the sedimentation reduction strategies indeed target
the largest contributing factors to the carbon footprint of the dock.
The steel piles that are part of the combi walls is another large contributing factor, as it accounts
for 17% of the total material production phase. While no further optimization has been made
regarding this aspect, reduction of the floor thickness can also lead to a reduction in pile length
since the installation depth of the floor will be less deep. This has not been taken into account
here but can potentially serve as another possibility to reduce the carbon footprint of the base
design.
The other goal of this LCA was to compare the design variants, initially by looking at the total
carbon footprint and then by zooming into the specific design aspects of each variant. Figure
55 illustrates the total emissions of the design variants, for each variant a reduction of
emissions can be seen compared to the base design.
The materials production stage and operational phase are still the predominant phases of the
lifecycle for the variants and the main reductions are caused by the alternative floor designs.
This is also illustrated in Figure 56 which compares the total emissions during the material
production, transport and installation phases of the UCF alternatives. The operational and
maintenance phases are irrelevant for the dock floor which is why these are omitted here.
Variant B shows the greatest reduction in emissions when compared to the base design and
the other design variants. The results show a strong correlation with the required amount of
reinforced concrete in the design, as this layer for Variant B is 50 mm less thick compared to
Variant A. This, together with the lower emissions associated with the basalt fibres compared
to steel fibres, results in the difference between Variants A and B. Variant C, which contains
47
only the concrete floor with traditional rebar reinforcement, actually shows an increase in CO2-
emissions compared to the base design.
As far as the sediment reduction strategies are considered, the comparison between the
various mitigation measures are depicted in Figure 57. The operational emissions from the
table are multiplied by a factor 100 to account for the complete lifetime of the dock.
The dominant contributing lifecycle phases differ among the design variants. For the base
design, no mitigation measures are installed and extensive cleaning works are necessary
which explains why emissions only stem from the operational phase here. For Variant A, the
external telescopic crane that removes the gate is the only source of emissions, this process
is estimated to be relatively short which explains the limited emissions. The required energy
for operation of the bubble screen makes up the majority of the emissions for Variant B and
the large amount of moving elements mean that increased maintenance account for the
emissions of the winch system for Variant C. All in all, the sediment reduction strategy of
Variant C scores best by saving 90% of the emissions that result from removal of the sediment,
Variant A saves around 85% of the emissions and Variant B causes 50% less emissions
compared to the base design.
Finally, the three dock gate types are compared as depicted in Figure 58
48
We can directly conclude that the caisson gate has the smallest carbon footprint compared to
the other types, due to the fact that is doesn’t require additional driving mechanisms that are
installed in the dock head structure that need to be constructed, installed and periodically
maintained. It only uses the power from the external crane to remove the gate.
Section 6.1.3: End-of-life phase
The end-of-life phase consists of the demolition of the dock and the residual value that
elements of the dock will still have at that point. The design elements that are unique to each
design variant will distinguish between the end-of-life value of the design variants, starting off
with the UCF variants. It is expected that the demolition of all three floor types will completely
crush the (reinforced) concrete and so naturally, no direct reuse of the floors are possible.
However, the concrete could be downcycled into other structural materials such as
construction rubble and create value in a different purpose.
The elements that set apart the floor types are the fibres inside of the UCF, starting off with the
steel fibres present in Variant A. After demolition of the UCF, the steel fibres are also expected
to be at the end of their lifetime and so recycling by reheating of the steel in a furnace seems
like the only option. This process, as well as separating the steel fibres from the concrete,
requires the input of additional energy which contributes to the total lifetime emissions of this
design variant A. The same can be said for the floor with rebar reinforcement of Variant C but
the process of separating the steel from the concrete should be a lot simpler. Since basalt
fibres typically have a longer lifespan than steel, the basalt fibres are expected to have a larger
residual strength. However, the basalt fibres cannot be recycled in the way steel can because
the production process of the basalt fibre is a one-way process (Slegers, 2022). All in all, the
demolition processes of the UCF variant are considered to be similar but the steel variant allow
for more possibilities for recycling than the basalt does and so the residual value of variants A
and C is considered to be greater than for Variant B.
For the sedimentation reduction strategies, the removal and demolition of the concrete pipe
requires additional processes and corresponding emissions compared to Variants B and C. All
elements, including the concrete pipe, the hydro jet installation and winches will be at the end
of their lifespan and therefore don’t offer a lot of residual value, the demolition process of
Variant B and C is a lot easier compared to Variant A.
As far as the gate operation systems are concerned, the demolition process of all three variants
can be expected to be similar, and where the caisson gate system does not include any
additional parts that requires removal or any residual value, some parts of the hydro jet system
or mitre gate system could still be recycled and reapplied in other projects. These variants also
cause the largest amount of waste products that needs to be treated since there are simply
more moving elements for the sliding and mitre gates and so these two variants are both
expected to have the least beneficial end-of-life phase.
All in all, the design variants do not offer a great deal of residual value since almost all elements
are considered to be at the end of their life span and so they would require additional treatment
before being able to serve another purpose. This can be prevented by taking the end-of-life
phase into account in the design process at an earlier stage, for example by applying the
‘design for disassembly’ method which allows for an easier demolition process or by
implementing sustainable building materials that have a lifespan that exceeds the lifespan of
the dock. This might be at the expense of the costs or structural integrity of the dock though
which calls for an accurate assessments of the risks and benefits of applying these sustainable
design methods.
Section 6.1.4: Conclusions
All things considered, a few conclusions can be drawn from this LCA:
- The developed design alternatives indeed target the largest contributing factors to the
carbon footprint of the base design, namely the amount of reinforced concrete for the
dock floor and the emissions during operations associated with the removal and
disposal of sediment.
49
- The amount of reinforced concrete is actually the predominant factor in the complete
carbon footprint of a graving dock, since the production of reinforced concrete accounts
for 53% of the total emissions during the complete lifecycle of the base design.
- The most sustainable design of the graving dock, purely looking at CO2-emissions,
would be the combination of the UCF-package of Variant B (containing basalt fibre),
the winch system of Variant C, and with the installation of a caisson gate.
The LCA is used to determine the most sustainable combination of design alternatives to
optimize the carbon footprint of the dock design. However, other criteria such as cost of
construction, speed of operation etc. also influence the performance of the design alternative
so it cannot directly be concluded that the final design should be the most sustainable
combination of design alternatives. A Multi Criteria Analysis should be performed which takes
the results from this LCA into account by weighing is against other criteria to eventually come
up with the final design.
In the multi criteria analysis, a fourth design variant will be taken into account. This
Variant D contains the elements that lead to the lowest carbon footprint with the exception that
a piped inlet is chosen due to its speed of operation compared to the winch system and the
fact that the carbon footprint of these two sedimentation reduction strategies is similar. So to
summarize: Variant D contains a basalt fibre reinforced UCF, a piped inlet and a caisson
gate, this combination of elements will also be included in the cost benefit-analysis and multi
criteria analysis in the next chapters.
The structure of this section is similar to that of the LCA in section 6.1; the inventory of costs
used per life as well as the actual calculations are placed in Appendix I. Here, a few
assumptions that form the base of the analysis are given as well as the interpretation of the
results.
50
extracting the raw material (A1), transport to the production location (A2) and the production
process itself (A3). For the transport phase, the costs solely include the fuel costs and these
have been determined based on the most recent diesel fuel prices. For barge transport, inland
transport has been assumed to be done by use of the Kempenaar vessel, which has been built
for Dutch inland navigation with a capacity of approximately 700 tons. Transport coming out of
Germany has been assumed to be done by a Dortmunder vessel with a maximum capacity of
1000 tons (Wereld van de Binnenvaart, 2023). Their fuel consumption in €’s per km is then
determined by using information from target prices stemming from a 2008 report by NEA which
have been adjusted to account for inflation over time. For the Kempenaar Vessel (CEMT II –
M2 class) this has been determined to be €5.30 per km and for the Dortmunder (CEMT III –
M4 class) this value is €9.30 per km (Van der Meulen, Quispel, & Dasburg, 2009). For transport
by truck, the fuel consumption has been determined to be at 40 litres per 100 km, with a diesel
litre price of €1.75 per litres results in €0.70 per kilometre (Webfleet, 2020).
The installation phase comprises of the fuel used during installation and the rent of some
external cranes that need to be used. For many installation processes of the base design (in
the information that has been provided by Adonin), the rent of the equipment is assumed to be
processed in the price that is used in the material phase and therefore omitted here. A diesel
price of €1.7 per litre is used again together with a kWh price of €0.8/kWh even though this
can be considered to be a conservative estimation when taking the developments in energy
prices into account.
The total cost of construction of the base variant has been found to be roughly €6.55 million
and after adding the 15% to account for design costs and other unforeseen factors, the costs
amount €7.55 million. For Variant A, the total construction costs amount to roughly €7.85
million, for Variant B this is roughly €7.99 million, for Variant C roughly €7.1 million and finally
for Variant D roughly €7.67 million.
The results for the construction costs can be compared in Figure 59.
When comparing the construction costs of the design variants, the costs of production of the
steel and basalt fibres appear to be larger than the reduction in costs from saving a certain
amount of concrete. Additionally, the implementation of the sedimentation reduction strategy
and the increased costs that this brings through the purchase of additional elements leads to
the result as depicted in figure 59. The design Variants A and B are approximately 5% more
expensive than the base design, while design variant C saves around 5% of these costs.
Variant D combines the base costs of Variant A with the implementation of the basalt fibres
which leads to a beneficial effects with regards to the construction costs. When looking at
merely the UCF variants as depicted in Figure 60 below we can conclude that in general, a
correlation exists between the reduction in carbon footprint of the dock and the costs of
construction. Larger reductions require a larger monetary investment and on the other side the
reinforced concrete floor present in variant C leads to a 30% cost reduction compared to the
base design but also leads to the largest carbon footprint in the LCA.
51
Figure 60: UCF construction costs comparison
The multicriteria analysis in the next chapter should decide which of these floor alternatives
should be chosen by weighing the relative importance of the reduction of carbon footprint and
minimization of construction costs that the three design variant bring. At first glance, the basalt
fibre reinforced UCF appears to score best among those two criteria (carbon footprint reduction
and construction costs).
A small note should be made regarding the constructability of the traditional reinforced UCF of
Variant C. In practice, placement of this floor type could require additional measures such as
the continuous drainage of the building pit before the reinforcement can be accurately placed
and the associated costs could drive up the total amount considerably. These additional costs
are currently not included in the CBA and therefore the results for Variant C might be
inaccurate.
The costs associated with the regulated maintenance strategy as described in chapter 6.1 will
be calculated next. Again the costs associated with the inspection of the design elements are
estimated to be 5% of the required estimations to construct, transport and install a new
element. The reader is referred to Appendix I for the calculations of the maintenance phase
per variant, the following paragraph will compare the calculation results.
The yearly operational costs are multiplied by a factor 100 to calculate the total operational
costs during the total lifetime of the dock. Figure 61 below shows that the operational costs of
the base design are a great deal larger than that of the design variants which proves that
prevention of the sedimentation of the dock does not only benefit the carbon footprint of the
design but will also drastically lower the total costs.
52
Figure 61: Total operational & maintenance costs
When only taking the costs associated with the sediment reduction strategies into account we
can see the same trend, as depicted in Figure 62.
The disposal of the sediment leads to the majority (78%) of the costs associated with the base
design and this is no longer needed when sedimentation of the dock is prevented. Variants A
and B show similar costs but the system of winches of Variant C is clearly the best-scoring
option because tugboats are not needed for this design variant.
As far as the costs involved with the different gate operation modes are concerned it is clear
from Figure 63 that the presence of hydro jets and winches for the sliding gate require a lot of
additional maintenance which makes the sliding gate variant the most costly.
All in all, the total lifetime costs of the design can be combined and plotted to collect all the
information into one graphical representation as shown in Figure 64.
53
Figure 64: Total lifetime cost comparison
The most cost effective design options would be variant C, but is has to be kept in mind that
this variant performed worst in the LCA and so the weighing of these and other criteria during
the multi criteria analysis will need to determine the actual optimal design combination. Apart
from that, design Variant D would score well on both the financial and sustainability criteria.
Section 6.2.3: Benefits – docking revenues
Throughout the lifetime of the dock, the main quantifiable benefit that will provide a yearly
revenue stems from the fee that ship owners have to pay for using the dock when the ship
maintenance works are performed. In collaboration with the client, these fees have been set
as follows:
Fee
Docking €5.000,-
procedure
Daily rent €2.000,-
A fixed amount of €5000,- is paid for docking the ship into the chamber and placing it on the
blocks, after that a daily fee €2000,- needs to be paid for each day that the ship is inside of the
chamber, a so-called ‘sitting-day’. Therefore, the total annual revenues depend on the duration
of the maintenance of one vessel and the number of vessels that will use the dock per year.
On average, each docking period will last between 10-14 days and 20 dockings per year will
take place which means a yearly income between €500.000,- and €660.000,- .
Section 6.2.4: Cost-benefit development over time
All monetary stream flows can now be combined in one overview to determine the development
of the net balance between costs and benefits over the lifetime of the structure. In this way,
the expected return of investment for the design variants can be made clear and it can be
determined whether the construction of the new dock is actually a viable choice over time.
In order to determine whether the investment that is required for the construction of the dock
is worth it, the Net Present Value (NPV) is calculated. In this way, the present value of future
cash flows is quantified by applying a certain discount rate that describes a required return on
an investment. The later the money stream appears, the larger the discount value and so the
smaller the value that money stream has in the present year. The NPV accounts for the time
value of money and it can be determined using the following equation (Jonkman, Steenbergen,
& Morales-Nápoles, Probabilistic Design: Risk and Reliability Analysis in Civil Engineering,
2017):
𝑇
𝐶𝑡
𝑁𝑃𝑉 = ∑
(1 + 𝑟)𝑡
𝑡=0
54
Where Ct is the nett cash flow (benefits minus costs) in year t and r is the discount rate. In year
0, the construction of the dock takes place and so the investment is made. After this, the yearly
flow of money consist of operational and maintenance costs and docking revenues. Applying
a discount rate of r=0.025 is common practice and implementing the described maintenance
strategies for the design variants, the development of the NPV over time can be seen in Figures
65 and 66 below.
The higher yearly operational costs for the base design results in a clearly less viable NPV
development over time, with an expected return of investment of approximately 90 years. For
all design variants this expected return of investment is shifted to approximately 20 years. The
maintenance costs for variants A, B, C and D are in the same order of magnitude and so the
main difference maker needs to be found in the difference in annual operational costs among
the three design variants, it turns out that the lack of tugboat renting costs saves around
€32000 in operational costs per year compared to variants A, B and D and this makes Variant
C the most beneficial option over time.
All in all, the main takeaways from this Cost-Benefit Analysis are:
- Comparing the results of the CBA and the LCA it appears that a reduction of carbon
footprint requires some additional investments. This is most clearly visible in the
55
results for the different dock floor packages since variant C clearly is the cheapest
option but also leads to an increase in CO2-emissions compared to the base design.
- Prevention of sedimentation pays off. The disposal of sediment throughout the 100
year lifetime is by far the largest expense and any mitigation method is worth its
investment. When comparing the sedimentation reduction measures, the absence of
tugboats is the deciding factor that makes the system of winches the most cost-
effective option.
- The expected return of investment is approximately 20 years.
56
The importance of the criteria can change during the lifetime of the dock, during the design
and construction phase, the sustainability might be more important but during the operational
phase, the ease of operation and ease of maintenance outweigh those other aspects. This
could’ve been accounted for by shifting the weighting factors for the comparison of the design
aspects.
The criteria and weighting factors are determined and so the next step is to assign scores to
the design variants for each criteria. Table 17 below shows the scores of the variants and
underneath the tables, an explanation of the allocated points is given in Appendix J.
Design alternative
Criteria Weight A B C D
Sustainability 35% 8 9 6 9
Return of Investment 10% 7 6 9 7
Ease of operation 30% 6 7 8 6
Ease of maintenance 20% 9 5 7 9
Constructability 5% 7 6 4 7
Scores 7.5 7.2 7 7.8
The scores for the sustainability criteria result from Figure 55, showing the total carbon footprint
of the design variants. The scores for the financial criteria are substantiated by the results from
Figure 66 and Figure 64 which shows the total lifetime costs. The remaining criteria require a
more subjective assessment, which is explained in Appendix J, which also shows MCA scores
for all design elements separately.
It can be concluded that alternative D offers the optimal combination of reduction in carbon
footprint, required monetary investments and ease of operation & maintenance. This design
variant is therefore chosen to be the recommended build-up for the new graving dock.
57
Chapter 7: Design Optimization
In this chapter, the results from the Multi Criteria Analysis in chapter 6.2 are used to determine
the optimal conceptual design for the dock. The final dock components such as the pumping
system, connection to the residual heat network and dock head structure are made.
58
The inclusion of a dock hall is recommended due to the benefits that controlling the
environment inside of the hall brings. Not only will the quality of some of the operations within
the dock be improved, the elimination of concrete shrinkage due to temperature fluctuations
throughout the year will be beneficial for the structural integrity of the dock floor and reduce
the probability of cracking. The structural design of the dock hall is not part of this thesis but
global requirements for the dimensions are determined and processed in the sketch, these
required dimensions are a length of 167 meters, a width of 50 meters and a height of 40 meters.
On top of this, it is highly recommended to include the use of the residual heat from the REC
for controlling the hall temperature. This would require the expansion of the infrastructure that
would bring the heated water to the dock but since both the equipment (large heat exchangers)
and knowhow is present within the Damen company, and it would create a sustainable win-
win situation for both Damen as well as the REC, these investments will likely be worth it.
The pumping unit to empty the dock chamber should consist of two centrifugal pumps that
have a maximum capacity of 10000 m3/h such as for example the Amarex KRT K pumps. This
pump capacity allow the dock to be pumped dry withing 3 hours which was one of the client’s
wishes. The pumping unit is installed near the dock gate so that the leakage discharge through
the gate can directly be removed from the dock again.
The hall should contain at least 6 workspaces to facilitate the maintenance works and an office
area, from which the dock master can operate the dock which is often build above the
workspaces. These workspaces should each roughly be 8 by 10 meters and 3.5 meters high.
Two overhead cranes are installed, consisting of 2 trolleys with a capacity of 20 tons per trolley.
The choice for overhead cranes over an entrance has been substantiated in section 5.2.5 and
the piled foundations of these cranes are not included in the current design.
While all of these elements require additional design steps which have been considered to be
outside the scope of this report, they have been included in the sketches below to show
indications of their dimensions and locations within the dock. These are by no means the final
designs for those elements.
59
Chapter 7.2: Concluding sketches
The final step is to combine all dock elements into some final sketches. Firstly, Figure 69 shows
a sketch with a cross section of the dock.
Figure 70 shows an overall side view of the dock, including pumping units near the gate and
the inlet pipe on the far right.
Finally, for completion, Figure 71 shows a top view of the final location of the dock in relation
with the current Damen shipyard.
60
Figure 71: Dock location top view
61
Chapter 8: Conclusion, discussion and
recommendations
This final chapter summarizes the main conclusions, discusses the results and offers
recommendations for further research in this field.
A base design that takes these requirements into account and serves as a basis for
comparison of the sustainable design variants contains the following elements:
• Retaining walls consisting of a combi-wall profile with GEWI63.5 anchors at a centre-
to-centre distance of 2.48 meters and tension piles that are installed at a grid of 2.5 by
2.4 meters.
• A floor package consisting of the ‘traditional’ buildup: a 1 meter thick underwater
concrete floor (UCF) that solely has a water retaining function, covered by a layer of fill
sand and a structural floor of 500 mm thickness containing rebars that carries the loads.
• The design gate has a width of 30 meters, a height of 13.6 meters and weighs 270
tons.
• The foundation of the graving dock should reach the deep sand layer that originates
from the Pleistocene era. This sand layer start at -24 meter NAP and has a high bearing
capacity in combination with a high over consolidation ratio (OCR), meaning that
settlement related damage to the dock can be prevented.
A life cycle analysis (LCA) identifies the following hotspots in the carbon footprint of this base
design:
• The amount of reinforced concrete for the dock floor package, accounting for 53% of
the total emissions of the base design.
• The removal and disposal of inflowing sediment that enters the dock through a pressure
gradient that is created when a vessel enters or leaves the dock.
Alternatives for these two hotspots firstly include the addition of fibers to the concrete mixture
of the UCF in order to also give it a structural function that can be integrated into the top floor
layer containing rebars, reducing the total concrete volume of the design. The UCF alternatives
should meet the watertightness requirements, which means that cracks should not pass
through the height of the floor.
Mitigation measures for the inflowing sediment are the construction of a piped inlet on the
opposite side of the chamber that aims to take away the pressure gradient during docking, a
bubble screen that filters the sediment out of the inflowing water and a system of winches to
dock the vessel instead of having to use sediment stirring tugboats. Finally, the mode of
operation of the dock gate is varied, which results in three design variant A, B and C:
62
• Variant A combines a steel fiber reinforced underwater concrete floor (SFRUCF) with
a total thickness of 1.13 m with the piped inlet and a caisson gate.
• Variant B has a basalt fiber reinforced underwater concrete floor (BFRUCF) with a total
thickness of 1.13 m, a bubble screen and a sliding gate.
• Variant C has a UCF containing rebars with a thickness of 850 mm, a system of winches
for docking and a miter gate.
An LCA for each of these design variant found the most sustainable design of the graving dock,
purely looking at CO2-emissions, to be the combination of the UCF-package of Variant B
(containing basalt fibers), the gated inlet system of Variant A, and the installation of a caisson
gate. These have been combined into a separate design variant D that has been found to
reduce the total carbon footprint by nearly 50%. Other takeaways from the evaluation of the
design variants by performing the LCA, cost benefit analysis (CBA) and multi criteria analysis
(MCA) are:
• Prevention of the sediment inflow should always be part of the design of a dock since
it pays off money wise, sustainability wise and the ease of operation is improved.
Actually, operational costs due to sediment disposal could make the construction of the
dock as a whole not economically viable.
• The use of ‘green’ construction equipment can reduce the emissions in the installation
phase by 25%, but this is only a small fraction of the total carbon footprint and the same
can be said for investing in local manufacturers to reduce emissions during the
transport phase
• The construction costs amount to roughly 7.6 million euros and the return period of
investment is 20 years on average.
• Implementation of an alternative UCF will require some additional investment costs but
leads to reduction in emissions, it is therefore up to the client to determine which of
these aspects is valued most.
All in all it is concluded that variant D gives the optimal combination of strengths of the different
design elements, since the basalt fiber reinforced UCF leads to an optimal combination of
construction costs and emission reduction benefits, the use of the gated inlet allows for the
most swift and easy operation of the dock and maintenance can easily be performed to the
caisson gate.
63
When determining the required height for the gate, it is found that the governing water level is
at +4.9m NAP while the ground level in this part of the harbour is at +4.3 m NAP. This means
that, under extreme circumstances, floodings can be expected to occur roughly every 100
years. This poses the question whether a heightening of the new dock location is required or
that we can accept a periodical flooding of the (area around) the dock. This is a matter of
economic optimization taking into account the risk of flooding, the associated economic
damages and the costs of the mitigation measures. Perhaps a solution could be to construct a
type of ‘terp’ for protection of equipment and allow the remaining infrastructure to be flooded
periodically. While it falls outside of the scope of this thesis to dive into this matter, it can be a
cause of concern and requires decision making from (local) authorities.
The watertightness of the floor is crucial for a durable use of the dock and in the design process
this is mentioned to be the governing boundary conditions for the UCF. It is mentioned that this
means that bending cracks in the bottom of the floor under ship loading conditions must not
be able to travel through the complete thickness of the floor and this is ensured by maintaining
a certain compression zone height at all times and it is concluded in the design calculations
that this is the case at the moment of cracking.
The same can be said for the efficiency of the sedimentation reduction strategies that have not
been tested. It is recommended to create physical models to perform scale test to the
mentioned strategies. In this report it is assumed that all measures completely take away the
inflow of sediment but in practice still some sediment might find its way into the dock chamber.
On the other hand the estimation of the sedimentation rate of the new dock might be
conservative, the question can be asked whether sedimentation at the location of the new dock
will actually pose that large of an issue. Since the sedimentation has been found to be the
predominant factor in both the LCA as well as the CBA results, the need for implementation of
the mitigation measures resulting from those experiments could have a large impact on the
final layout of the dock.
The development of design variants also leaves a lot of room for discussion. Chapter 3
investigated numerous methods to apply more sustainable design alternatives to the base
design and eventually the focus is chosen to be on the UCF and sedimentation of the dock. It
is not necessary that this choice leads to the best results in practice and so other alternatives
that focus on the end-of-life stage of the dock, on sustainable alternatives for other dock
elements such as the retaining walls or on alternative construction materials could form the
basis for further research in this area and might lead to more beneficial designs. Besides,
many combinations of the design elements have not been made. For example the combination
of the inlet and the winches could use the benefits of both elements. Fast operation of the
winches due to the presence of the inlet combined with the absence of tugboats.
The assessment of the environmental impact of the design variants also requires some
assumptions, firstly by setting the boundaries for the LCA and the elements that are and are
not included in the assessment. These assumptions are mentioned in Chapter 6.1 but might
also alter the accuracy of the results. Similarly, the emission factors used in the LCA stem from
multiple different sources that each are based on varying assumptions on their own.
The quality of the LCA largely depends on the quality of the data available in the databases.
For DuboCalc, the information is only available for a few elements that may not completely
match the profiles in the design and therefore need to be converted which can lead to
inaccuracies. Furthermore, information stemming from separate sources may lead to
inaccuracies since different producers could use different methods to determine the emissions
depending on the type of EPD (Environmental Product Declaration) that is maintained, the
standards that are used in different countries and the recency of the data. These factors may
cause the need for a critical review of the LCA study. Most accurate emission factors are
assumed to stem from the most recent EPDs of every element but again, this might not be
available for every element and in practice the distributors that supply the EPDs might not be
used. The accuracy of the emission factors used in the LCA has a huge influence on the final
design of the dock. The fact that the amount of reinforced concrete is the predominant factor
in the carbon footprint of the could potentially be attributed to the fact that this emission factor
is manually determined based on data for concrete and steel separately from varying EPD’s.
64
In case this emission factor proves to be significantly less from a more applicable or recent
EPD, other elements of the dock might prove to be a bigger contributor and sustainable
alternatives will have to focus on these elements, ultimately influencing the final design of the
dock as a whole.
All in all, the need for a critical review of the LCA study is large but since the goal of the LCA
is mostly about comparison of results among the design variants, it is considered to be
sufficient for this report. The LCA is made based on the Global Warming Potential (GWP), in
practice other categories such as nitrogen emissions could be more influential and the
correlation between amount GWP and amount nitrogen might not be as clear.
The results from the cost benefit analysis do not take into account the costs for the hall
structure and its foundations, the workplaces, finalization etc. and the factor of 15% might not
cover all of this. In presence, the investment could therefore come out considerably larger and
the results might be a lot less optimistic as depicted in chapter 6.2 and additional cost
calculations are required. Finally, some comments need to be made about the accuracy of the
multi criteria analysis. The weighting factors and assigned scores will always contain a sense
of subjectivity and consequently contain a lot of room for discussion. In chapter 6.3, the
weighting factors have been determined in collaboration with employees of the client’s
company and aim to represent the core of the project which is to develop a sustainable dock
design that still meets all requirements for structural integrity and smooth operation.
65
References
Allianz. (n.d.). Lock & Dock gates. London: Allianz.
Apon, M. (2019). Soil-structure interaction of a permanent steel fibre reinforced underwater
concrete floor system. Delft.
Arcadis. (2012). Structuurvisie gemeente Harlingen 2025.
ArcelorMittal Commercial RPS. (2021). Sustainable Ports - Life Cycle Assessment. Esch zur
Alzette: ArcelorMittal Commercial RPS.
Basalt Reinforced Composites. (2022). Benefits. Retrieved from BRC website:
https://basaltreinforcedcomposites.com/benefits-
comparisons/#:~:text=Basalt%20Reinforced%20Rebar%20has%20less,for%20concr
ete%20reinforcement%20construction%20applications.
Bheel, N. (2020). Basalt fibre-reinforced concrete: review of fresh and mechanical properties.
Springer Nature Switzerland.
Bij12. (2022). Stikstof en Natura 2000. Retrieved from Natura 2000 website:
https://www.bij12.nl/onderwerpen/stikstof-en-natura2000/vergunningen-en-
toestemmingsbesluiten/vergunning-aanvragen-of-niet/
British Standard. (1988). Code of Practice for Maritime structures. BSI.
Brundtland, G. H. (1987). Our Common Future. Oslo.
Cheng, Y. (2021). Research on the plume stability of air bubble curtains under low transverse
flow velocity environment in dredging engineering. Ocean Engineering, 109-133.
Church, J. P. (2013). 2013: Sea Level Change. Cambridge: Cambridge University Press.
CIE4363 reader. (2018). Reader Deep Excavations. Delft.
Confeeder Shipping & Chartering. (2022). MV Endeavor. Retrieved from JR shipping website:
https://www.jrshipping.com/our-fleet/m-s-endeavor/
CUR. (2018). CROW-CUR Aanbeveling 111 Staalvezelbetonbedrijsvloeren op palen.
CUR166. (n.d.). CUR-Publication 166 - damwandconstructies. Gouda.
CUR-Aanbeveling 77. (2014). het onwerpen van ongewapende onderwaterbetonvloeren.
CUR-rapport 98-9. (2014). Ontwerpregels voor trekpalen.
Damen Marine Components. (2022). Capstans. Retrieved from Damen Marine Components
Web site: https://www.damenmc.com/en/products/deck-equipment/capstans
Damen Marine Components. (2022). Tugger Winches. Retrieved from DMC web site:
https://www.damenmc.com/en/products/deck-equipment/tugger-winches
Deltares. (2013). Harbour Sedimentation Harlingen - Delft3D Flexible Mesh pilot. Retrieved
from Deltares website: https://www.deltares.nl/en/projects/harbour-sedimentation-
harlingen-pilot-delft3d-flexible-mesh/
Deltares. (2013). Kenmerkende waarden Kustwateren en Grote Rivieren.
Dillingh, D. (2013). Kenmerkende waarden Kustwateren en Grote Rivieren. Delft: Deltares.
DINOLoket. (2022). Ondergrondgegevens. Retrieved from DINOloket website:
https://www.dinoloket.nl/ondergrondgegevens
66
DuboCalc. (n.d.). Dubocalc webapp. Retrieved from https://app6.dubocalc.nl/
DYWIDAG-Systems international. (2012). GEWI - accesoires.
El Hamdi, A. (2011). Sedimentation in the Botlek Harbour - A research into driving water
exchange mechanisms.
Feraldi group. (2011, 03 22). Environdec. Retrieved from EPD database:
https://www.environdec.com/library/epd6689
Fort, J., Koci, J., & Cerny, R. (2021). Environmental Efficiency Aspects of Basalt Fibers
Reinforcement in Concrete Mixtures. Prague: Czech Technical University.
Fort, J., Koci, J., & Cerny, R. (2021). Environmental Efficiency Aspects of Basalt Fibers
Reinforcement in Concrete Mixtures. Prague: Czech Technical University.
Franco, L., De Gerloni, M., & Van der Meer, J. (1994). Wave overtopping on vertical and
composite breakwaters.
Geng, Y., & Doberstein, B. (2008). Developing the circular economy in China: Challenges and
opportunities for achieving 'leapfrog development'.
Gerkema, T. (2017). Interannual variability of mean sea level and its sensitivity to wind climate
in an inter-tidal basin. Earth Syst. Dynam., 1223-1235.
Global Compact Network Netherlands. (2022). Sustainable Development Goals. Retrieved
from GC Netherlands website: https://gcnetherlands.nl/sdgs/
Google Maps. (2022). Retrieved from https://www.google.nl/maps/@53.1859131,5.43404,14z
IDEMATapp. (2023, February). Idemat app. Delft, Netherlands.
IN2-concrete. (2023, February 22). IN2-fiber. Retrieved from https://in2-
concrete.com/product/bouwmaterialen/wapening/in2-fiber-basalt/
Jonkers, H. (2018). Reader 'Sustainability'. Delft.
Jonkman, S. (2021). Flood Defences. Delft: TU Delft.
Jonkman, S., Steenbergen, R., & Morales-Nápoles, O. (2017). Probabilistic Design: Risk and
Reliability Analysis in Civil Engineering. Delft.
Kadasterdata. (2022). Kadastrale kaart. Retrieved from Kadasterdata:
https://www.kadasterdata.nl/kadastrale-
kaart?q=Lange%20Lijnbaan%2028,%208861NW%20Harlingen&step=1
Khatri, D. (2016). Fatigue Analysis of Concrete Structures.
Kim, S.-W., Jang, S.-J., & Kang, D.-H. (2015). Mechanical Properties and Eco-Efficiency of
Steel Fiber Reinforced Alkali-Activiated Slag Concrete. Daejeon, Korea: Department of
Construction Engineering Education.
KLIC. (2022, July 28). KLIC notification. Retrieved from Kadaster.
KSB. (2022). KSB website. Retrieved from Amarex KRT: https://www.ksb.com/nl-
nl/lc/producten/pomp/dompelpomp/amarex-krt/A30B
Maher, A. (2013). Preparation of a Manual for Management of Processed Dredge material at
Upland Sites.
Marin Teknikk AS. (1999). Kommandor Susan - Stability Booklet.
67
Marine Insight. (2021, January 9). Dry Dock, Types of Dry Docks & Requirements for Dry
Docks. Retrieved from Marine Insight website:
https://www.marineinsight.com/guidelines/dry-dock-types-of-dry-docks-requirements-
for-dry-dock/
Molenaar, W. (2020). Locks. Delft: TU Delft.
Molenaar, W., & Voorendt, M. (2020). General lecture notes Hydraulic Structures. Delft: TU
Delft.
Natura2000. (2017). PAS-gebiedsanalyse Waddenzee (001).
Natura2000. (2022). Waddenzee. Retrieved from Natura2000 website:
https://www.natura2000.nl/gebieden/friesland/waddenzee
NEN. (2005). Eurocode 2: Design of concrete structures. Delft: Nederlands Normalisatie-
instituut.
NEN. (2006). NEN-EN 1992-3:2006+NB:2011. NEN.
NEN. (2020). National Annex to NEN-EN1991-1-4+A1+C2: Wind actions.
NEN-EN. (2011). NEN-EN 1992-3:2006 Constructies voor keren en opslaan van stoffen.
NEN-EN1992-1-1. (1992). Design of concrete structure. Eurocode.
Oenema, O. (2019). Factsheet 'stikstofbronnen'. Wageningen: Wageningen University &
Research.
PB Lifttechnik GmbH. (2022). Scissor lifts. Retrieved from PB Lifttechnik GmbH.
PH bouwadvies. (2021). Stikstofproblemen en gevolgen bouwsector. Retrieved from PH
bouwadvies website: https://ph-bouwadvies.nl/stikstofproblemen-gevolgen-
bouwsector/
PH bouwadvies. (2022, May). Stikstofproblemen en gevolgen bouwsector. Retrieved from PH
bouwadvies web site: https://ph-bouwadvies.nl/stikstofproblemen-gevolgen-
bouwsector/
Port of Harlingen. (2022). Havenkaart. Retrieved from Port of Harlingen:
https://www.portofharlingen.nl/havengebied/havenkaart/
Potting, J., Hekkert, M., Worrell, E., & Hanemaaijer, A. (2017). Circular economy: measuring
innovation in the product chain. The Hague: PBL Publishers.
Radermacher, M. (2013). The art of screening: effectivesness of silt screens. Terra et Aqua,
3-12.
Ramesh, B., & Eswari, S. (2021). Mechanical behaviour of basalt fibre reinforced concrete: An
experimental study. Pondicherry: Pondicherry Engineering College.
Ravestein Shipyard & Construction Company. (2017, October 12). Animation Caisson Dock
Gate D/16 Devonport Bn 474.
Readymix Industries Ltd. (2022, 03 29). Environdec. Retrieved from EPD database:
https://www.environdec.com/library/epd5720
Rensen, J. (2013). Staalvezelbeton in de iQwoning. Eindhoven: Unit Structural Design.
Ricardo-Engine. (2019). How to Calculate the Fuel Consumption of Diesel Generators.
Retrieved from Ricardo-Engine website: http://www.ricardo-engine.com/new/new-41-
765.html
68
Rijksoverheid. (2020, October 13). Stikstofaanpak: sterkere natuur, perspectief voor de bouw.
Retrieved from Website van Rijksoverheid:
https://www.rijksoverheid.nl/actueel/nieuws/2020/10/13/stikstofaanpak-sterkere-
natuur-perspectief-voor-de-bouw
Rijksoverheid. (2021, July 1). De bouwvrijstelling gaat in op 1 juli 2021. Retrieved from
Rijksoverheid website: https://www.rijksoverheid.nl/actueel/nieuws/2021/07/01/de-
bouwvrijstelling-gaat-in-op-1-juli-2021
Rijksoverheid. (2022, May). Climate policy. Retrieved from Government of the Netherlands:
https://www.government.nl/topics/climate-change/climate-policy/
Rijkswaterstaat. (2007). Hydraulische Randvoorwaarden primaire waterkeringen.
Rijkswaterstaat, Waterdienst.
Rijkswaterstaat. (2022). Wet milieubeheer. Retrieved from Rijkswaterstaat website:
https://www.rijkswaterstaat.nl/water/wetten-regels-en-vergunningen/natuur-en-
milieuwetten/wet-milieubeheer
RIVM. (2021, September). Stikstof. Retrieved from RIVM web site: https://www.rivm.nl/stikstof
Royal Haskoning DHV. (2019). Dock & Lock Gates. Royal Haskoning DHV.
Royal Haskoning DHV. (2022). Dry Docks - key facilities in Smart Shipyards.
Sagström, J. (2017). Streamlined LCA model and complete assessment of a hydraulic drive
system. Stockholm: KTH Industrial Engineering and Management.
SEB. (2021). Routekaart Schoon en Emissieloos Bouwen. Retrieved from SEB website:
https://www.opwegnaarseb.nl/
Slegers, R. (2022). Basalt to replace steel in concrete quay wall aprons. Delft: TU Delft.
Smits, I. (2001). Analysis of the Rijkaart-Weibull model. KNMI.
Staaltabellen. (n.d.). balkstaal.
Teekay Corporation. (2018, April 18). Step-by-Step: A Glimpse Into The Dry-Docking Process.
Retrieved from Teekay Coorporation Website:
https://www.teekay.com/blog/2016/04/18/step-step-glimpse-dry-docking-process/
Tran, N. K. (2015). An empirical study of fleet expansion and growth of ship size in container
liner shipping. International Journal of Production Economics, 241-253.
UNDP. (2022). What are the Sustainable Development Goals. Retrieved from UNDP web site:
https://www.undp.org/sustainable-development-goals
Van der Meulen, S., Quispel, M., & Dasburg, N. (2009). Kostkengetallen binnenvaart 2008.
Zoetermeer: NEA.
Van Leeuwen verankeringen B.V. (2008). GEWI-paal.
Van Olst, L. (2019). Parametric design-tool to optimize preliminary design of navigation lock
chambers. Delft: TU Delft.
van 't Wout, F., Groot, M., Sminia, M., & Haas, M. (2010). Functionele Specificatie DuboCalc.
Vanlede, J. (2014). A geometric method to study water and sediment exchange in tidal harbors.
Ocean Dynamics, 1631-1641.
Vereniging Ondernemers Technisch Bodemonderzoek. (2006). Ondergrond - Handboek
geotechnisch bodemonderzoek. Purmerend: Rijser Grafische Communicatie.
69
Voorendt, M., & Molenaar, W. (2020). Manual Hydraulic Structures. Delft: Tu Delft.
Waddenvereniging. (2022). De 10 grootste zorgen van de Waddenvereniging. Retrieved from
Waddenvereniging website: https://waddenvereniging.nl/werelderfgoedmoetbeter
Webfleet. (2020, februari 28). Hoeveel diesel verbruikt een vrachtwagen per kilometer.
Retrieved from webfleet website:
https://www.webfleet.com/nl_nl/webfleet/blog/hoeveel-diesel-verbruikt-een-
vrachtwagen-per-
kilometer/#:~:text=Gemiddeld%20brandstofverbruik%20vrachtwagen&text=Gemiddel
d%20verbruikt%20een%20vrachtwagen%2030,of%20juist%20tussen%20meerdere%
20steden.
Wereld van de Binnenvaart. (2023, february 22). Schepen. Retrieved from
https://wereldvandebinnenvaart.nl/schepen/
Winterwerp, J. (2005). Reducing Harbor Siltation. I: Methodology. Journal of Waterway, Port,
Coastal, and Ocean Engineering.
Zheng, Y., Wu, X., He, G., Shang, Q., Xu, J., & Sun, Y. (2018). Mechanical Properties of Steel
Fiber-Reinforced Concrete by Vibratory Mixing Technology. Hindawi.
70
Appendix A: Design vessels
This appendix gives information on the design vessels that are used for determining the
required dimensions of the dock and the hall. All information is retrieved from the JR shipping
website (Confeeder Shipping & Chartering, 2022).
MV Endeavour:
71
MV Elysee, governing for the required height of the hall:
72
73
Appendix B: Boundary conditions
Both the 3 locations where the CPT tests were performed and the results can be seen in the
following figures. Firstly, the CPT test which is performed at the most west-located spot is
shown, from now on labeled as location 1, can be seen in Figure B1.
Figure shows the corresponding soil profile, as interpreted from the CPT test, as well as the
values for relevant soil parameters which have been determined using table 2b from NEN9997.
Figure B 2: Soil profile location 1
74
The same is now repeated for the most north-eastern location, from now on labeled as location
2.
Finally for location 3, which is the most south-eastern location, the CPT result is shown in
Figure B5.
75
Figure B 5: CPT result location 3
When overlapping these CPT tests, a total soil profile can be determined as illustrated in Figure
B7, where the fit of the overlap of the CPT tests is not perfect, but still allows for identification
of the main soil layer package.
76
Figure B 7: CPT test overlap leading to identification of the total soil package
Figure B8 shows both the location of the measurements and the results of the groundwater
monitoring.
77
The various factors that make up the total design water level are:
- Storm surge level and astronomic tide
- Wind setup
- Wave setup
- Long term effects (sea level rise and land subsidence)
- Freeboard (determined by allowable overtopping discharge)
The storm surge level and astronomic tide are included in the ‘Hydraulische randvoorwaarden
primaire waterkeringen’ by Rijkswaterstaat, which gives a design water level in Harlingen of
+4.9 m NAP (Rijkswaterstaat, 2007).
The wind setup has been determined using equation (4.3) from the reader Flood Defences,
which reads:
𝑢2
𝜕ℎ1 = 0.5𝜅 𝐹𝑐𝑜𝑠(𝜑)
𝑔ℎ
With the following parameters:
- 𝜕ℎ1 is the wind setup in meters (m)
- 𝜅 is a friction factor
- 𝑢 is the wind velocity at an altitude of 10 meters (m/s)
- ℎ is the water depth (m)
- 𝐹 is the fetch (m)
- 𝛷 is the angle between the shore normal and the wind direction (deg)
The wind information as projected in Figure 19 is used to determine the governing wind
direction, which is West as this situation gives the largest combination of fetch and wind
velocity. Figure B9 below shows a Fetch of 35700 meters, as determined using Google Maps.
Figure B 9: Longest Fetch across the Wadden sea, source: Google Maps
The corresponding maximum wind velocity is found to 31.4 m/s, the water depth equals 12.4
meters and the value for κ is 3.2*10-6. Filling in the equation yields:
31.42
𝜕ℎ1 = 0.5 ∗ 3.2 ∗ 10−6 ∗ 35700 ∗ 𝑐𝑜𝑠(0) = 0.5 𝑚
9.81 ∗ 12.4
For determination of the significant wave height, the Young & Verhagen (1996) formulas are
used. In this instance, the maximum fetch is taken within the port of Harlingen with regards to
the orientation of the dock gate. The corresponding design wind velocity from a NNE,
78
maintaining a reduction factor of 0.64 (Smits, 2001) is vNNE=33.4 * 0.64=21.38 m/s. The Young
& Verhagen equations yield:
0.000441𝐹̃ 0.79
̃ = 0.24(tanh(0.343𝑑̃1.14 ) tanh (
𝐻 ))0.572
̃
tanh(0.343𝑑 )1.14
With:
𝑑𝑔 12.4 ∗ 9.81
𝑑̃ = 2 = = 0.27
𝑢10 21.382
𝐹𝑔 370 ∗ 9.81
𝐹̃ = 2 = 21.382 = 7.94
𝑢10
This gives:
̃ = 1.19 ∗ 10−6
𝐻
𝑇̃ = 1.25
Transforming the dimensionless wave height and period to the significant wave height and
period gives:
̃ ∗ 𝑢10
𝐻 2
𝐻𝑠 = = 0.2 𝑚𝑒𝑡𝑟𝑒𝑠
𝑔
𝑇̃ ∗ 𝑢10
𝑇𝑠 = = 1.7𝑠
𝑔
In order to retrieve the design wave height from this, a Rayleigh distribution is used. This is
accepted for small wave heights. The probability that the design wave height Hd is exceeded
during a storm with N waves is:
𝐻
−2( 𝑑 )2
−𝑁𝑒 𝐻𝑠
Pr(𝐻 > 𝐻𝑑 ) = 1 − 𝑒
Where N is calculated using the following relation:
𝑇𝑠𝑡𝑜𝑟𝑚
𝑁=
𝑇𝑠
Ts follows from the earlier Brettschneider equations and Tstorm for a storm along the coast is
around 2 hours (Jonkman, Flood Defences, 2021), leading to:
2 ∗ 3600
𝑁= = 4116 𝑤𝑎𝑣𝑒𝑠
1.75
The allowed exceedance probability is taken as 0.01 as it is assumed that the combination of
a 1/100 wave and the 1/100 water level leads to a sufficiently conservative design. Filling in
these parameters yields:
𝐻𝑑 2
−2( )
0.99 = 1 − 𝑒 −4141𝑒 0.26
𝐻𝑑 = 0.52 𝑚
79
For the long term effects a value 0.5 meters is taken, which includes sea level rise, where a
relative sea level rise of the North Sea of at least 40 centimeters per century can be expected
according to research in 1990 (Voorendt & Molenaar, 2020). An additional 10 centimeters
accounts for land subsidence, which is in accordance with a prognosis as made by De Lange
& Gunnik from 2011 (Voorendt & Molenaar, 2020).
Finally, the required amount of freeboard is determined by the maximum allowable overtopping
discharge during the design scenario. For elements of building structures, the European
Overtopping Manual determined a maximum specific discharge of 1 l/s/m, which is equal to
0.001 m3/s/m. For vertical walls, such as the dock gate, Franco and colleagues proposed the
following equation (Franco, De Gerloni, & Van der Meer, 1994):
𝑞 𝑅
−𝑏 𝑐
= 𝑎𝑒 𝛾𝐻𝑠
√𝑔 ∗ 𝐻𝑠3
Where:
Rc [m] = Overtopping height
q [m3/m/s] = specific discharge
a,b [-] = empirical coefficients, (a=0.192, b=4.3)
γ [-] = geometrical coefficient (for rectangular shapes, γ=1)
Hs [m] = significant wave height
The long term effects and the freeboard can now be added to the design water level to
determine the total required height for the gate with regards to NAP:
𝛥ℎ𝑔𝑎𝑡𝑒 = 4.9𝑚 + 0.5𝑚 + 0.62 𝑚 = +6.02 𝑚 𝑁𝐴𝑃 ≈ +6.10 𝑚 𝑁𝐴𝑃
Knowing that the bottom of the dock is at -7.5m NAP, the total height of the gate will be:
ℎ𝑔𝑎𝑡𝑒 = 13.6 𝑚
80
Figure B 10: Cable and pipeline situation, source: (KLIC, 2022)
Figure B11 shows the cross sectional location of the industrial water pipeline owned by North
Water along the water. A depth of 16 meters below bottom level is reached (KLIC, 2022).
Figure B 11: Water pipeline situation along the water, source: (KLIC, 2022)
Figure B12 shows the same for data and electricity cables, which reaches a depth of 13 meters
(KLIC, 2022).
81
Figure B 12: Data and electricity cable situation along the water, source: (KLIC, 2022)
82
Appendix C: Base design calculations
This Appendix contains the calculations behind the base design, starting with the retaining
wall.
Step 1: Determine the representative cross section, all relevant external loads and water
pressures and the safety class
Figure C 1 shows the representative cross section for the wall, with a water level of +3.5 m
NAP as determined in chapter 4. The structure is classified as an RC2 safety class structure
and so a surcharge load of 20 kPa is applied.
83
The design values for the properties of the combi wall are presented in Figure C 3, where the
material factor for steel is equal to γm=1.0. Since the sheet pile wall consists of Z-profiles where
the interlocks are not in the neutral bending axis, correction factors to account for oblique
bending do not have to be taken into (CUR166).
84
Figure C 6: minimum embedment depth of combiwall
Step 6.5 for stage 6 (dewatering) is governing for the maximum bending moment, shear force
and deflection of the wall. Note that the values from the graphs in Figure C 8 still need to be
multiplied by a factor of 1.2 to reach the design values in Table C 1.
Value Stage
Bending moment Med [kNm] 2257 Stage 6 – Step 6.5 * 1.2
Shear Force Ved [kN] 1600 Stage 6 – Step 6.5 * 1.2
Deflection [mm] 42 Stage 6 – Step 6.5
Anchor Force Fanchor;d [kN] 947 Stage 6 – Step 6.5 * 1.2
85
Figure C 8: Graphs showing maximum design values for forces and displacements
86
Since 𝐹𝑟,𝑚𝑎𝑥,𝑔𝑟;𝑑>𝐹𝑎,𝑚𝑎𝑥,𝑔𝑟;𝑑, the designed grout element is strong enough and needs to be
elongated.
Next up, the strength of the anchor rod is tested. The value of the design yield force is equal
to 1584 kN, so 𝐹𝑟, 𝑚𝑎𝑥, 𝑟𝑜𝑑, 𝑑 = 1584 𝑘𝑁
The design value of the force acting on the rod is: 𝐹𝑎, 𝑚𝑎𝑥, 𝑟𝑜𝑑, 𝑑 = 1,25 ∗ 𝐹𝑎, 𝑚𝑎𝑥 = 1184𝑘𝑁
Since 𝐹𝑟,𝑚𝑎𝑥,𝑟𝑜𝑑,𝑑<𝐹𝑎,𝑚𝑎𝑥,𝑟𝑜𝑑,𝑑, it can be concluded that the anchor rod is sufficiently strong.
Furthermore, the anchor was prestressed at a value of 124 kN which influences the maximum
occurring deflection that will be evaluated in step 10.
Since all design steps are performed and all check are successfully performed, we can
conclude that the current design fulfils all needs.
Finally, the waling that ties together the anchors needs to be designed, the design moment in
the waling is described by:
1
𝑀𝑤𝑎𝑙;𝑑 = ∗𝐹 ∗𝑎
10 𝑎;𝑑
Where 𝐹𝑎;𝑑 is the design value of the anchor force, which is equivalent to the maximum force
of the grout body, 𝐹𝑎;𝑑 = 1042 𝑘𝑁 and a is the centre to centre distance of the anchors, 𝑎 =
2.48 𝑚. Filling this in yield:
1
𝑀𝑤𝑎𝑙;𝑑 = ∗ 1042 ∗ 2.62 = 273 𝑘𝑁𝑚
10
87
The required section modules Wd can now be determined by stating that the maximum
allowable moment 𝑀𝑟𝑑;𝑤𝑎𝑙 = 𝑀𝑤𝑎𝑙;𝑑 = 𝜎𝑣𝑙;𝑑 ∗ 𝑊𝑑 . Assuming a steel quality S430, the minimal
required section modulus is equal to:
𝑀𝑤𝑎𝑙;𝑑 2.73 ∗ 106 𝑁𝑚𝑚
𝑊𝑑 ≥ = = 6.3 ∗ 105 𝑚𝑚3
𝜎𝑣𝑙;𝑑 430𝑁
𝑚𝑚2
This means that 2 UNP260 profiles are necessary to make up the waling, as they have a Wy;el
of 3*105 mm3 each (Staaltabellen).
UCF design
First of all, some decisions have to be made regarding the thickness of the underwater
concrete floor and its strength properties. CUR77 requires a minimal average thickness of 800
millimetres and often a thickness of 1 meter is chosen in practice and the same is done for this
design. The minimal thickness of the UCF taking placement tolerances into account with
tolunder=150 mm for a UCF resting on a sand layer and tolover=75 mm for the use of the hop-
dobbler method leads to:
2 2
ℎ𝑚𝑖𝑛 = ℎ𝑎𝑣𝑔 − √(𝑡𝑜𝑙𝑢𝑛𝑑𝑒𝑟 + 𝑡𝑜𝑙𝑜𝑣𝑒𝑟 = 833 𝑚𝑚
A concrete strength class of C20/25 is chosen which has the following relevant properties:
The first check that will be performed considers bending cracks in the ‘center region’ of the
UCF in the long direction. This Check A from CUR77 is also used to determine the maximum
centre-to-centre distance for the tension piles in long direction Ly by examining the following
equation:
where
and
1
𝑀= ∗ 𝑞 ∗ 𝐿2𝑦
8
Q is the representative value for the loads which is equal to the water pressure at a level of -
8.5 m NAP minus the self-weight of the concrete, which is equal to:
𝑘𝑁
𝑘𝑁 23 3
𝑞 = ((8.5 𝑚 + 3.66 𝑚) ∗ 10 3 ) − (1 𝑚 ∗ 𝑚 ) = 100 𝑘𝑁/𝑚2
𝑚 1.1
Where the volumetric weight of the concrete is corrected with a material factor 1.1.
This leads to Ly,max=3.08 meters, for simplicity, a value of 2.5 meters is chosen.
Next, check G is performed which concerns the punching shear check of the connections
between the anchors and the concrete floor. The following must be true (CUR-Aanbeveling 77,
2014):
88
VEd is equal to the upward pressure qEd multiplied by the area that is carried by each tension
pile. Lx is chosen to be 2.4 metres, this means:
𝑉𝐸𝑑 = 𝑞𝐸𝑑 ∗ 𝐿𝑥 ∗ 𝐿𝑦 = (100 ∗ 1.35) ∗ 2.4 ∗ 2.5 = 816𝑘𝑁
GEWI63.5 anchors are again chosen as the type of tension pile, for which the CUR77 has
prescribed the following checks:
Here, p is the distance between the average level of the top of the floor and the average level
of the bottom of the dish. The value of p should be chosen based on:
Where tolanchoring=100 millimeters. This leads to pmin=175 mm and pmax=500 mm, p=200 mm is
chosen here, leading to dmin=550 mm and kr=0.833. For the anchor dish, a cast anchor piece
with a diameter of 350 mm is chosen (DYWIDAG-Systems international, 2012).
This means a value for u1=8000 mm and since vmin=0.32 N/mm2, the maximum allowable shear
force is equal to:
1.25∗816
VRd=1166 kN, leading to a u.c. of 1167 = 0.87.
89
Finally, check B2 is performed which checks the bending moment in the short direction without
the membrane effect. CUR77 prescribed the following check (CUR-Aanbeveling 77, 2014).
The normal force in the concrete floor is determined from D-sheetpiling where the floor is
modelled as a strut with E=30000 Mpa, A = 1m2/m’ and L = 15 metres. This leads to a force in
the strut of Fstrut=2460 kN, as can be found in Figure C 10.
90
The value of z1 and z2 are computed to be 212 mm and 183 mm respectively and so the
value for MRd= Ned* z2 = 448 kNm, meaning that this check is also verified with a unity check
98
of 448 = 0.22.
It can therefore be concluded that the design of the UCF is sufficiently strong.
The design value of the bearing capacity has to be corrected for using multiple safety factors:
For screwed piles with grout injection or grout mixing, αt=0.009 and the most conservative
value for γm;var;qc=1.5 is taken. The value for ξ=1.3 for 3 CPT test in a non-stiff soil (CUR-rapport
98-9, 2014). The effect of excavations on the bearing capacity of the soil have to be taken into
account as follows:
Next, the value for f1 is determined from the degree of densification due to the installation of
the pile field, described in an increase in the relative density Re.
91
The value for r in this case is equal to 5 in the long direction and r=4.8 in the short direction,
n=4.
The calculation of f1 for a varying length of the tension pile can be seen in Table C 2 below.
92
The next step is to determine the value of the tensional load factor f2 which takes the decrease
in effective stress in the layers from which the pile obtains its tensional bearing capacity into
account.
If the values for f2 are known, we can calculate the bearing capacity of the pile group and also
that of a single pile, since for a single pile the equation mentioned earlier is used with f1=f2=1.0.
These calculations are all performed within Excel, the results can be seen in Table C 3 below,
where the last column gives the maximum bearing capacity for a single pile and the second to
last column gives the bearing capacity of the pile group. We can conclude that the pull-out of
a single pile is the governing failure mechanism and the minimal depth is equal to -33 metres,
which is highlighted by the green cells which indicate when Fr;tension;d is larger than Ved, with Ved
is equal to the upward water pressure, which is equal to: qed=100 kPa *1.1 = 110 kPa.
Ved=qed*Lx*Ly=665 kN
93
Finally, the clump criterion at an installation depth of -33 metres should be checked, which
takes the maximum strength of the mobilized soil volume by each pile into account. It can be
found that this strength is also sufficient and therefore the governing failure mechanism is
indeed pull-out of a single pile, since Fclump;max;d=1345 kN, the calculations can be seen from
Table C 4
Gate design
A schematization of the situation that is used in the Sainflou method to determine the wave
loads can be seen in Figure 28, source: (Voorendt & Molenaar, 2020).
The still water level in front of the gate will increase with h0:
94
1 2
ℎ0 = ∗ 𝑘 ∗ 𝐻𝑖𝑛 ∗ coth (𝑘 ∗ 𝑑)
2
Where:
h0 = increase of mean water level in front of the structure
Hin= height of the incoming wave
d = water depth in front of the gate
k = wave number of the incoming wave
For our situation Hin=Hd=0.52 meters, d = 12.40 meters and the wave number k will be
determined assuming deep water, which will have to be checked. The wave length is calculated
first using:
𝑔 ∗ 𝑇 2 𝑔 ∗ 1.752
𝐿= = = 4.78 𝑚
2𝜋 2𝜋
Deep water check:
ℎ 1 12.40 1
> → ≈ 2.6 > , 𝑑𝑒𝑒𝑝 𝑤𝑎𝑡𝑒𝑟 𝑎𝑠𝑠𝑢𝑚𝑝𝑡𝑖𝑜𝑛 𝑖𝑠 𝑣𝑎𝑙𝑖𝑑
𝐿 2 4.78 2
𝜌 ∗ 𝑔 ∗ 𝐻𝑖𝑛
𝑝0 =
cosh (𝑘 ∗ 𝑑)
We can now compute these values using the parameters we know:
𝑝1 = 1000 ∗ 9.81 ∗ (0.52 + 0.18) = 6.79𝑘𝑁/𝑚2
The characteristic value of the corresponding load per meter width of the gate are now as
follows:
1 1
𝑄𝑤𝑎𝑣𝑒,𝑟 = ∗ 𝑝1 ∗ (𝐻𝑖𝑛 + ℎ0 + 𝑑) = ∗ 6.79 ∗ (0.52 + 0.18 + 12.4) = 44.4 𝑘𝑁/𝑚
2 2
95
Total gate in long direction
For ULS conditions, the bending moment line can be seen in the MatrixFrame screenshot
below.
Elements Amount Thickness Width Total Total Distance Distance Iown Isteiner Itotal
per width area to top to NC
element fibre
[-] [mm] [mm] [mm] [mm2] [mm] [mm] [mm4] [mm4] [mm4]
Skin plate 1 20 13600 13600 272000 10 878 9.07 *106 2.1 *1011 2.1
*1011
Girders 7 1900 30 210 399000 970 82 1.2*1011 2.68*109 1.2
*1011
Flanges 7 40 700 4900 19600 1940 1052 2.61*107 2.2 *1011 2.2
*1011
Total IRd= 5.5
*1011
Since WRd>WEd, the design is sufficiently strong, with a unity check of 0.67. The displacements
follow from Matrixframe and it equal to 73 mm, leading to a unity check of 0.97.
96
Figure C 13: Deformation total gate long direction
below.
Elements Amount Thickness Width Total Total Distance Distance Iown Isteiner Itotal
per width area to top to NC
element fibre
[-] [mm] [mm] [mm] [mm2] [mm] [mm] [mm4] [mm4] [mm4]
Skin 1 20 30000 30000 600000 10 721 2 *107 3.1 *1011 3.1
plate *1011
Posts 5 1900 30 150 285000 970 239 8.6*1010 1.6*1010 1.0
*1011
Stiffeners 18 400 20 360 144000 1720 989 1.9*109 1.4*1011 1.4*1011
Flanges 4 40 920 4600 18400 1940 1209 2.5*107 2.7 *1011 2.7
*1011
Total IRd= 8.3 *1011
97
𝐼𝑅𝑑 8.3 ∗ 1011 𝑚𝑚4
𝑊𝑅𝑑 = = = 6.7 ∗ 108 𝑚𝑚3
ℎ𝑑𝑜𝑤𝑛 1229 𝑚𝑚
Since WRd>WEd, the design is sufficiently strong. The displacements follow from Matrixframe
and is equal to 0.2 mm. The unity checks might hint towards an overdesigned gate but the
strength and deformation of the skin plate would prove to be governing and requires all these
elements to be present.
Since WRd>WEd, the design is sufficiently strong with a unity check of 0.31, the maximum
deflection each span is equal to:
𝐿 1500
𝛿𝑚𝑎𝑥 = = = 3.75 𝑚𝑚
400 400
98
Matrixframe gives the following deflection, with a maximum value of 3.5 millimetres, meaning
a sufficiently stiff design, the unity check is 0.93.
The bending moment line as a result of the loading can be seen in the figure below.
The maximum bending moment can be found from this figure:
𝑀𝐸𝑑 = 48.44𝑘𝑁𝑚
For a steel quality S355, this leads to a required section modulus for the skin plate of:
𝑀𝐸𝑑 4.84 ∗ 107
𝑊𝐸𝑑 = = = 1.36 ∗ 105 𝑚𝑚3
𝜎 355 𝑁/𝑚𝑚2
From table 2, we know that the moment of Inertia of the skin plate is equal to I=9*106 mm4 and
since hup =hdown=10 mm, the actual section modulus of the skinplate equals to:
𝐼𝑅𝑑 9 ∗ 106 𝑚𝑚4
𝑊𝑅𝑑 = = = 9 ∗ 105 𝑚𝑚3
ℎ𝑑𝑜𝑤𝑛 10 𝑚𝑚
Since WRd>WEd, the design is sufficiently strong with a unity check of 0.15, the maximum
deflection each span is equal to:
𝐿 1750
𝛿𝑚𝑎𝑥 = = = 4.375 𝑚𝑚
400 400
Matrixframe gives the following deflection, with a maximum value of 3.8 millimetres, meaning
a sufficiently stiff design, the unity check is 0.87.
99
Appendix D: Development of design
alternatives
This appendix collects all background information and design calculations for the development
of the design alternatives, starting off with the MCA for the sedimentation reduction strategies
as described in chapter 5.2.1.
The criteria that are included in the MCA for the sedimentation reduction strategies are:
• Constructability: the ease of construction of the mitigation measure, including the
required construction expertise and knowledge for realization.
• Effectiveness: how well does each alternative manage to prevent the inflow of
sediment or enable for an easy disposal of the inflowing sediment. Since this is not
accurately known for each measure, an qualitative estimation will be made.
• Costs: Including the cost of construction and the costs of operation and maintenance.
• Sustainability: The environmental impact that is involved with this mitigation measure
which includes the required amount of energy, emissions during construction and
maintenance and total lifetime of the solution.
• Maintenance: Frequency of maintenance and maintenance accessibility.
Each of these criteria is then given a weight based on their relative importance, effectiveness
is chosen to be the most important criteria and is given a weight of 40, with feasibility and
sustainability being assigned a weight of 20 and costs and maintenance a weight of 10. This
distribution of the weights is a subjective decision and is based on the time and money that
can be saved when the sediment clean up works are effectively prevented and considering the
additional environmental impact that is involved with the disposal of the sediment after it has
been cleaned up, since the sediment is typically heavily contaminated and cannot be disposed
back into the harbor.
Each mitigation measure will be scored on a scale of 1-5 for each of the 5 criteria and a
weighted average is then taken to determine a ranking among the alternatives. Table 5 below
shows the scores given and the results of the MCA.
Alternative
Criteria Weight Water inlet Jetty Sediment Bubble Sediment
obstacle screen trap
Constructability 20 3 2 3 4 5
Effectiveness 40 5 4 1 4 2
Costs 10 2 2 4 2 3
Sustainability 20 3 3 3 1 1
Maintenance 10 4 5 3 2 2
Total 3.8 3.3 2.3 3 2.5
score
The sediment trap option is given the highest score for the constructability criteria since it does
not include the design of a structure of some sort, merely the excavation of the trap in front of
the gate is sufficient. The other options are not extremely complex engineering projects but the
constructability of the jetty is considered to be worse due to the possible conflicts with port
regulations. The effectiveness of the water inlet is assumed to be best but should be tested,
for example by setting up a physical model. The sediment obstacle option scores worst on the
effectiveness criteria due to the turbulence of the flow, but is the cheapest option construction
100
wise and the same can be said for the siltation trap. The water inlet and jetty would require a
significant amount of construction costs, the bubble screen is costly during operation
procedures and the sediment trap requires additional costs during maintenance when the
sediment trap has to be excavated again after a period of time. As far as the sustainability of
the alternatives are considered, the water inlet, jetty and sediment obstacles would mainly
cause emissions during construction but no additional energy is necessary and the lifetime of
these concepts is the longest. The bubble screen requires continuous energy and is expected
to have a shorter lifetime, the sediment trap requires excavations works every few years and
a lot of contaminated sediment will need to be disposed of, these two alternatives therefore
score worst for this criteria. Finally, the least amount of maintenance is required for the jetty
and water inlet, the latter might only require the disposal of residue sand near the inlet and
outlet of the pipeline. The bubble screen has a risk of clogging and maintenance works will
have to be performed to the power unit of the installation. As mentioned before, the sediment
trap would require continuous periodical maintenance and therefore also scores low on this
criteria.
Steel fibres
For the steel fibre reinforced UCF, the average floor thickness is taken as 800 millimetres and
so:
2 2
ℎ𝑚𝑖𝑛 = ℎ𝑎𝑣𝑔 − √(𝑡𝑜𝑙𝑢𝑛𝑑𝑒𝑟 + 𝑡𝑜𝑙𝑜𝑣𝑒𝑟 = 633 𝑚𝑚 and as mentioned before, the concrete strength
where
and
1
𝑀= ∗ 𝑞 ∗ 𝐿2𝑦
8
q is the representative value for the loads which is equal to the water pressure at a level of -
8.5 m NAP minus the self-weight of the concrete, which is equal to, which due to the decrease
of the thickness of the floor increases to:
𝑘𝑁
𝑘𝑁 23 3
𝑞 = ((8.5 𝑚 + 3.66 𝑚) ∗ 10 3 ) − (0.8 𝑚 ∗ 𝑚 ) = 104.8 𝑘𝑁/𝑚2
𝑚 1.1
Where the volumetric weight of the concrete is corrected with a material factor 1.1.
This leads to Ly,max=2.56 meters, so a value of 2.5 meters is still sufficient.
101
Next, check G is performed which concerns the punching shear check of the connections
between the anchors and the concrete floor. The following must be true (CUR-Aanbeveling 77,
2014):
VEd is equal to the upward pressure qEd multiplied by the area that is carried by each tension
pile. Lx is chosen to be 2.4 metres, this means:
𝑉𝐸𝑑 = 𝑞𝐸𝑑 ∗ 𝐿𝑥 ∗ 𝐿𝑦 = (104.8 ∗ 1.35) ∗ 2.4 ∗ 2.5 = 849𝑘𝑁
Again, the value of VRd is given as follows:
But here, an additional component vRfd that accounts for the presence of the steel fibre can be
added, this components can be calculated as follows:
𝑓𝑒𝑞𝑘,3 𝑓𝑐𝑡𝑘,𝑠𝑓
𝑣𝑅𝑓𝑑 = 0.18 ∗ = 0.18 ∗ = 0.2 𝑁/𝑚𝑚2
1.4 ∗ 𝛾𝑓𝑡 1.4 ∗ 1.25
So the total vnew=vmin+vRfd.
The value of p is changed from 200 mm in the base design to pmin=175 mm, which results in
u1=5812 mm and since vmin=0.39 N/mm2, vnew=0.59 N/mm2 and the maximum allowable shear
force is equal to:
1.25∗849
VRd=1213 kN, leading to a u.c. of 1213 = 0.88.
Finally check B2 is performed again, where the strut force in the concrete remains the same,
but the value of z changes to 63 mm, resulting in a value for MRd= Ned* z2 = 156 kNm.
102
MEd=1/8*qd*Lx = 102 kNm meaning that this check is also verified with a unity check of 156 =
0.66.
It can therefore be concluded that the design of the UCF including the steel fibres is still
sufficiently strong.
Basalt fibres
The procedure can again be done for the basalt fibre reinforced UCF, this time an average
floor thickness of 850 millimetres is taken and so:
2 2
ℎ𝑚𝑖𝑛 = ℎ𝑎𝑣𝑔 − √(𝑡𝑜𝑙𝑢𝑛𝑑𝑒𝑟 + 𝑡𝑜𝑙𝑜𝑣𝑒𝑟 ) = 683 𝑚𝑚
And the increased concrete strength properties are:
Check A leads to a maximum value for Ly of 2.75 so the chosen c.t.c. distance of 2.5 meters
can be maintained. The representative value for the water pressure is equal to:
102
𝑘𝑁
𝑘𝑁 23 3
𝑞 = ((8.5 𝑚 + 3.66 𝑚) ∗ 10 3 ) − (0.85 𝑚 ∗ 𝑚 ) = 103.8 𝑘𝑁/𝑚2
𝑚 1.1
Next, check G is performed again but here, the additional component vRfd that accounts for the
presence of the basalt fibre can be added, this components can be calculated as follows:
𝑓𝑒𝑞𝑘,3 𝑓𝑐𝑡𝑘,𝑏𝑓
𝑣𝑅𝑓𝑑 = 0.18 ∗ = 0.18 ∗ = 0.2 𝑁/𝑚𝑚2
1.4 ∗ 𝛾𝑓𝑡 1.4 ∗ 1.25
So the total vnew=vmin+vRfd.
The value of p is changed from 200 mm in the base design to pmin=175 mm, which results in
u1=6440 mm and since vmin=0.35 N/mm2, vnew=0.55 N/mm2 and the maximum allowable shear
force is equal to:
1.25∗841
VRd=1368 kN, VRd= (qrep *1.35)*Lx *Ly=841 kN leading to a u.c. of 1368 = 0.77.
For check B2, value of z changes to 73 mm, resulting in a value for MRd= Ned* z2 = 180 kNm.
101
MEd=1/8*qd*Lx = 101 kNm meaning that this check is also verified with a unity check of 180 =
0.56.
It can therefore be concluded that the design of the UCF including the basalt fibres is still
sufficiently strong.
103
Where A is the horizontal area of the dock chamber.
By setting (5) equal to zero, the total filling time of the dock chamber can be found to be:
2 ∗ 𝐴 ∗ √ℎ0
𝑇𝑡𝑜𝑡𝑎𝑙 =
𝑚𝑠 ∗ 𝑓 ∗ √2𝑔
The total filling time can be plotted against the total area of the valves, using A = 30*150 =
4500 m2, ms=0.8, g = 9.81 m2/s and ℎ0 =7.5 m. . The value for h0 is equal to the average water
level in the harbour which is at +0.00 m NAP, this is the minimum water level to start the
docking procedure.
This plot can be seen in Figure D1 below.
We can conclude from this graph that the minimum valve area should be 5 m2 as the filling
time exponentially increases for smaller values. Taking this minimal valve area, the graphs
showing the development of the inflowing discharge Q and the water level difference between
the harbor and the dock chamber can be plotted, shown in Figure D2 and Figure D3.
The time required for filling up the dock is around 1400 seconds, which is roughly 25 minutes.
104
Appendix E: UCF loads
This appendix collects the information behind determining the design loads on the UCF.
Ship load
The ship load has been determined for the governing design vessel Kommandor Susan which
is a British supply vessel that would lead to the largest concentrated load on the floor of the
dock. The loading condition that is taken is the ballast arrival loading condition which means a
total ship displacement of 3534 tons according to the Stability Booklet that is provided by
Damen Harlingen (Marin Teknikk AS, 1999). The corresponding shear force diagram can be
seen in the figure E1 below, from this it is possible to estimated the distributed loads caused
by the weight of the vessel including the upward water pressure, since the shear force diagram
shows the nett shear force acting on the outmost fibre of the vessel hull itself in a submerged
condition which means that the water pressure must be subtracted to reach the total distributed
weight of the ship.
The nett distributed can now be retrieved by investigated the slope in the shear force diagram,
which yield:
𝑞1 = 500 𝑡 / 28 𝑚𝑒𝑡𝑒𝑟 ≈ 18 𝑡 / 𝑚
𝑞2 = 1000 𝑡 / 46 𝑚𝑒𝑡𝑒𝑟 ≈ 22 𝑡 / 𝑚
𝑞3 = 500𝑡/35 𝑚𝑒𝑡𝑒𝑟 ≈ 14 𝑡 / 𝑚
105
Figure E 2
To estimate the upward directed water pressure due to the displacement of the vessel, it is
assumed that this is distributed equally over the length of the vessel, which is equal to roughly
130 meter resulting in a distributed load of 3534/130 ≈27 t/m. This will lead to an overestimation
of the governing peak load since the displacement of the vessel is not equally distributed over
the length as well. The vessel is widest in the middle section and becomes slimmer near the
bow and the stern and so the estimation in this appendix will lead to an overly dimensioned
but structurally safe floor. The final weight distribution can therefore be retrieved by subtracting
the upward directed pressure from the diagram in figure E2, to reach the final pressure
distribution as can be seen in figure E3 below.
Blocks are typically placed at a centre to centre distance of 2 meters or less, meaning that the
governing load on one ‘row’ of blocks, consisting of a wider centre block and two edge blocks,
is equal 90 tons.
The middle block carries roughly 75% of the total load and the other 25 % of the load is
distributed over the two outer blocks, which means that qmid=45 t/m=450 kN/m and Fedge=11.25
t = 112.5 kN. The result can be seen in the diagram below.
Shrinkage
The total shrinkage of the floor consists of the autogenous and drying shrinkage, meaning that
the following equation needs to be determined:
𝜀𝑐𝑠 = 𝜀𝑐𝑎 + 𝜀𝑐𝑑
Firstly, the autogenous shrinkage strain 𝜀𝑐𝑎 is determined using the relations in NEN-EN 1992-
1-1:
106
𝜀𝑐𝑎 (∞) = 2.5(𝑓𝑐𝑘 − 10) ∗ 10−6 = 0.025 ∗ 10−3
The drying shrinkage strain is harder to determine:
𝜀𝑐𝑑 = 𝛽𝑑𝑠 ∗ 𝑘ℎ ∗ 𝜀𝑐𝑑,0
Where:
𝑓𝑐𝑚
𝜀𝑐𝑑,0 = 0.85 ∗ ((220 + 110 ∗ 𝛼𝑑𝑠1 ) ∗ exp (−𝛼𝑑𝑠2 ∗ )) ∗ 10−6 ∗ 𝛽𝑅𝐻
𝑓𝑐𝑚0
𝑅𝐻 3
𝛽𝑅𝐻 = 1.55 ∗ (1 − ( ) )
𝑅𝐻0
With:
- Fcm = 28 N/mm2
- Fcm0 = 10 N/mm2
- 𝛼𝑑𝑠1 = 4
- 𝛼𝑑𝑠2 =0.12
- 𝑅𝐻=70 %
- 𝑅𝐻0=100 %
Where t is taken as 100 years (365000 days) and ts = 28 days, leading to 𝛽𝑑𝑠 = 0.999956.
And so 𝜀𝑐𝑑 = 0.286 ∗ 10−3 and 𝜀𝑐𝑠 = 0.31 ∗ 10−3 .
The effect of this concrete shrinkage on floor can be taken into account in two ways, according
to CUR 111:
- Load case 1: an equally spread tensile stress resulting from 1.0𝜀𝑐𝑠 leading to a tensile
load of Fshrinkage=Ec * kφ* hfloor*𝜀𝑐𝑠
- Load case 2: a combination of a tensile force and a bending moment caused by a
distribution of the shrinkage strain of 0.9 εcs at the top of the floor and 0.6εcs at the
bottom of the floor, leading to a tensile force of Fshrinkage=Ec * kφ* hfloor*0.75𝜀𝑐𝑠 and
Mshrinkage = Ec * kφ* hfloor2/12 * (0.15𝜀𝑐𝑠 /h).
kφ is determined to be 0.43 according to the relations in NEN-1992-1-1 which leads to:
Load case 1: Fshrinkage=964 kN
Load case 2: Fshrinkage=723 kN and Mshrinkage=12 kNm.
Thermal load
The temperature in the dock hall can increase to a maximum of temperature of Tout,max=30 ֯C
in summer and a minimum temperature Tout,max=-25 ֯C in winter according to Table 5.2 of NEN-
EN 1991-1-1-5. The temperature at the bottom side of the floor stays constant around T in=10
107
֯C conform table 5.3. The uniform temperature component of a constructive element follow from
equation (5.1) in chapter 5.2 as follows:
𝛥𝑇 = 𝑇 − 𝑇0
Where T0 is the initial temperature of the structural element, conform Appendix A, T0=10 ֯C.
The change in temperature at either side of the floor during summer and winter is:
𝛥𝑇𝑡𝑜𝑝,𝑠𝑢𝑚𝑚𝑒𝑟 = 𝑇𝑜𝑢𝑡,𝑚𝑎𝑥 − 𝑇0 = 30 − 10 = 20 ֯C
𝛥𝑇𝑏𝑜𝑡𝑡𝑜𝑚 = 𝑇𝑖𝑛 − 𝑇0 = 10 − 10 = 0 ֯C
Leading to the following profiles over the height of the floor, which is linear according to
Appendix D.
108
This strain distribution lead to compressive forces in the concrete in summer and a bending
moment leading to compression in the top fibre. In winter however, the thermal effects lead to
a Tensile force acting on the concrete cross section with the magnitude:
1
𝐹𝑡ℎ𝑒𝑟𝑚𝑎𝑙 = ∗ 𝑏 ∗ (𝜀 ∗ 𝐸) ∗ ℎ𝑓𝑙𝑜𝑜𝑟
2
And a bending moment leading to tensile stresses in the top fibre with a magnitude of:
ℎ𝑓𝑙𝑜𝑜𝑟 ℎ𝑓𝑙𝑜𝑜𝑟
𝑀𝑡ℎ𝑒𝑟𝑚𝑎𝑙 = 𝐹𝑡ℎ𝑒𝑟𝑚𝑎𝑙 ∗ ( − )
2 3
Taking this into account during the design of the floor has lead to significant stresses in the
concrete and the requirement regarding no cracks in the top fibre of the concrete can no longer
be met. In conclusion: temperature decrease during winter can lead to cracks in the top fibre
of the concrete. Actually when the temperature inside the hall is below -6 C ͦ for an extended
period of time, the concrete tensile stress in the top fibre will exceed the cracking limits. This
further substantiates the idea of placing a hall on top of the floor where the temperature can
be regulated.
109
Appendix F: UCF Design calculations
This Appendix contains the calculations behind the design of the UCF variants.
Fatigue
The first check is performed for the reinforcement steel in the UCF, where section 6.8.6 of
NEN-EN1992-1-1 norm (NEN, 2005) mentions the following requirement:
This value for Δσs for the UCF can be found to be Δσs=42 MPa under an assumed E-modulus
for cracked concrete of E=10000 N/mm2. For the verification of concrete under compression,
the following check is performed:
With σc,min=0 N/mm2 and σc,max=7.28 N/mm2, the value for fcd,fat is computed to be:
𝑓𝑐𝑘
𝑓𝑐𝑑,𝑓𝑎𝑡 = 0.85 ∗ 𝑓𝑐𝑑 (1 − ) = 10.4 𝑁/𝑚𝑚2
250
Filling this in yields:
𝜎𝑐,𝑚𝑎𝑥
= 0.7 > 0.5
𝑓𝑐𝑑,𝑓𝑎𝑡
And so this requirement is not met and the concrete strength class should be increased to a
C30/37. However, the checks described in the norm are more applicable to short duration
cycles such as traffic loads over a bridge deck or railways, the design amount of cycles that is
prescribed is in the order of 106. For the UCF, with an expected lifetime of 100 years and an
average of 20 load cycles per year, this amount of cycles is in the order of 2000, so a factor
500 less. The effect of fatigue can therefore also be expected to be less severe, it is assumed
that the current concrete strength class C20/25 is sufficient.
Finally, the shear checks that should be performed are as follows (NEN, 2005):
110
- VRd,c=777 kN (see appendix C)
1 1 1 1
= + +
𝑘𝑡𝑒𝑛𝑠𝑖𝑜𝑛𝑝𝑖𝑙𝑒 𝑘𝑠ℎ𝑎𝑓𝑡 𝑘𝑒𝑙𝑎𝑠𝑡𝑖𝑐 𝑘𝑢𝑛𝑑𝑒𝑟𝑠𝑜𝑖𝑙
Where:
Kshaft is the spring stiffness of the shaft friction
Kelastic is the elastic extension of the pile
Kundersoil is the rise of the soil under the pile tip due to the relief
For anchor piles, the representative value for the spring stiffness of the shaft friction equals:
𝑘𝑠ℎ𝑎𝑓𝑡;𝑟𝑒𝑝 = 120 ∗ 𝑅𝑠;𝑐𝑎𝑙;𝑚𝑎𝑥
Where the value for Rs;cal;max, which is the shaft resistance of the pile calculated in the design
of the tension piles, is equal to 721 kN. This means that:
𝑘𝑠ℎ𝑎𝑓𝑡;𝑟𝑒𝑝 = 120 ∗ 721 = 86520 𝑘𝑁/𝑚
For the determination of the elastic spring stiffness of the pile it is assumed that the grout body
as a length of 10 meters so that the part of the pile that is installed in the strong sand layer is
fully grouted. The value for kelastic can be calculated as follows:
𝐸𝐴
𝑘𝑒𝑙𝑎𝑠𝑡 =
1
𝐿1 + 2 (𝐿 − 𝐿1 )
111
The modulus of elasticity of the pile is equal to 200 * 106 kN/m2 and with a pile area of 0.00317
m2, EA amounts to 634*103 kN. L1 is the ungrouted length of the pile, so L1=14.5 m and L=24.5
m. Filling this in yields:
634 ∗ 106
𝑘𝑒𝑙𝑎𝑠𝑡 = = 32500 𝑘𝑁/𝑚
1
14.5 + 2 (10)
For the determination of kundersoil, CUR77 recommends that it of secondary importance
compared to the shaft friction resistance and that is can therefore be neglected, meaning that
kundersoil=∞
This means that ktensionpile=23625 kN/m and after including a factor of variation of √2 results in:
𝑘𝑡𝑒𝑛𝑠𝑖𝑜𝑛𝑝𝑖𝑙𝑒 = 16706𝑘𝑁/𝑚
A similar procedure is followed for the combi-wall, where the shaft friction resistance has been
retrieved from the D-sheetpiling calculation and amount to Rs,cal,max=601 kN. This means that:
𝑘𝑠ℎ𝑎𝑓𝑡,𝑤𝑎𝑙𝑙 = 120 ∗ 601 = 72120𝑘𝑁/𝑚
For the elastic spring stiffness of the combi wall, the coating area of the piles and the sheet
pile section combined is found from the D-sheetpiling database and amounts to 1.33 m2/m.
This has to be transformed into a coating area per unit meter width of the wall section and so
it is divided by the length of the wall section which is equal 2.68 meters. The area of wall per
meter width and per meter height is therefore 0.5 m2/m/m. The combined section runs from -8
m NAP till -18 m NAP and so L=10 meters. L1=1 meter, which gives:
0.5 ∗ 200 ∗ 106
𝑘𝑒𝑙𝑎𝑠𝑡 = = 18181 ∗ 103 𝑘𝑁/𝑚
1
1 + ∗ (9)
2
So:
1 1 1
= +
𝑘𝑠ℎ𝑒𝑒𝑡𝑝𝑖𝑙𝑒 72120 18181 ∗ 103
Which amounts to ksheetpile=71835 kN/m, and including the factor of variation again yields:
𝑘𝑠ℎ𝑒𝑒𝑡𝑝𝑖𝑙𝑒 = 101590 𝑘𝑁/𝑚
SFRUCF – loading phase 2
For determining the cracking moment of the UCF it is assumed that the tensile stress in the
top fibre of the concrete is equal to fctm=2.21 N/mm2 and the resulting compressive strain in the
bottom fibre of the concrete εc,bottom<1.75 ‰. The corresponding stress and strain diagrams
can be seen in Figure F2 below.
Figure F 2
The E-modulus of the concrete, required for determining the strain distribution of the concrete
is:
𝑓𝑐𝑑 15.7
𝐸= = = 8968 𝑁/𝑚𝑚2
𝜀𝑐 1.75 ∗ 10−3
112
The height of the compression zone x is determined by an horizontal equilibrium of forces, so:
𝑁𝑠 − 𝑁𝑐 − 𝑁𝐸𝑑 − 𝑁𝑐,𝑡𝑟𝑒𝑘 + 𝑁𝑠ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = 0
1
𝑁𝑐,𝑡𝑟𝑒𝑘 = ∗ 𝑓 ∗ (ℎ − 𝑥) ∗ 𝑏
2 𝑐𝑡𝑚
Where b is the unit width of 1000 mm and h is the total height of the cross section, h=havg-
tolunder+ 330 mm = 980 mm.
ℎ−𝑥−𝑐
𝑁𝑠 = 𝐴𝑠 𝜎𝑠 = 𝐴𝑠 ∗ ∗ 𝜀𝑐,𝑡𝑜𝑝 ∗ 𝐸𝑠
ℎ−𝑥
Where c is the concrete cover of 60 mm and Es=200000 N/mm2.
𝑁𝑠ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = (0.75 ∗ 𝜀𝑐𝑠 ∗ 𝑘𝜑 ) ∗ 𝐸𝑐 ∗ ℎ ∗ 𝑏 = 814 𝑘𝑁
NEd = 2213 kN and so the value for x can be found to be 590 mm. Filling this into the equations
for Nc and Ns gives:
𝑁𝑐 = 986 𝑘𝑁 and 𝑁𝑠 = 215 𝑘𝑁 and the εc,bottom= 0.4‰ < 1.75 ‰.
Figure F 3
(𝐸𝑐 ∗ 𝑘𝜑 ∗ ℎ2 ) 0.15𝜀𝑐𝑠
𝑀𝑠ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = ∗ = 14.18 𝑘𝑁𝑚
12 ℎ
113
𝑀𝑐𝑟 = 𝑁𝑐 ∗ 𝑧𝑐 + 𝑁𝑠 ∗ 𝑧𝑠 + 𝑁𝑐,𝑡𝑟𝑒𝑘 ∗ 𝑧𝑐,𝑡𝑟𝑒𝑘 − 𝑀𝑠ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = 522 𝑘𝑁𝑚
The acting bending moments from the upward water pressure qrep=101 kN/m/m’ and the
moments at the supports can be seen in the M-line below from matrixframe.
The maximum bending moment is MEd=368 kNm and since this is smaller than Mcr, no cracks
are expected to occur, with a u.c. of 0.70.
For the bearing capacity of the concrete, it is assumed that the reinforcement steel yields,
meaning that εs>εyd=2.175‰, on the concrete side, the concrete compression zone has a bi-
linear shapes and εc=εcu=3.5‰, resulting in the following stress and strain diagrams:
Figure F 4
Again, a horizontal equilibrium of forces leads to the height of the compression zone xu:
𝑁𝑠 − 𝑁𝑐 − 𝑁𝐸𝑑 + 𝑁𝑠ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = 0
114
𝑁𝑐 = 3533 𝑘𝑁 , the resulting strain diagram can be used to check the first assumption of the
yielding reinforcement steel.
𝜀𝑐𝑢
𝜀𝑠 = ∗ (ℎ − 𝑥𝑢 − 𝑐) = 7.22 ‰ > 𝜀𝑦𝑑
𝑥𝑢
Figure F 5
The acting bending moments from the upward water pressure qEd=1.35*101=136 kN/m/m’ and
the moments at the supports can be seen in the M-line below from matrixframe.
The maximum bending moment is MEd=518 kNm and since this is smaller than MRd, the cross
section is sufficiently strong with a u.c. of 0.23.
SFRUCF – loading phase 3
SLS checks: The governing bending moment inside the UCF in this loading phase is equal to
MEd=654 kNm as can be seen in the bending moment line below corresponding to a
combination of the shipload and the ground water level of +0.00 m NAP since this would lead
to the largest bending moment in the bottom fibre of UCF.
115
The cracking moment Mcr is again found from an equilibrium of forces between the strut force
NEd from the combi-wall and the compressional and tensile forces in the concrete cross section.
For determining the cracking moment of the UCF it is assumed that the tensile stress in the
bottom fibre of the concrete is equal to fctm=2.21 N/mm2 and the resulting compressive strain
in the bottom fibre of the concrete εc,bottom<1.75 ‰. The corresponding stress and strain
diagrams can be seen below.
Figure F 6
The E-modulus of the concrete, required for determining the strain distribution of the concrete
is:
𝑓𝑐𝑑 15.7
𝐸= = = 8968 𝑁/𝑚𝑚2
𝜀𝑐 1.75 ∗ 10−3
The height of the compression zone x is determined by an horizontal equilibrium of forces, so:
𝑁𝑐,𝑡 − 𝑁𝑐,𝑐 − 𝑁𝐸𝑑 + 𝑁𝑠ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = 0
116
Figure F 7
MEd>Mcr and so cracks can be expected in the bottom of the floor. After the concrete is cracked,
the tensile strength of the steel fibres is activated, the stress distribution in the cross section
will include a term of these steel fibre according to CUR 111:
Figure F 8
The magnitude of the contribution of the steel fibre is therefore determined by the value of
fft,rep,2 which is determined by performing a 3 point bending test corresponding to the residual
strength of the test member at an occurring crack width of 2.5 mm. Test results from a report
that reported such 3 point bending tests with a fibre ratio in the test element of 40 kg/m3,
corresponding to roughly 1.5 volume %, is used. Using the results, a ratio between the flexural
strength of an uncracked member and fR2, which is the flexural strength of the test member at
a 2.5 mm crack width, is determined. Afterwards, a conversion factor is used from the same
report to transform fR2 into fft,rep2 which will be used in the design of the floor. The first ratio to
117
determine fR2 is done to account for the different concrete strength class between the tests in
the report and the C20/25 in the design of the concrete.
The ratio between fctm and fR,2 is found to be roughly 1.6, stemming from a summary of the test
results from the report (Rensen, 2013):
And so the value for fft,rep that is used in the design of the SFRUCF is equal to fft,rep=0.37 * fR =
1.34 N/mm2
The maximum allowable crack width is 0.2 mm and the occurring crack width in the steel-fibre
reinforced concrete can be determined by the following equation from CUR 111 (CUR, 2018):
𝑤𝑚𝑎𝑥 = 2(ℎ − ℎ𝑥 ) ∗ 𝜀𝑓𝑡,𝑚𝑎𝑥
Where hx is the height of the compression zone and 𝜀𝑓𝑡,𝑚𝑎𝑥 is the maximum tensile strain in the
SFRUCF.
As previously determined, the acting bending moment is larger than the cracking moment. The
height of the compression zone and consequently the tensile strain in the bottom will be altered
118
by inclusion of the additional bending moment term MEd-Mcr that increases the strain in the top
fibre, assuming a cracked E-modulus of 10 GPa, as follows:
𝑀𝑎𝑑𝑑𝑖𝑡𝑖𝑜𝑛𝑎𝑙 1
𝜀𝑡𝑜𝑝,𝑐𝑟𝑎𝑐𝑘𝑒𝑑 = 𝜀𝑡𝑜𝑝,𝑢𝑛𝑐𝑟𝑎𝑐𝑘𝑒𝑑 + 𝑊
∗𝐸 = 0.29‰
𝑐𝑟𝑎𝑐𝑘𝑒𝑑
After performing another equilibrium calculation with the use of Maple, the height of the
compression zone is found to be 666 mm and so the resulting crack width amounts to:
𝑤𝑚𝑎𝑥 = 2(980 − 655) ∗ 0.29‰=0.1992 mm
Figure F 10
This is an acceptable crack width, since wmax<0.2 mm. As far as the watertightness of the floor
is concerned, it must be ensured that the crack doesn’t pass through the complete height of
the floor, in other words, a compressional zone must still be present under these
circumstances. Figure F10 proves that this is indeed the case, equilibrium is reached even
under cracked conditions, and so the floor is indeed sufficiently watertight for this loading
condition.
For the ultimate bearing capacity under ULS conditions, the acting bending moment is retrieved
by applying a safety factor of 1.5 on the ship load and 0.9 on the upward water pressure,
leading to a governing bending moment of 1111 kNm, as visible on the M-line below:
119
The resisting bending moment results from an equilibrium of forces inside the the concrete
floor, including the effect of the steel fibres, as can be seen in the figure below from CUR 111.
Figure F 11
In our case, the strut force from the combiwalls has to be included and the terms T1 is not
present since there is no traditional reinforcement in the bottom of the floor. The total cross
sectional forces and stress and strain diagrams can be seen below:
Figure F 12
The height of the compression zone x is determined by an horizontal equilibrium of forces, so:
𝑁𝑐 − 𝑇𝑓𝑖𝑏𝑟𝑒 − 𝑁𝐸𝑑 + 𝑁𝑆ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = 0
120
𝑁
𝑇𝑓𝑖𝑏𝑟𝑒 = (ℎ − 𝑥) ∗ 𝑏 ∗ 1.34
𝑚𝑚2
NEd = 2213 kN, Nshrinkage=814 kN and so the value for x can be found to be:
𝑥𝑢 = 207 𝑚𝑚
Filling this into the equations for Nc:
𝑁𝑐 = 2436 𝑘𝑁 and 𝑇𝑓𝑖𝑏𝑟𝑒 = 1033 𝑘𝑁
Figure F 13
121
BFRUCF – loading phase 2
For determining the cracking moment of the UCF it is assumed that the tensile stress in the
top fibre of the concrete is equal to fctm=2.21 N/mm2 and the resulting compressive strain in the
bottom fibre of the concrete εc,bottom<1.75 ‰. The corresponding stress and strain diagrams
can be seen below.
Figure F 14
The E-modulus of the concrete, required for determining the strain distribution of the concrete
is:
𝑓𝑐𝑑 13.9
𝐸= = = 7942 𝑁/𝑚𝑚2
𝜀𝑐 1.75 ∗ 10−3
The height of the compression zone x is determined by an horizontal equilibrium of forces, so:
𝑁𝑠 − 𝑁𝑐 − 𝑁𝐸𝑑 − 𝑁𝑐,𝑡𝑟𝑒𝑘 + 𝑁𝑠ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = 0
122
NEd = 2213 kN and so the value for x can be found to be 707 mm. Filling this into the equations
for Nc and Ns gives:
𝑁𝑐 = 2023 𝑘𝑁 and 𝑁𝑠 = 224 𝑘𝑁 and the εc,bottom= 0.72‰ < 1.75 ‰.
Figure F 15
The acting bending moments from the upward water pressure qrep=101 kN/m/m’ is MEd=368
kNm and since this is smaller than Mcr, no cracks are expected to occur, with a u.c. of 0.51.
For the bearing capacity of the concrete, it is assumed that the reinforcement steel yields,
meaning that εs>εyd=2.175‰, on the concrete side, the concrete compression zone has a bi-
linear shapes and εc=εcu=3.5‰, resulting in the following stress and strain diagrams:
Figure F 16
Again, a horizontal equilibrium of forces leads to the height of the compression zone xu:
123
𝑁𝑠 − 𝑁𝑐 − 𝑁𝐸𝑑 + 𝑁𝑠ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = 0
Figure F 17
The acting bending moments from the upward water pressure qEd=1.35*101=136 kN/m/m’ is
MEd=518 kNm and since this is smaller than MRd, the cross section is sufficiently strong with a
u.c. of 0.23 .
BFRUCF – loading phase 3
124
SLS checks: The governing bending moment inside the UCF in this loading phase is equal to
MEd=654 kNm. The cracking moment Mcr is again found from an equilibrium of forces between
the strut force NEd from the combi-wall and the compressional and tensile forces in the concrete
cross section. For determining the cracking moment of the UCF it is assumed that the tensile
stress in the bottom fibre of the concrete is equal to fctm=2.21 N/mm2 and the resulting
compressive strain in the bottom fibre of the concrete εc,bottom<1.75 ‰. The corresponding
stress and strain diagrams can be seen below.
Figure F 18
The height of the compression zone x is determined by an horizontal equilibrium of forces, so:
𝑁𝑐,𝑡 − 𝑁𝑐,𝑐 − 𝑁𝐸𝑑 + 𝑁𝑠ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = 0
125
Figure F 19
MEd>Mcr and so cracks can be expected in the bottom of the floor. After the concrete is cracked,
the tensile strength of the steel fibres is activated, the stress distribution in the cross section
will include a term of these steel fibre according to CUR 111:
Figure F 20
The magnitude of the contribution of the steel fibre is therefore determined by the value of
fft,rep,2 which is determined in the same way as described for the steel fibre variant but now
applying the ratios to the strength properties of the basalt fibre reinforced concrete leading to
to fR,2=1.6*2.21 * (1.226/1.5) = 3.54 N/mm2. And so the value for fft,rep that is used in the design
of the BFRUCF is equal to fft,rep=0.37 * fR = 1.31 N/mm2
The occurring crack width in the steel-fibre reinforced concrete can be determined by the
following equation from CUR 111 (CUR, 2018):
126
𝑤𝑚𝑎𝑥 = 2(ℎ − ℎ𝑥 ) ∗ 𝜀𝑓𝑡,𝑚𝑎𝑥
Where hx is the height of the compression zone and 𝜀𝑓𝑡,𝑚𝑎𝑥 is the maximum tensile strain in the
BFRUCF.
As previously determined, the acting bending moment is larger than the cracking moment. The
height of the compression zone and consequently the tensile strain in the bottom will be altered
by inclusion of the additional bending moment term MEd-Mcr that increases the strain in the
fibre, assuming a cracked E-modulus of 10 GP, as follows:
𝑀𝑎𝑑𝑑𝑖𝑡𝑖𝑜𝑛𝑎𝑙 1
𝜀𝑏𝑜𝑡𝑡𝑜𝑚,𝑐𝑟𝑎𝑐𝑘𝑒𝑑 = 𝜀𝑏𝑜𝑡𝑡𝑜𝑚,𝑢𝑛𝑐𝑟𝑎𝑐𝑘𝑒𝑑 + 𝑊
∗𝐸 = 0.33‰
𝑐𝑟𝑎𝑐𝑘𝑒𝑑
After performing another equilibrium calculation with the use of Maple, the height of the
compression zone is found to be 689 mm and so the resulting crack width amounts to:
𝑤𝑚𝑎𝑥 = 2(980 − 671) ∗ 0.33=0.204 mm
Figure F 21
127
For the ultimate bearing capacity under ULS conditions, the acting bending moment 1111 kNm.
The resisting bending moment results from an equilibrium of forces inside the the concrete
floor, including the effect of the steel fibres, as can be seen in the figure below from CUR 111.
Figure F 22
In our case, the strut force from the combiwalls has to be included and the term T 1 is not
present since there is no traditional reinforcement in the bottom of the floor. The total cross
sectional forces and stress and strain diagrams can be seen below:
Figure F 23
The height of the compression zone x is determined by an horizontal equilibrium of forces, so:
𝑁𝑐 − 𝑇𝑓𝑖𝑏𝑟𝑒 − 𝑁𝐸𝑑 + 𝑁𝑆ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = 0
128
Where b is the unit width of 1000 mm and h is the total height of the cross section is 980 mm.
The force in the tensile zone amounts to:
𝑁
𝑇𝑓𝑖𝑏𝑟𝑒 = (ℎ − 𝑥) ∗ 𝑏 ∗ 1.31
𝑚𝑚2
NEd = 2213 kN and so the value for x can be found to be:
𝑥𝑢 = 238 𝑚𝑚
Filling this into the equations for Nc:
𝑁𝑐 = 2048 𝑘𝑁 and 𝑇𝑓𝑖𝑏𝑟𝑒 = 973 𝑘𝑁
Figure F 24
With a narrow unity check of 0.99, this ULS check is found to be governing in the design of the
floor.
The shear force VEd acting on the floor amounts to 475, from the V-line below:
129
Rebar reinforced – loading phase 2
For determining the cracking moment of the UCF it is assumed that the tensile stress in the
top fibre of the concrete is equal to fctm=2.21 N/mm2 and the resulting compressive strain in the
bottom fibre of the concrete εc,bottom<1.75 ‰. The corresponding stress and strain diagrams
can be seen below.
Figure F 25
The E-modulus of the concrete, required for determining the strain distribution of the concrete
is:
𝑓𝑐𝑑 13.3
𝐸= = = 7600 𝑁/𝑚𝑚2
𝜀𝑐 1.75 ∗ 10−3
The height of the compression zone x is determined by an horizontal equilibrium of forces, so:
𝑁𝑐,𝑐𝑜𝑚𝑝𝑟𝑒𝑠𝑠𝑖𝑜𝑛 + 𝑁𝑆,𝑐𝑜𝑚𝑝𝑟𝑒𝑠𝑠𝑖𝑜𝑛 − 𝑁𝑐,𝑡𝑒𝑛𝑠𝑖𝑜𝑛 − 𝑁𝑆,𝑡𝑒𝑛𝑠𝑖𝑜𝑛 − 𝑁𝐸𝑑 + 𝑁𝑆ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = 0
Where b is the unit width of 1000 mm and h is the total height of the cross section, h=havg-
tolunder = 700 mm.
𝑥−𝑐
𝑁𝑆,𝑐𝑜𝑚𝑝𝑟𝑒𝑠𝑠𝑖𝑜𝑛 = 𝐴𝑠,𝑏𝑜𝑡𝑡𝑜𝑚 ∗ ∗𝜀 ∗ 𝐸𝑠
ℎ − 𝑥 𝑐,𝑡𝑜𝑝
ℎ−𝑥−𝑐
𝑁𝑆,𝑡𝑒𝑛𝑠𝑖𝑜𝑛 = 𝐴𝑠,𝑡𝑜𝑝 ∗ ∗ 𝜀𝑐,𝑡𝑜𝑝 ∗ 𝐸𝑠
ℎ−𝑥
130
𝑁𝑆ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = (0.75 ∗ 𝜀𝑐𝑠 ∗ 𝑘𝜑 ) ∗ 𝐸𝑐 ∗ ℎ ∗ 𝑏 = 480 𝑘𝑁
NEd = 2213 kN and so the value for x can be found to be 493 mm. Filling this into the equations
the other force terms gives.
𝑁𝑐,𝑐𝑜𝑚𝑝𝑟 = 1419 𝑘𝑁 and 𝑁𝑠,𝑐𝑜𝑚𝑝𝑟𝑒𝑠𝑠𝑖𝑜𝑛 = 731 𝑘𝑁 and the εc,bottom= 0.75‰ < 1.75 ‰.
Figure F 26
The acting bending moments from the upward water pressure qrep=102.3kN/m/m’ and the
moments at the supports can be seen in the M-line below from matrixframe.
131
The maximum bending moment is MEd=377 kNm and since this is smaller than Mcr, no cracks
are expected to occur, with a u.c. of 0.66.
For the bearing capacity of the concrete, it is assumed that the reinforcement steel yields,
meaning that εs,top>εyd=2.175‰, on the concrete side, the concrete compression zone has a
bi-linear shapes and εc=εcu=3.5‰, resulting in the following stress and strain diagrams:
Figure F 27
Again, a horizontal equilibrium of forces leads to the height of the compression zone xu:
𝑁𝑐 + 𝑁𝑠,𝑐𝑜𝑚𝑝𝑟𝑒𝑠𝑠𝑖𝑜𝑛 − 𝑁𝑡𝑒𝑛𝑠𝑖𝑜𝑛 − 𝑁𝐸𝑑 + 𝑁𝑠ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = 0
132
The resulting strain diagram can be used to check the first assumption of the yielding
reinforcement steel as is required according to Eurocode norm 1992-1-1 (NEN-EN1992-1-1,
1992):
𝜀𝑐𝑢
𝜀𝑠 = ∗ (ℎ − 𝑥𝑢 − 𝑐) = 10.16 ‰ > 𝜀𝑦𝑑
𝑥𝑢
Figure F 28
The acting bending moments from the upward water pressure qEd=138 kN/m/m’ and the
moments at the supports can be seen in the M-line below from matrixframe.
The maximum bending moment is MEd=527 kNm and since this is smaller than MRd, the cross
section is sufficiently strong with a u.c. of 0.31.
Rebar reinforced – loading phase 3
SLS checks: The governing bending moment inside the UCF in this loading phase is equal to
MEd=660 kNm as can be seen in the bending moment line below corresponding to a
combination of the shipload and the ground water level of +0.00 m NAP since this would lead
to the largest bending moment in the bottom fibre of UCF.
133
The cracking moment Mcr is again found from an equilibrium of forces between the strut force
NEd from the combi-wall and the compressional and tensile forces in the concrete cross section.
For determining the cracking moment of the UCF it is assumed that the tensile stress in the
bottom fibre of the concrete is equal to fctm=2.21 N/mm2 and the resulting compressive strain
in the bottom fibre of the concrete εc,bottom<1.75 ‰. The corresponding stress and strain
diagrams can be seen below.
Figure F 29
Where b is the unit width of 1000 mm and h is the total height of the cross section, h=havg-
tolunder = 700 mm.
𝑥−𝑐
𝑁𝑆,𝑐𝑜𝑚𝑝𝑟𝑒𝑠𝑠𝑖𝑜𝑛 = 𝐴𝑠,𝑏𝑜𝑡𝑡𝑜𝑚 ∗ ∗𝜀 ∗ 𝐸𝑠
ℎ − 𝑥 𝑐,𝑡𝑜𝑝
ℎ−𝑥−𝑐
𝑁𝑆,𝑡𝑒𝑛𝑠𝑖𝑜𝑛 = 𝐴𝑠,𝑡𝑜𝑝 ∗ ∗ 𝜀𝑐,𝑡𝑜𝑝 ∗ 𝐸𝑠
ℎ−𝑥
134
NEd = 2213 kN and so the value for x can be found to be 497 mm. Filling this into the equations
the other force terms gives.
𝑁𝑐,𝑐𝑜𝑚𝑝𝑟 = 1473 𝑘𝑁 and 𝑁𝑠,𝑐𝑜𝑚𝑝𝑟𝑒𝑠𝑠𝑖𝑜𝑛 = 673𝑘𝑁 and the εc,top= 0.78‰ < 1.75 ‰.
Figure F 30
135
The maximum allowable crack width is 0.2 mm, following a frequent load combination for XS3
according to Table 7.1N in the NEN-EN 1992-1-1 and the occurring crack width in the
reinforced concrete can be determined by the following equation from the same norm (NEN-
EN1992-1-1, 1992):
In our case, we have long term loading under the stabilized cracking stage, meaning that α=0.3
and β=1, τbm=2fctm. The effective reinforcement ratio has to be used in this equation, which
takes into account the effective tensile zone under bending:
ℎ−𝑥
ℎ𝑐,𝑒𝑓𝑓 = min (2,5(ℎ − 𝑑); ( )) = 117 𝑚𝑚
3
This is an acceptable crack width, since wmax<0.2 mm. As far as the watertightness of the floor
is concerned, it must me ensured that the crack doesn’t pass through the complete height of
the floor, in other words, a compressional zone must still be present under these
circumstances.
136
For the ultimate bearing capacity under ULS conditions, the acting bending moment is retrieved
by applying a safety factor of 1.5 on the ship load and 0.9 on the upward water pressure,
leading to a governing bending moment of 1117 kNm, as visible on the M-line below:
The resisting bending moment results from an equilibrium of forces inside the the concrete
floor, including the effect of the steel fibres, as can be seen in the figure below:
Figure F 31
The height of the compression zone x is determined by an horizontal equilibrium of forces, so:
𝑁𝑐 + 𝑁𝑆,𝑐𝑜𝑚𝑝𝑟 − 𝑁𝐸𝑑 − 𝑁𝑆,𝑡𝑒𝑛𝑠𝑖𝑜𝑛 + 𝑁𝑠ℎ𝑟𝑖𝑛𝑘𝑎𝑔𝑒 = 0
137
The resulting strain diagram can be used to check the first assumption of the yielding
reinforcement steel, as is required according to Eurocode norm 1992-1-1 (NEN-EN1992-1-1,
1992):
𝜀𝑐𝑢
𝜀𝑠 = ∗ (ℎ − 𝑥𝑢 − 𝑐) = 8.08 ‰ > 𝜀𝑦𝑑
𝑥𝑢
Figure F 32
The maximum bending moment is MEd=1117 kNm and since this is smaller than MRd, the cross
section is sufficiently strong with a u.c. of 0.64.
The shear force VEd acting on the floor amounts to 478kN, from the V-line below:
138
Appendix G: Sediment reduction
strategies
This Appendix contains additional information regarding the sediment reduction strategies. n
Inlet short side
The dimensions of the pipeline running from the port into the dock are determined by the
required discharge that has to flow through the pipe to prevent the inflow of sediment into the
dock chamber. This Qreq is determined by the displacement of the vessel and the velocity of
the vessel when it leaves the dock. The nett cross sectional area of the design vessel is
determined by the following equation:
𝐴𝑠ℎ𝑖𝑝 = 𝐷𝑟𝑎𝑢𝑔ℎ𝑡 ∗ 𝑊𝑖𝑑𝑡ℎ ∗ 𝐶𝑀
Where the draught of the ship is determined for a Ballast-Departure loading condition of 3.5
meters and the CM is the midship coefficient, retrieved from the Hydrostatical Calculations
chapter of the Ship’s Stability booklet, for a trim of +0.00 m and a draught of 3.5 meters,
Cm=0.997 and Width = 20 meters. While this is the cross sectional area for the midship section
and the hull shape becomes slimmer near the fore and the aft of vessel, the results will give
the extreme value of the displacement for which the inlet system will be designed. Filling in
these values yields:
𝐴𝑠ℎ𝑖𝑝 = 3.5 ∗ 20 ∗ .997 = 69.3𝑚2
In order to determine the flow rate entering the dock as the vessel leaves, this nett cross section
needs to be multiplied with the velocity of the vessel during undocking. Based on conversation
with Damen employees, this velocity is estimated at 0.1 m/s, resulting in a Qreq=69.3*0.1=6.93
m3/s.
The mean velocity of the return current ur flowing into the dock can now be determined as well,
under average docking conditions (water level at +0.00 m NAP):
𝑄𝑟𝑒𝑞 = 𝑄𝑖𝑛 = 𝐴𝑑𝑜𝑐𝑘,𝑛𝑒𝑡𝑡 ∗ 𝑢𝑟
With:
𝐴𝑑𝑜𝑐𝑘,𝑛𝑒𝑡𝑡 = 𝐴𝑑𝑜𝑐𝑘 − 𝐴𝑠ℎ𝑖𝑝 = (7.5 ∗ 30) − 69.3 = 225 𝑚2
𝑢𝑟 = 0.045 𝑚/𝑠
The flow rate entering the dock through the inlet pipes must at least be equal to Qreq. The pipe
transferring this flow from the port into the dock consists of coarse concrete pipe with a
diameter of 2 meters, that bifurcates into 3 smaller pipes to spread the inflow of water along
the short side of the chamber.
With a diameter of D=2 m and thus a cross sectional area of the pipe of Apipe=3.14 m2, the
required flow velocity near the outlet of the pipe is:
𝑄𝑟𝑒𝑞
𝑉𝑟𝑒𝑞 = = 2.21 𝑚/𝑠
𝐴𝑝𝑖𝑝𝑒
In terms of velocity head:
2
𝑉𝑟𝑒𝑞
𝐻𝑟𝑒𝑞 = = 0.25 𝑚
2∗𝑔
139
Along the pipelines, a number of head losses will cause the loss of velocity. In order to meet
the required head at the entrance, the following equation for the velocity head at the inlet of
the pipeline Hin needs to be met:
𝐻𝑟𝑒𝑞 + 𝛥𝐻 = 𝐻𝑖𝑛
Where the head loss is a some of all minor and friction losses:
2
𝑣𝑖𝑛
𝛥𝐻 = (∑ 𝐾𝐿 ) ∗
2∗𝑔
Where ∑ 𝐾𝐿 consists of losses at the inlet, friction losses over the length of the pipe and a 90
degree bend.
The loss coefficient for a slightly rounded entrance is KL 0.2 (source hydraulic manual). For a
90 degree bend with a factor R/D=2, where R is the radius of the bend and D is the diameter
of the pipe, and a factor ε=0.25 mm for coarse concrete, the loss coefficient is found to be
KL=0.2, source: vano engineering.
The major loss in terms of friction will be determined by using the Darcy-Weisbach equation:
𝐿 𝑢2
𝛥𝐻𝑓 = 𝑓
𝐷 2𝑔
Where f is a friction factor, L is the length of the pipe section and D is the diameter of the pipe.
The length of the pipe section has an estimated length of 80 meters, as can be seen in Figure
G1 below.
Figure G 1
140
For determining the friction factor f, the Moody diagram is used, which also requires the
Reynolds number that in turn depends on the flow velocity in the pipe. Iterations have been
performed based on a first estimated flow velocity of 10 m/s, which eventually leads to a friction
factor of 0.012, for a Re=1.27*107 and ε/D = 1.25*10-5.
2𝑔𝐻𝑟𝑒𝑞 𝑚
𝑢𝑖𝑛 = √ = 6.37 𝑎𝑛𝑑 𝐻𝑖𝑛 = 2.1 𝑚
0.12 𝑠
Meaning that the chosen pump system needs to have sufficient capacity to add 2.1 m of head
to a flow rate of 7 m3/s. The centreline of the pipe should be installed at a depth of -4 m NAP
at the port side and while the calculations have be made for a pipeline that is assumed to be
horizontal along its trajectory, it is also possible to apply a certain slope as the pipeline head
towards the dock and install the outlet of the pipe at -6 m NAP for example, as this would also
decrease the required pump head. A number of gates should be installed near the outlets, for
example by applying a ‘knife gate valve’.
141
Appendix H: LCA details
Firstly, the source of information for the Life Cycle Inventory (LCI) is further explained in the
following paragraphs.
In the following sub-sections, the emissions factors per relevant lifetime phase are listed in
tables.
Since only limited information is available in the DuboCalc database, the emission factors for
the steel piles, sheet piles and anchors have been adjusted to better represent the profiles
applied in the dock design. For example, DuboCalc has data for the use of steel pile with a
diameter of 700 mm and a wall thickness of 12 mm, a factor of 2 is added to this emission
factor to represent the larger profile used in our design since more steel is required. The same
is done for all other elements for which the data in the DuboCalc library are not exactly
matching the design.
For the reinforced concrete the information from an Environmental Product Declaration (EPD),
which contains all LCA information for one specific product of one specific distributor, for
reinforcement steel is combined with information from the EPD from a ready-mix concrete
distributor. While the question arises whether the reinforcement steel will actually be used from
this Italian distributor, the information does give an indication. The EPD for the reinforcement
steel mentions a total GWP of 587 kg CO2-eq/ton steel, which equal to 4579 kg CO2-eq/m3
steel (Feraldi group, 2011), and for C25/30 ready mix concrete containing CEMII cement the
GWP value is 261 kg CO2-eq/m3 (Readymix Industries Ltd., 2022). Again, the strength class
and cement type might not complete represent the one used in the design and so the accuracy
of this data is debatable. However, for this conceptual design it is assumed to be sufficient.
Combining this data, together with a 1.5% reinforcement ratio leads to a value of 0.33 kg CO2-
eq/dm3
The steel coating will be applied to the inner side of the combiwall that is exposed inside the
dock chamber, the complete inner side of the dock gate and on the outer side of the dock gate
from -2 m NAP till the top of the dock gate. This is because the coating will only be applied to
142
steel components that are susceptible to corrosion by being in alternating contact with water
and air. Steel elements that are continuously in contact with water (the bottom part of the outer
side of the dock gate) or in continuous contact with the soil (the outside of the combi wall) will
be protected by applying a cathodic protection.
For the pumping unit, the total emissions per pump is determined by taking the emission factor
per kg of crude iron steel (2.35 kg CO2-eq per kg, (IDEMATapp, 2023)) and the process of
rolling steel (2.35 kg CO2-eq per kg, (IDEMATapp, 2023)) multiplied by the weight of a single
pump which is estimated to be 225 kg (KSB, 2022) . In this estimation, processes such as
painting the steel, manufacturing of the power unit and other aspects are disregarded. The
pump of choice is a Amarex KRT K pump which is a centrifugal pump with a maximum capacity
of 10000 m3/h, 2 pumps are chosen for emptying the dock chamber.
The same emission factors for the dock pump are also applied for the inlet pumps (variant A)
and the hydro jets (variant B) for simplicity even though the capacity of these pumps can be
chosen to be less than the inlet pumps, the weight of these pumps are expected to be of a
similar order of magnitude.
For the winches (the dock winches in Variant C and the gate-operating winches in Variant B)
the emissions are determined in a similar way as the pumps. The chosen winches are so-
called tugger winches with a weight of 650 kg each (Damen Marine Components, 2022).
Variant C requires 6 of these in total and Variant B require 2.
The emissions of the hydraulic motor are taken from an LCA that has been performed by Johan
Sagström for a hydraulic motor with a capacity of 45 kW and a weight of 1200 kg (Sagström,
2017).
143
For the materials, their place of origin has been chosen as either the most commonly applied
distributor in practice (steel items from the TATA-steel factory in IJmuiden, cement from the
ENCI IJmuiden, Gravel from near the Meuse river in Limburg), or by searching for the nearest
fabricator of the structural elements. Many of these structural elements such as the combi-
wall, anchors and fibres originate from Germany. The LCA data from a EPD from an Italian
distributor is used and so the reinforcement steel is assumed to originate from this factory.
The steel that is used for the construction of the gate is first transported from IJmuiden to Heeg,
where the gate is made at Nauta Heeg and then transported from Heeg to Harlingen.
The pumps and hydro jets are produced at Pentair Nijhuis Pompen BV in Winterswijk, the
winches at Damen Marine Components in Hardinxveld-Giessendam and the hydraulic motor
at Holland Hydraulics BV in Hengelo. Finally, the concrete pipe is constructed at LB
Betonproducten in Burgum.
Installation stage
Table H3 shows the different installation activities, the equipment required and the fuel/energy
consumption of the machinery. In every case, the more climate friendly electrical option is
chosen if available.
Table H 3: Installation phase inventory
144
Placement of Placement Telescopic crane 1500 kg/h 400 All
structural floor of
reinforceme
nt
Mitre gate driving system Mobile crane 1 unit per day 101 C
installation
The value for the productivity is needed for calculating the total amount of energy consumption
for the construction of the dock. For equipment running on fuel, the emission factor for diesel
is used for simplicity. In practise, some equipment could potentially also run on more
sustainable fuel alternatives such as HVO, but this has been neglected. To estimate the diesel
consumption per hour, it is assumed that diesel generators consume around 0.2 litres/kW/hour
(Ricardo-Engine, 2019).
Operational phase
Per Variant, an overview of what the operational phase comprises can be found in the tables
below.
Table H 4: Operational phase – Base design
145
Gate operation Telescopic crane 0.5 hours per 505 Electric telescopic
docking activity crane with 450 t
capacity
Dock emptying Pumps 3 hours per docking 800
Inlet operation Inlet pumps 1 hour per docking 400
activity
Tugboats support during Tugboat 2 h/docking activity 94 1 tugboat running at
docking 20% of max capacity
Maintenance stage
Each of the design variants require a regulated maintenance strategy for the pumping unit. It
is chosen to perform dock pump inspections every 5 years, where the emissions associated
with this are estimated to be 5% of the required estimations to construct, transport and install
a new pump unit. The dock pumps unit has an expected lifetime of 25 years and so new pumps
need to be installed three times through the lifetime of the dock. Other ‘moving’ parts of the
design variant will need to be regularly inspected and replaced as well, this maintenance
strategy can be seen below. The steel elements that need to be coated will be recoated every
20 years.
Table H 8: Maintenance strategy
146
Hydraulic motor replacement Every 25 years C
The material that contributes most to the total emissions is the reinforced concrete, which is
mainly due to the contribution of cement and the fact that a significant amount of this material
is required in all of the designs. It makes sense to target this main contributor when trying to
decrease the emissions of the design, which is done in the design variants by opting for a
different dock floor build-up.
The following step is calculating the amount of emissions associated with transportation of the
products from the place of manufacturing to the construction site. The table below shows the
total transport emissions for the base design.
147
Table H 10: Transport phase – Base design
Activity Equipment Productivity Diesel Energy Hours Total diesel Total Total Contr
consumption consumption consumption Electricity emissions ibutio
[Litres/hour] [kWh/h] [litres] consumption [ton CO2-eq] n
[kWh]
Pile Installation Hydraulic crane 2 piles per day - 141 500 70500 11.8 6%
installation Vibratory Hammer 2 piles per day 48 - 500 24214 39.4 20%
Welding Welding aggregate 2 piles per day 9 62.5 561 0.9 <1%
Sheet pile wall Installation Hydraulic crane 7 meters per day 141 411 58011 9.8 5%
installation Vibratory hammer 7 meters per day 48 411 19925 32.4 17%
Welding Welding aggregate 2 sections per hour 9 62 557 0.9 <1%
Finishing Cleaning Pressure washer 64 locks per day 0.8 31 25 0.04 <1%
locks Aerial work platform 64 locks per day 12 31 372 0.06 <1%
Welding Welding aggregate 2 sections per hour 9 62 557 0.9 <1%
Aerial work platform 2 sections per hour 12 62 744 0.1 <1%
Excavation Cutter suction 7000 m3/hour 594 9 5159 8.4 4%
dredger
Anchor/tension pile installation Installation rig 1 anchor per hour 49 844 41356 67.3 35%
Grout pump 1 anchor per hour 0.8 844 675 1.1 <1%
UCF pouring Concrete pump 100 m3/hour 15 45 662 1.1 1%
Dewatering Pumping Drainage pump 6000 m3/day 10 90 900 1.465 1%
Crane vessel 24 90 2120 3.50 2%
Cleaning High pressure 200 m2/day 0.8 180 144 0.2 <1%
washer
Sediment removal 85 m2/day 24 53 1249 2 1%
machine
3
Land filling Hydraulic crane 1300 m /day 141 14 1952 0.3 <1%
Placement of Placement of Telescopic crane 1500 kg/h 400 177 70650 11.9 6%
structural floor reinforcement
Pouring Concrete pump 100 m3/hour 15 22.5 331 <1% 0%
concrete
Gate placement Floating crane 1 unit per day 132 8 1056 0.2 <1%
Pump placement Mobile crane 1 unit per day 101 8 807 0.1 <1%
Total 194.6 100%
The combination of large volumetric weight and distance make the transport of Gravel from
Limburg to Harlingen the predominant factor in the emissions during transportations. In
absolute terms, the total emissions during this phase are only a fraction of the material
construction phase but when aiming to decrease the transport emissions, reduction in the
amount of concrete needed will again be the most effective.
Table H 11: Installation phase – Base design
Materials Total Place of origin Transport Distanc Emissio Unit Total Contributio
amount mode e [km] n Factor emissions n
[ton] [ton CO2-
eq]
Concr Cement 972 IJmuiden Barge (river) 120 0.203 kg CO2- 2.4 3%
ete eq/t.10km
Sand 6300 Harlingen Barge (river) 2 0.203 kg CO2- 0.5 <1%
eq/t.10km
Gravel 7155 Limburg Barge (river) 350 0.203 kg CO2- 50.8 69%
eq/t.10km
Steel Reinforcement 265 Brescia (Italy) Barge (river) 1200 0.203 kg CO2- 6.4 9%
steel eq/t.10km
Steel piles 1026 Essen (Germany) Barge (river) 283 0.203 kg CO2- 5.9 8%
eq/t.10km
Sheet piles 323 Essen (Germany) Barge (river) 283 0.203 kg CO2- 1.9 3%
eq/t.10km
Gate 270 IJmuiden to Heeg Truck 127 0.078 kg CO2-eq/t.km 2.7 4%
270 Heeg to Harlingen Barge (river) 45 0.203 kg CO2- 0.2 <1%
eq/t.10km
Anchors & 506 Langefeld Barge (river) 319 0.203 kg CO2- 3.3 4%
tension piles (Germany) eq/t.10km
Pumps 0.466 Winterswijk Truck 290 0.078 kg CO2-eq/t.km 0.007 <1%
Total 74 100%
The main contributors in the installation phase are the installing of the anchors and tension
piles and the combi-wall. These elements form the basis of the design of the dock for which
optimizations will not be done.
By choosing electrical equipment where possible, 15% of installation emissions can be
saved compared to traditional equipment running on diesel.
148
Table H 12: operational phase – base design
Total 79 100%
The emissions during operations are based on 20 ships entering and leaving the dock on
average per year, where emptying the dock requires 3 hours and docking procedures itself
require around 1 hour per activity. The table shows the emissions per year, so during a lifetime
of 100 years, 5800 tons of CO2-eq are expected, where transport and disposal of the sediment
is the main contributing factor. Mitigation measures that aim to prevent the need for cleaning
procedures could therefore prove to be effective.
The emissions associated with the maintenance phase can be seen in table H13.
Table H 13: Maintenance phase base design
Activity Equipment Productivity Diesel Energy Hours per Emissions per Contributi
year [ton CO2- on
consumptio consumptio year eq/year]
n n [kWh/h]
[Litres/hour
]
Dock emptying Pumps 3 hours per 800 60 16.2 23%
docking
Sediment Cleaning High pressure 10000 m2/day 0.8 72 0.2 <1%
removal washer
Sediment 85 m3/h 24 53 4.1 6%
removal
machine
Sedimen Truck 270 tons - 25.2 36%
t
transport
Tugboats support during Tugboat 2 h/docking activity 94 40 24.4 35%
docking
Total 70 100%
The limited amount of mechanical structures and moving parts lead to a small contribution of
the maintenance phase of the base design.
Impact assessments – Variant A
The same calculation procedures are now performed for the different design variants, starting
with Variant A.
Compared to the base design, this variant consists of less reinforced concrete, UCF concrete
but adds the contribution of steel fibres and the construction of the inlet.
149
Table H 14: Material phase – variant A
Materials Total Unit Emission Unit Total Contribution
amount factor emissions
[ton CO2-eq]
Reinforced 14850000 [dm3] 0.33 kg CO2-eq / dm3 4900 42%
concrete
Steel piles 2775 [m’] 1196 kg CO2-eq / m’ 33181 29%
Sheet pile wall 174 [m’] 554 kg CO2-eq / m’ 96 1%
Steel 269622 [kg] 0.65 kg CO2-eq / kg 175 1%
Steel coating 4509 [m2] 4 kg CO2-eq / m2 18 <1%
UCF 3600 [m3] 105 kg CO2-eq / m3 378 3%
Anchors 124 [# of anchors] 2393 kg CO2-eq / 297 3%
anchor
Tension piles 720 [# of piles] 2393 kg CO2-eq / pile 1723 15%
Pumps 2 [# of pumps] 724 kg CO2-eq / pump 1.5 <1%
Steel fibre 423900 [kg] 1.6 kg CO2-eq / kg 678.2 6%
Concrete pipe 80 [m’] 238.2 kg CO2-eq / m’ 19 <1%
Inlet pumps 2 [# of pumps] 724 kg CO2-eq / pump 1.5 <1%
Total 11605 100%
The total amount of emissions is significantly lower than the base design, stemming from the
fact that less reinforced concrete is needed which is the predominant factor in this lifecycle
phase. The inclusion of the steel fibres and inlet construction are significantly less than the
effects from concrete saving floor design.
The emissions during the transport phase of Variant A can be seen in Table H15
Table H 15: Transport phase – variant A
Materials Total Place of origin Transport Distance Emission Unit Total Contrib
amount mode [km] Factor emissions ution
[ton] [ton CO2-
eq]
Concrete Cement 732 IJmuiden Barge (river) 120 0.203 kg CO2- 2.4 3%
eq/t.10km
Sand 2034 Harlingen Barge (river) 2 0.203 kg CO2- 0.5 <1%
eq/t.10km
Gravel 5390 Limburg Barge (river) 350 0.203 kg CO2- 50.8 60%
eq/t.10km
Pipe 137 Burgum Barge (river) 45 0.203 kg CO2- 0.1 <1%
eq/t.10km
Steel Reinforceme 175 Brescia (Italy) Barge (river) 113 0.203 kg CO2- 4.2 7%
nt steel eq/t.10km
Steel piles 1026 Essen (Germany) Barge (river) 283 0.203 kg CO2- 5.9 9%
eq/t.10km
Sheet piles 323 Essen (Germany) Barge (river) 283 0.203 kg CO2- 1.9 3%
eq/t.10km
Gate 270 IJmuiden to Heeg Truck 127 0.078 kg CO2- 2.7 4%
eq/t.km
270 Heeg to Harlingen Barge (river) 45 0.203 kg CO2- 0.2 <1%
eq/t.10km
Anchors & 506 Langefeld Barge (river) 319 0.203 kg CO2- 3.3 5%
tension piles (Germany) eq/t.10km
Fibres 424 Neidenstein Barge (river) 623 0.203 kg CO2- 5.3 8%
(Germany) eq/t.10km
Pumps 0.9 Winterswijk Truck 290 0.078 kg CO2- 0.015 <1%
eq/t.km
Total 64 100%
150
Finishing Cleaning Pressure washer 64 locks per day 0.8 - 31 25 - 0.04 <1%
locks Aerial work platform 64 locks per day - 12 31 - 372 0.06 <1%
Welding Welding aggregate 2 sections per hour 9 - 62 557 - 0.9 <1%
Aerial work platform 2 sections per hour - 12 62 - 744 0.1 <1%
Excavation Cutter suction 7000 m3/hour 594 - 9 5159 - 8.4 4%
dredger
Anchor/tension pile installation Installation rig 1 anchor per hour 49 - 844 41356 - 67.3 35%
Grout pump 1 anchor per hour 0.8 - 844 675 - 1.1 <1%
UCF pouring Concrete pump 100 m3/hour 15 - 36 662 - 0.8 1%
Dewatering Pumping Drainage pump 6000 m3/day 10 - 90 900 - 1.465 <1%
Crane vessel 24 - 90 2120 - 3.50 <1%
Cleaning High pressure 200 m2/day 0.8 - 180 144 - 0.2 1%
washer
Sediment removal 85 m2/day 24 - 53 1249 - 2 2%
machine
Placement of Placement of Telescopic crane 1500 kg/h - 400 117 - 70650 7.9 <1%
structural floor reinforcement
Pouring Concrete pump 100 m3/hour 15 - 14.9 331 - 0.3 1%
concrete
Gate placement Floating crane 1 unit per day - 132 8 - 1056 0.2 0%
Pump placement Mobile crane 1 unit per day - 101 8 - 807 0.1 0%
Excavation Hydraulic crane 1300 m3/day - 141 6 - 833 0.1 <1%
Pipe Mobile crane - 101 40 - 4040 0.7 <1%
placement
Pump Mobile crane 1 unit per day - 101 8 - 808 0.1 <1%
placement
Total 191 100%
No major reductions can be found in this phase, the quantities of elements that make up the
base of the dock chamber structure are unchanged.
The emissions during the operational phase of Variant A can be seen in Table H17.
Table H 17: Operational phase – variant A
Activity Equipment Productivity Diesel Energy Hours per Emissions per Contribution
consumption consumption year [ton CO2-
[Litres/hour] [kWh/h] year eq/year]
Gate operation Telescopic 0.25 hours per - 505 20 0.85 3%
crane docking
Dock emptying Pumps 3 hours per docking - 800 60 16.2 66%
Inlet operation Pumps/gate 1 hour per docking - 400 72 2.7 11%
Tugboats support Tugboat 2 h/docking activity 94 - 40 4.8 20%
during docking
Total 25 100%
As expected, the emissions during operational phase are significantly reduced by the
prevention of sediment cleaning and disposal activities.
The emissions during the maintenance phase of Variant A can be seen in Table H18
Table H 18: Maintenance phase – variant A
Presence of more pumps will naturally ask for more maintenance of the design, but the total
increase in emissions is relatively small.
151
Reinforced 12600000 [dm3] 0.33 kg CO2-eq / dm3 4158 41%
concrete
Steel piles 2775 [m’] 1196 kg CO2-eq / m’ 33181 32%
Sheet pile wall 174 [m’] 554 kg CO2-eq / m’ 96 1%
Steel 269622 [kg] 0.65 kg CO2-eq / kg 175 2%
Steel coating 4509 [m2] 4 kg CO2-eq / m2 18 <1%
UCF 3825 [m3] 105 kg CO2-eq / m3 401 4%
Anchors 124 [# of anchors] 2393 kg CO2-eq / anchor 297 3%
Tension piles 720 [# of piles] 2393 kg CO2-eq / pile 1723 17%
Pumps 2 [# of pumps] 724 kg CO2-eq / pump 1.5 <1%
Basalt fibre 123 [ton] 398 kg CO2-eq / ton 49 <1%
Hydrojet 4 [# of hydrojets] 724 kg CO2-eq / 2 1.5 <1%
hydrojet
Gate operating 2 [# of winches] 2111 kg CO2-eq / winch 4.2 <1%
system 2 [# of 9670 kg CO2-eq / motor 19 <1%
hydromotor]
Total 10262 100%
The total amount of emissions is significantly lower than the base design, stemming from the
fact that less reinforced concrete is needed which is the predominant factor in this lifecycle
phase. The inclusion of the basalt fibres and hydro jet construction are significantly less than
the effects from the concrete saving floor design. Compared to Variant A, the presence of
basalt fibres as opposed to steel fibre leads to a reduction in total emissions.
The emissions during the transport phase of Variant B can be seen in Table H20.
Table H 20: Transport phase – Variant B
Again, the reduction in emissions can be allocated to the reduction of gravel transport.
The emissions during the installing phase of Variant B can be seen in Table H21
Table H 21: Installation phase – variant B
Activity Equipment Productivity Diesel Energy Hours Total diesel Total Total Contribution
152
Sheet pile wall Vibratory hammer 7 meters per day 48 - 411 19925 - 32.4 17%
installation Welding Welding aggregate 2 sections per 9 - 62 557 - 0.9 <1%
hour
Finishing Cleaning Pressure washer 64 locks per day 0.8 - 31 25 - 0.04 <1%
locks Aerial work platform 64 locks per day - 12 31 - 372 0.06 <1%
Welding Welding aggregate 2 sections per 9 - 62 557 - 0.9 <1%
hour
Aerial work platform 2 sections per - 12 62 - 744 0.1 <1%
hour
Excavation Cutter suction dredger 7000 m3/hour 594 - 9 5227 - 8.5 5%
Anchor/tension pile installation Installation rig 1 anchor per 49 - 844 41356 - 67.3 35%
hour
Grout pump 1 anchor per 0.8 - 844 675 - 1.1 <1%
hour
UCF pouring Concrete pump 100 m3/hour 15 - 38.25 563 - 0.9 <1%
Dewatering Pumping Drainage pump 6000 m3/day 10 - 90 900 - 1.465 1%
Crane vessel 24 - 90 2120 - 3.5 2%
Cleaning High pressure washer 200 m2/day 0.8 - 180 144 - 0.2 <1%
Sediment removal 85 m2/day 24 - 53 1249 - 2 1%
machine
Placement of Placement of Telescopic crane 1500 kg/h - 400 99 - 39564 6.7 4%
structural floor reinforcement
3
Pouring Concrete pump 100 m /hour 15 - 12.6 185 - 0.3 <1%
concrete
Gate Hydrojet Mobile crane 4 unit per day - 101 8 - 808 0.1 <1%
placement placement
Gate Mobile crane 2 unit per day - 101 8 - 404 0.1 <1%
operation
system
Gate Floating crane 1 unit per day - 132 8 - 1056 0.2 <1%
installation
Pump placement Mobile crane 1 unit per day - 101 8 - 808 0.1 <1%
Total 189 100%
The fact that that the hydro jets seem to have an efficient energy consumption compared to
the inlet pumps in Variant A leads to a reduction in emissions compared to the other variants.
Maintenance
The emissions during the maintenance phase of Variant B can be seen in Table H23
Maintenance emissions of this Variant are significantly larger than Variant A, due to the
presence of more electrical/mechanical elements that require maintenance throughout the
lifetime of the dock.
153
Impact assessments – Variant C
Compared to the base design, this variant consists of a little less reinforced concrete and no
underwater concrete but it adds the contribution of the 6 winches and the gate operation
system. The total emissions for this phase can be found in table H24
Table H 24: Material phase – Variant C
The emissions for this phase are less than the base designs, since the required amount of
reinforced concrete is less. However, the design requires a thicker layer of reinforced concrete
due to the absence of the UCF layer and therefore the total amount of CO 2 emissions are
significantly larger for Variant C than for the other two design Variants.
The emissions during the transport phase of Variant C can be seen in Table H25
Table H 25: Transport phase – Variant C
Materials Total Place of origin Transport Distanc Emissio Unit Total Contribution
emissions
amount mode e [km] n Factor [ton CO2-eq]
[ton]
Concrete Cement 551 IJmuiden Barge (river) 120 0.203 kg CO2- 1.3 2%
eq/t.10km
Sand 1530 Harlingen Barge (river) 2 0.203 kg CO2- 0.06 <1%
eq/t.10km
Gravel 4055 Limburg Barge (river) 350 0.203 kg CO2- 28.8 52%
eq/t.10km
Steel Reinforcement 450 Brescia (Italy) Barge (river) 1200 0.203 kg CO2- 11 20%
steel eq/t.10km
Steel piles 1026 Essen Barge (river) 283 0.203 kg CO2- 5.9 11%
(Germany) eq/t.10km
Sheet piles 323 Essen Barge (river) 283 0.203 kg CO2- 1.9 3%
(Germany) eq/t.10km
Gate 270 IJmuiden to Truck 127 0.078 kg CO2-eq/t.km 2.7 5%
Heeg
270 Heeg to Barge (river) 45 0.203 kg CO2- 0.2 1%
Harlingen eq/t.10km
Anchors & 506 Langefeld Barge (river) 319 0.203 kg CO2- 3.3 6%
tension piles (Germany) eq/t.10km
Pumps 0.5 Winterswijk Truck 200 0.078 kg CO2-eq/t.km 0.007 <1%
Gate Hydraulic 2.5 Hengelo Barge (river) 180 0.203 kg CO2- 0.046 <1%
operation motor eq/t.10km
system
Winches 3.9 Hardinxveld- Barge (river) 200 0.203 kg CO2- 0.015 <1%
Giessendam eq/t.10km
Total 55 100%
The emissions during the installation phase of Variant C can be seen in Table H26
154
Table H 26: Installation phase – Variant C
Activity Equipment Productivity Diesel Energy Hours Total diesel Total Total Contr
consumption consumption required consumption Electricity emissions ibutio
[Litres/hour] [kWh/h] [litres] consumption [ton CO2-eq] n
[kWh]
Pile installation Installation Hydraulic crane 2 piles per day - 141 500 - 70500 11.8 6%
Vibratory Hammer 2 piles per day 48 - 500 24214 - 39.4 20%
Welding Welding aggregate 2 piles per day 9 - 62.5 561 - 0.9 <1%
Sheet pile wall Installation Hydraulic crane 7 meters per day - 141 411 - 58011 9.8 5%
installation Vibratory hammer 7 meters per day 48 - 411 19925 - 32.4 16%
Welding Welding aggregate 2 sections per hour 9 - 62 557 - 0.9 <1%
Finishing locks Cleaning Pressure washer 64 locks per day 0.8 - 31 25 - 0.04 <1%
Aerial work platform 64 locks per day - 12 31 - 372 0.06 <1%
Welding Welding aggregate 2 sections per hour 9 - 62 557 - 0.9 <1%
Aerial work platform 2 sections per hour - 12 62 - 744 0.1 <1%
Excavation Cutter suction 7000 m3/hour 594 - 9 5159 - 8.4 4%
dredger
Anchor/tension pile installation Installation rig 1 anchor per hour 49 - 844 41356 - 67.3 33%
Grout pump 1 anchor per hour 0.8 - 844 675 - 1.1 <1%
Placement of Placement Telescopic crane 1500 kg/h - 400 300 - 120105 20.2 10%
structural floor of
reinforceme
nt
Pouring Concrete pump 100 m3/hour 15 - 38.25 563 - 0.9 <1%
concrete
Dewatering Pumping Drainage pump 6000 m3/day 10 - 90 900 - 1.465 1%
Crane vessel 24 - 90 2120 - 3.5 2%
2
Cleaning High pressure 200 m /day 0.8 - 180 144 - 0.2 <1%
washer
Sediment removal 85 m2/day 24 - 53 1249 - 2 1%
machine
Gate placement Gate Floating crane 1 unit per day - 132 8 - 1056 0.2 <1%
installation
Pump placement Mobile crane 1 unit per day - 101 8 - 808 0.1 <1%
Winch installation Mobile crane 1 unit per day - 101 8 - 808 0.1 <1%
Total 202 100%
Activity Equipment Productivity Diesel Energy Hours per Emissions per Contribution
year [ton CO2-
consumptio consumptio year eq/year]
n n [kWh/h]
[Litres/hour
]
Gate operation Hydraulic 30 minutes per - 50 3 0.007 <1%
power unit docking
Dock emptying Pumps 3 hours per docking - 800 60 16.2 87%
Docking/undocking Winches 1 h/docking activity - 180 40 2.4 13%
vessels
Total 18.6 100%
Due to the short duration of gate operation and absence of tugboats for this variant, the
emissions during operation of the dock throughout its lifetime is relatively low.
The emissions during the maintenance phase of Variant C can be seen in Table H28.
Table H 28: Maintenance phase – Variant C
More mechanical components mean more maintenance and 6 winches also means that 6 new
winches are needed after 50 years.
155
Appendix I: Cost-Benefit analysis details
This appendix gives the details behind the cost benefit analysis, starting off with the cost
inventory per life cycle stage.
Table I 1: Material costs inventory
The prices for the elements that make up the base design, such as the combi walls, UCF,
reinforced concrete, anchors, tension piles and sand have been retrieved from a reference
project provided by ADONIN where the prices have been adjusted to 2023 by accounting for
price changes due to inflation over time.
The gate price per ton has been estimated at 2000 euros per ton which includes the price of
steel and the construction costs for the gate.
The price for steel coating of the gate and combi wall has been estimated with help of
experience from DAMEN.
LBN Betonproducten B.V. provided the costs for the concrete pipe per running meter for the
pipe with the largest available diameter in their product range.
The prices for the steel and basalt fibres originate from online research (IN2-concrete, 2023)
and the prices for the hydro jets, winches and hydraulic driving unit are estimated based on
the estimated price per pump unit and online research.
156
Pumps Winterswijk Truck 290 0.7 All
Hydrojet Winterswijk Truck 290 0.7 B
Winch Hardinxveld- Kempenaar 200 5.3 B,C
Giessendam
Hydraulic motor Hengelo Kempenaar 180 5.3 B,C
Table H3 shows an overview of the fuel consumptions for the installation phase from the fuel
costs for the installation phase is determined. For the rent of equipment, the following
estimated prices are maintained based on estimations and online research as shown in Table
I3:
Table I 3: Equipment rent prices
The calculations of the construction costs are performed for each of the design variants,
starting with the base design.
For the materials phase of the base design, the costs per element can be determined as
follows:
Table I 4: Material costs – base design
The largest contributing factor to the total construction costs are found to be the anchors and
tension piles and the floor package.
Transport costs for the base design are given in the table below
Table I 5: Transport costs – base design
Materials Place of origin Transport Distanc Amount Amount of Total Costs Total costs Contribution
mode e [km] [ton] trips [km] [€/km] [€]
Concrete Cement IJmuiden Kempenaar 120 972 2 480 5.3 2544 3%
Sand Harlingen Kempenaar 2 6300 10 40 5.3 312 <1%
Gravel Limburg Kempenaar 350 7155 11 7700 5.3 40810 51%
Steel Reinforcement Brescia (Italy) Kempenaar 1200 265 1 2400 5.3 12720 16%
steel
Steel piles Essen (Germany) Dortmunder 283 1026 2 1132 9.3 10528 13%
Sheet piles Essen (Germany) Dortmunder 283 323 1 566 9.3 5264 7%
Gate IJmuiden to Heeg Truck 127 270 10 2540 0.7 1778 2%
Heeg to Harlingen Kempenaar 45 270 1 90 5.3 477 1%
Anchors Langefeld (Germany) Dortmunder 319 506 1 638 9.3 5933 7%
Pumps Winterswijk Truck 290 0.47 1 400 0.7 280 <1%
Total €80.546 100%
Similar to the life cycle analysis, the transport of the gravel that is used to produce concrete is
the predominant factor in the total costs of the transport phase. Furthermore, the contribution
of the transport phase to the total is minimal compared to the material phase.
The costs of the installation phase consist of fuel and rent prices where the number of hours
that equipment is required depends on the productivity of that equipment. An overview of the
costs can be seen in the table below.
157
Table I 6: Installation costs – base design
The costs for performing the excavations result from the beforementioned reference project
and is divided into the removal of dry and wet soil.
The total cost of construction of the base variant has been found to be roughly €6.55 million
and after adding the 15% to account for design costs and other unforeseen factors, the costs
amount €7.55 million. Comparison between the costs of construction of the design variant can
be done after the construction costs of the other design variant have been calculated.
The reduction in costs that result from the reduced volume of concrete is negated by the
increased costs of the steel fibres. This has resulted in a price increase compared to the base
design.
Transport costs for this variant are given in table I8.
Table I 8: Transport costs – Variant A
Materials Place of origin Transport Distanc Amount Amount of Total Costs Total costs Contribution
mode e [km] [ton] trips [km] [€/km] [€]
158
Concrete Cement IJmuiden Kempenaar 120 732 2 480 5.3 2544 3%
Sand Harlingen Kempenaar 2 2034 3 12 5.3 64 <1%
Gravel Limburg Kempenaar 350 5390 8 5600 5.3 29680 36%
Pipe Burgum Kempenaar 45 137 1 90 5.3 477 1%
Steel Reinforcement Brescia (Italy) Kempenaar 1200 175 1 2400 5.3 12720 16%
steel
Steel piles Essen (Germany) Dortmunder 283 1026 2 1132 9.3 10528 13%
Sheet piles Essen (Germany) Dortmunder 283 323 1 566 9.3 5264 6%
Gate IJmuiden to Heeg Truck 127 270 10 2540 0.7 1778 2%
Heeg to Harlingen Kempenaar 45 270 1 90 5.3 477 1%
Anchors Langefeld (Germany) Dortmunder 319 424 1 638 9.3 5933 7%
Fibres Neidenstein (Germany) Dortmunder 623 506 1 1246 9.3 11588 14%
Pumps Winterswijk Truck 290 .47 1 400 0.7 280 <1%
Total €81.332 100%
The costs of the installation phase consist of fuel and rent prices where the number of hours
that equipment is required depends on the productivity of that equipment. An overview of the
costs can be seen in table I9.
Table I 9: Installation costs – Variant A
The total cost of construction of this variant, including the 15% for unforeseen costs has been
found to be roughly €7.85 million.
Now for Variant B, starting off with the material costs in table I10.
Table I 10: Material costs – Variant B
159
Basalt fibres [kg] 10 123356 1233560 20%
Hydro jet [#] 10000 4 40000 1%
Winches [#] 10000 2 20000 <1%
Hydraulic motor [#] 2000 2 4000 <1%
Total €6.189.820 100%
Materials Place of origin Transport Distanc Amount Amount of Total Costs Total costs Contribution
mode e [km] [ton] trips [km] [€/km] [€]
Concrete Cement IJmuiden Kempenaar 120 732 2 480 5.3 2544 3%
Sand Harlingen Kempenaar 2 2034 3 12 5.3 64 <1%
Gravel Limburg Kempenaar 350 5390 8 5600 5.3 29680 35%
Steel Reinforcement Brescia (Italy) Kempenaar 1200 175 1 2400 5.3 12720 15%
steel
Steel piles Essen (Germany) Dortmunder 283 1026 2 1132 9.3 10528 13%
Sheet piles Essen (Germany) Dortmunder 283 323 1 566 9.3 5264 6%
Gate IJmuiden to Heeg Truck 127 270 10 2540 0.7 1778 2%
Heeg to Harlingen Kempenaar 45 270 1 90 5.3 477 1%
Anchors Langefeld (Germany) Dortmunder 319 123 1 638 9.3 5933 7%
Basalt Fibres Neidenstein (Germany) Dortmunder 623 506 1 1246 9.3 10528 13%
Pumps Winterswijk Truck 200 .47 1 400 0.7 280 <1%
Hydro jet Winterswijk Truck 200 .47 1 400 0.7 280 <1%
Gate operating Hydraulic motor Hengelo Kempenaar 180 2.5 1 360 5.3 1908 2%
system Winches Hardinxveld- Kempenaar 200 1.3 1 400 5.3 2120 3%
Giessendam
Total €84.103 100%
The total cost of construction of this variant has been found to be roughly €7.99 million
including the factor of 15% for unforeseen costs. Next, the construction costs for Variant C is
calculated, starting with the material costs in table I13.
160
Table I 13: Material costs – Variant C
Materials Place of origin Transport Distanc Amount Amount of Total Costs Total costs Contribution
mode e [km] [ton] trips [km] [€/km] [€]
Concrete Cement IJmuiden Kempenaar 120 551 1 240 5.3 1272 2%
Sand Harlingen Kempenaar 2 1530 3 12 5.3 64 <1%
Gravel Limburg Kempenaar 350 4055 6 4200 5.3 22260 34%
Steel Reinforcement Brescia (Italy) Kempenaar 1200 450 1 2400 5.3 12720 20%
steel
Steel piles Essen (Germany) Dortmunder 283 1026 2 1132 9.3 10528 16%
Sheet piles Essen (Germany) Dortmunder 283 323 1 566 9.3 5264 8%
Gate IJmuiden to Heeg Truck 127 270 10 2540 0.7 1778 3%
Heeg to Harlingen Kempenaar 45 270 1 90 5.3 477 1%
Anchors Langefeld (Germany) Dortmunder 319 506 1 638 9.3 5933 9%
Pumps Winterswijk Truck 200 .47 1 400 0.7 280 1%
Gate operating Hydraulic motor Hengelo Kempenaar 180 2.5 1 360 5.3 1908 3%
system Winches Hardinxveld- Kempenaar 200 3.9 1 400 5.3 2120 3%
Giessendam
Total €64.603 100%
The total cost of construction including the uncertainty range of 15%, has been found to be
roughly €7.1 million.
Finally, the construction costs for Variant D are calculated starting off with the material costs
in table I16. This is a combination of Variant A and Variant B.
161
Table I 16: Material costs – Variant D
Materials Place of origin Transport Distanc Amount Amount of Total Costs Total costs Contribution
mode e [km] [ton] trips [km] [€/km] [€]
Concrete Cement IJmuiden Kempenaar 120 732 2 480 5.3 2544 3%
Sand Harlingen Kempenaar 2 2034 3 12 5.3 64 <1%
Gravel Limburg Kempenaar 350 5390 8 5600 5.3 29680 37%
Pipe Burgum Kempenaar 45 137 1 90 5.3 477 1%
Steel Reinforcement Brescia (Italy) Kempenaar 1200 175 1 2400 5.3 12720 16%
steel
Steel piles Essen (Germany) Dortmunder 283 1026 2 1132 9.3 10528 13%
Sheet piles Essen (Germany) Dortmunder 283 323 1 566 9.3 5264 7%
Gate IJmuiden to Heeg Truck 127 270 10 2540 0.7 1778 2%
Heeg to Harlingen Kempenaar 45 270 1 90 5.3 477 1%
Anchors Langefeld (Germany) Dortmunder 319 424 1 638 9.3 5933 7%
Basalt Fibres Sangerhausen Dortmunder 566 123 1 1132 9.3 10528 13%
(Germany)
Pumps Winterswijk Truck 290 .47 1 400 0.7 280 <1%
Total €80.272 100%
162
The total cost of construction of this variant, including the 15% for unforeseen costs has been
found to be roughly €7.67 million.
Discussion and comparison of the result can be seen in section 6.2, the following section here
show the calculations for the operational costs of each variant.
Firstly for the base design, the operation costs are displayed in Table I 19
Activity Equipment Productivity Diesel Energy Hours per Costs per Contribution
consumption consumption year year [€]
[Litres/hour] [kWh/h]
Dock emptying Pumps 3 hours per 800 60 19200 6%
docking
Sediment Cleaning High pressure 10000 m2/day 0.8 72 49 <1%
removal washer
Sediment removal 85 m3/h 24 53 1062 <1%
machine
Sediment Truck 270 tons - 840 <1%
transport
Sediment - 270 tons 270000 89%
disposal
Tugboats support during Tugboat fuel 2 h/docking 94 80 12762 4%
docking activity
Tugboat rent 2 h/docking - - 80 64000 21%
activity
Total €367.913 100%
For design variants A, and D the costs per year are identical and stem from energy use of the
pumps and the rent of a single tugboat to balance the ship during docking. Table I20 below
shows an overview.
Table I 20: Operational costs – Variant A&D
Activity Equipment Productivity Diesel Energy Hours per Costs per year [€] Contribution
consumption consumption
[Litres/hour] [kWh/h] year
Gate operation Telescopic 0.25 hours per - 505 10 2020 3%
crane docking
Dock emptying Pumps 3 hours per docking - 800 60 19200 31%
Inlet operation Pumps/gate 1 hour per docking - 400 20 3200 5%
Tugboat support Tugboat fuel 2 h/docking activity 94 - 80 6381 10%
during docking Tugboat rent 2 h/docking activity - - 80 32000 51%
Total €62.801 100%
The fact that no sediment needs to be disposed here leads to a large reduction in operational
costs, this can also be seen in the operational costs for Variant B in table I21.
The operational costs per year for this variant is in the same order of magnitude of variant A,
meaning a considerable saving compared to the base design.
Finally, the operational costs per year for Variant C are calculated below in table I22.
163
Table I 22: Operational costs – Variant C
Activity Frequency Costs per activity [€] Costs during lifetime [€] Contribution
Dock pump inspection Every 5 years 1556 29569 10%
Dock pump Every 25 years 31125 93374 31%
replacement
Gate & Combi-wall Every 20 years 45000 180000 59%
coating
Total €302.943 100%
The design variants will have larger maintenance costs since there are simply more
maintenance-requiring elements. For variants A and D this results in the following costs:
Table I 24: Maintenance costs – Variant A&D
Activity Frequency Costs per activity [€] Costs during lifetime Contribution
[€]
Dock pump inspection Every 5 years 1556 29569 7%
Dock pump Every 25 years 31125 93374 22%
replacement
Inlet pump inspection Every 5 years 1556 29569 7%
Inlet pump Every 25 years 31125 93374 22%
replacement
Gate & Combi-wall Every 20 years 45000 180000 42%
coating
Total €425.886 100%
And similarly for variant B, the maintenance costs are shown in table I25.
Table I 25: Maintenance costs – Variant B
Activity Frequency Costs per activity [€] Costs during lifetime [€] Contribution
Dock pump inspection Every 5 years 1556 29569 5%
Dock pump replacement Every 25 years 31125 93374 16%
Hydrojet inspection Every 5 years 2046 40926 7%
Hydrojet replacement Every 25 years 40926 122779 21%
Hydraulic motor inspection Every 5 years 312 6231 1%
Hydraulic motor replacement Every 25 years 6231 18694 3%
Gate driving system inspection Every 5 years 1122 22443 4%
Gate driving system Every 25 years 22443 67330 12%
replacement
Gate & Combi-wall coating Every 20 years 45000 180000 31%
Total €581.346 100%
Finally for Variant C, the maintenance costs are displayed in table I26.
Activity Frequency Costs per activity [€] Costs during lifetime [€] Contribution
Dock pump inspection Every 5 years 1556 29569 5%
Dock pump replacement Every 25 years 31125 93374 16%
164
Hydraulic motor inspection Every 5 years 312 5920 1%
Hydraulic motor replacement Every 25 years 6231 18693.6 3%
Winch inspection Every 5 years 3138 59627 10%
Winch replacement Every 25 years 62765 188296 33%
Gate & Combi-wall coating Every 20 years 45000 180000 31%
Total €575.479 100%
165
Appendix J: Multi Criteria Analysis details
This appendix gives all relevant background information for the multi criteria analysis, starting
off with determining the weighting factors of the criteria which is done as follows. The criteria
are compared in pairs, for each cell, a ‘0’ or ‘1’ indicates which of the two criteria is more
important. After filling out the entire matrix, the scores per criteria are added up per row and
the weighting factors can be determined based on these scores.
Ease of operation and sustainability are the most important criteria and are assigned a weight
of 35% and 30% respectively, followed by ease of maintenance which is given a weighing
factor of 20%, costs is then given a factor of 10% and finally constructability is given a factor
of 5%.
The tables below give scores to each design element separately, for explanations chapter 6.3
can be consulted.
For the UCF, the scores are given below. Since ease of operation and ease of maintenance
are not relevant for UCF design, all alternative are given a neutral score of 6.
Design alternative
Criteria Weight A B C D
Sustainability 30% 8 9 3 9
Costs 10% 5 6 8 6
Ease of operation 30% 6 6 6 6
Ease of
maintenance 25% 6 6 6 6
Constructability 5% 6 6 1 6
Scores 6.6 7.1 4.9 7.1
For the sedimentation reduction strategies, the scores are given in table J3
166
Table J 3: Sedimentation reduction strategies scores
Design alternative
Criteria Weight A B C D
Sustainability 30% 8 7 9 8
Costs 10% 6 6 8 6
Ease of operation 30% 8 7 4 8
Ease of
maintenance 25% 8 4 7 8
Constructability 5% 5 5 7 5
Scores 7.7 6.2 6.9 7.7
And finally for the gate type, the scores are given in table J4
Design alternative
Criteria Weight A B C D
Sustainability 30% 8 4 6 8
Costs 10% 7 3 8 7
Ease of operation 30% 5 7 7 5
Ease of
maintenance 25% 9 4 5 9
Constructability 5% 8 5 6 8
Scores 7.2 4.9 6.3 7.2
167
automatic operation of the gate and therefore score a little bit higher. Variant C scores a little
bit higher due to the low risks and automated processes.
For the ease of maintenance the caisson gate scores best since the gate can be completely
removed to reapply the coating and no mechanical driving unit is present that would require
maintenance. Furthermore, maintenance within the pipe can be assessed easily by closing off
the pipe with gates. This leads to the best score for variants A and D for this criterion. The
sliding gate of Variant B might be hard to access and frequent cleaning and unclogging of the
hydro jets is needed. Mitre gates are expected to require less maintenance than a sliding gate
but is harder to apply maintenance than for the caisson gate. The winches are easily accessible
since they are located on the dock walls and the winch cords can be replaced easily. All in all,
Variant B scores worst for this criteria with a 5 followed by Variant C with a score of 7. Variants
A and D clearly score better with a 9.
For the constructability criteria, the UCF with rebars scores worst since it is by far the hardest
to construct and never applied in practice. On top of that, mitre gate are commonly applied for
locks but less common for dock gate since they typically have a larger width that complicates
the design. The simplicity of the winch system compensates but still Variant C scores poorly
for the constructability criteria with a 4. For Variants A and D, the gate installation is the most
simple due to the lack of gate operating systems but the installation of the pipe is more
complicated than other sedimentation reduction strategies. The hydro jet system is also not
commonly applied and effective operation of the bubble screen could complicate the design
process. All in all, the simplicity of the caisson gate design of variants A and D result in a score
of 7 for constructability. Variant B scores a little bit worse with a 6 due to the complexity in the
hydro jet design.
168
Appendix K: Dock operations at the
current shipyard
This appendix aims to give an insight into the operation procedure at the current Damen
shipyard as well as the floating docks that were present in the past.
In the current shipyard’s ship lift docking procedure, the vessel’s mooring lines are connected
from the dock to the ship. Two tug boats then help to maneuver the vessel in the right position
to enter the basin. Winches are located at the sides of the dock and the mooring lines get
connected to the bollards on the vessel’s deck and are tightened. While the vessel might also
contain strong winches that could tighten mooring lines, these winches are often too powerful
to ensure adequate mooring and thus the dock’s winches are used to do so. In the meantime,
the docking blocks are positioned appropriately conform the vessel hull shape to ensure a
stable docking position for the vessel.
In the 1980’s, four floating docks were in use, all originally manufactured in Sweden and
purchased by the shipyard to increase the capacity of the shipyard. However, over the course
of time a structural failure lead to the collapse of one of the floating docks. At that time, the
docks were already near the end of their lifetime and so it was decided that all four docks were
to be demolished. A picture of the dock failure can be seen in figure K1 below.
169
Appendix L: Location-specific design
aspects
This appendix gives additional background information regarding the location specific design
aspects such as sedimentation of the dock and nitrogen emissions near the intended
construction location.
Located in the Wadden Sea, the port of Harlingen is known to suffer from high yearly
sedimentation rates that reduce the effective depth of the harbour basin over time and
subsequently complicate vessel navigation through the port. This means that expensive
dredging works are necessary which needs to be prevented as much as possible (Deltares,
2013). Even though the expected location of the new dock and its entry are located relatively
far downstream of the port entrance, local sedimentation can still be expected to be present.
In general, this sedimentation is caused by three main processes, which are tidal filling and
emptying, horizontal eddies and density-driven vertical flow (Winterwerp, 2005). From the
mentioned causes, Deltares has found that for the specific case of the Port of Harlingen, the
main drivers for siltation are the tidal filling and emptying of the port basin and the density-
driven circulation (Deltares, 2013). This density difference can be driven by a horizontal
gradient in salinity, temperature or sediment concentration (Vanlede, 2014). In our case, the
dominant cause of the density-driven circulation is found to be the discharges of freshwater
into the port of Harlingen. However, the effect of the tidal variation on these sedimentation
processes that exist for the complete port of Harlingen are assumed to be limited at the location
of the new dock. The location of the new dock is at a distance of approximately 3 kilometers
from the entrance of the port, as visible on Figure L1 (Google Maps, 2022).
Figure L 1: Distance between the entrance of the Port of Harlingen and Damen
Shiprepair, source: (Google Maps, 2022)
Research from Deltares shows that the effect of the sedimentation processes have already
reached a minimum at a distance of approximately 1300 meters. This can be seen in Figure
L2 below, at the peak of the tidal cycle, the sedimentation at the first harbor basin is at 0.1 g/L
and therefore this value is assumed to be considerably lower at the location of the dock
170
(Deltares, 2013). The density-driven circulation is more likely to play a factor in the
sedimentation of the port.
Various mitigation measures exist to reduce the amount of siltation, including a siltation trap or
siltation screen. A siltation trap is a local deepening of the bed where sediments prefer to settle
due to the reduced flow velocity caused by the locally increased water column that is present
(El Hamdi, 2011). Silt screen can take various forms, most of which either will form a barrier
for navigation (standing or hanging silt screen (Radermacher, 2013)), or will require a surplus
of energy for operation, as for example an air bubble silt curtain (Cheng, 2021). For the siltation
of the graving dock, the solution should possibly be searched in an efficient disposal of the
inflowing silt, possibly by allowing an outflow of water from the backside of the dock when the
vessel enters the dock that reduces the turbulence near the entrance of the dock chamber due
to the reduced return flow.
An aspect that should also be considered when discussing the sedimentation of a dock is the
opportunities for re-use of the silt upon dredging. Depending on the chemical and physical
composition of the silt, it could potentially be used for the following uses:
• Engineering purposes like land reclamation or beach nourishment.
• Agricultural uses due to the typical fertile characteristics of the silt.
• Environmental enhancement, including the restoration of nature areas such as wetland.
It has to be mentioned that the local silt might be contaminated and additional costs for
decontamination of the silt need to be included (Maher, 2013).
Another location specific design issue is the emissions of nitrogen. While nitrogen is naturally
present in the atmosphere, it can cause problems when reacting with other substances. The
present nitrogen problems are caused by the emission of nitrogen dioxide (NOx) after reacting
with oxygen during combustion processes caused by for example road vehicles (cars, trucks),
gas boilers and large industrial processes. The other harmful gas is ammonia (NH3) which is
created when nitrogen reacts with hydrogen, as is the case in the agricultural sector through
the evaporation of manure. In case large amounts of these gasses are emitted, it leads to an
enhanced growth of certain specific plant species which results in deterioration of other species
due to the competition among species in a habitat. This deterioration of diversity leads to a
more limited food supply for animal species with overall harm to the flora and fauna in the
Netherlands as a result (PH bouwadvies, 2021). Simultaneously, large concentrations in
groundwater and surface water is detrimental for the water quality and the water life (Oenema,
2019).
171
For the construction of the graving dock, the emissions during the construction phase do not
have to be considered when determining the environmental footprint of the dock for applying
for a building permit but it should still be minimized in view of the other emission gasses. The
guidelines SEB (Schoon en Emissieloos Bouwen) collect researches and initiatives, focusing
on the reduction of emissions during the construction phase, that are under development or
already in execution throughout the Netherlands. Aspects such as clean building machinery
and zero-omission building logistics are discussed and serve as a possibility for the
construction phase of the graving dock as well (SEB, 2021).
172
Appendix M: Dock components
specification
This appendix gives background information regarding the most commonly applied dock gate
types and explains the hydraulic operation of the two kinds of emptying/filling system.
Gate types
A short introduction of these most commonly applied gate types is given below:
- Mitre gates: Being one of the oldest gate types, they are most often used as lock gates
or impounding gats in ports or docks. Their mode of operation is a rotation around
hinges that are installed on the vertical axis (Royal Haskoning DHV, 2019). The main
advantage of this gate type is its low cost and short opening time but downsides include
the sensitivity for blockage by for example debris and ice (Molenaar, Locks, 2020).
Figure M1 shows an example of this gate type (Allianz). Single leaf gates are
comparable to mitre gates in terms of operation and characteristics, but are usually
more applicable for locks or docks with a small width.
- Sector gates: These rotational gates can be applied in pairs or as a single gates and
have the main advantages that they can retain water in two ways and they can be
operated in case of a water head difference across the gate which makes them very
applicable as lock gates or storm surge barriers but for docks these characteristics are
not of importance. The large amount of material that is required during the construction
of the gates makes them less favorable in comparison to other gate types (Molenaar,
Locks, 2020). A well-known example of the application of sector gates in the
Netherlands is the Maeslantkering.
- Lift gates: Using a vertical translational movement, lift gates are usually opened by
means of winches and wire ropes that pull the gates upwards. This gate type requires
little space for operation and are easy to control and repair. However, the limited air
draught that is provided for the vessels in combination with the large and heavy
superstructure that is needed for application of this gate type likely makes it less
applicable for docks. Furthermore, deflection of the bottom of the gates under full
loading conditions can cause difficulties with the roller tracks that guide the gate during
operation (Molenaar, Locks, 2020).
- A sliding or rolling gate moves horizontally into a gate chamber in the lock wall during
operation. The main advantages are the watertightness of the gate in closed condition
and its easy operation and maintenance. On the downside, there is a possibility of
jamming of the operational system of the gate during opening or closure of the gate
and the fact that a large door recess area is necessary adjacent to the gate which
increases the costs of this gate type.
173
- Caisson gates need to be completely removed in case of operation. This is usually
done by using large cranes or other machinery but floating variants exist as well, which
can be filled and pumped dry and removed by dragging them across the water surface.
The ability for maintaining a considerable reverse hydrostatic head in combination with
the lack of need for an operational system (which means no chance of malfunction
during operation) and cheap and effective design means that this type of gate is
common among shipping docks. On the other hand, additional equipment is necessary
for removal or placement of the dock which increases the amount of preparation and
time that is necessary for operation. This type of gate is present at the dock of Icon
Yachts, that was mentioned before. The fact that the dock is used for construction of
vessels means that the frequency with which the dock gate needs to be removed is
relatively low and thus this gate type is very suitable for this dock. As mentioned before,
in figure 4, this gate can be seen as well as the crane rails that are located on the
outside of the dock so that the dock cranes can be used for removal or placement of
the gates. In figure M2, the procedure of placement of a floating caisson dock can be
seen, (Ravestein Shipyard & Construction Company, 2017)
Figure 73: Procedure of caisson gate placement, source: (Ravestein Shipyard & Construction Company, 2017)
For the longitudinal inlet system, a certain energy head loss has to be taken into account for
the intake of the culvert and a friction loss for the transportation of the flow through the culvert,
additional losses due to bends in the culvert exist as well and in total, a significant amount of
energy loss will be present. The culvert can be located in or adjacent to the lock wall, in, below
or above the lock chamber floor.
174
Appendix N: Sustainable design
alternative details
This appendix gives further background information regarding the 9R framework and illustrates
how this can be applied in the design of the graving dock.
- Refuse (R0) in this case could mean to not construct a graving dock at all and try to
increase Damen’s capacity in a different way, for example by purchasing an already
existing floating dock which has the same dimensions as the intended graving dock.
Naturally the climate impact of this strategy would nearly be zero (except for the
transport of the dock to the shipyard). However, the decision of constructing a new
graving dock has already been made at an earlier stage and at a ‘higher level’ of
decision making. For the purpose of this thesis, the designer therefore has no say in
this and this strategy will therefore not be considered.
- Rethink (R1) could mean making the dock multi-functional and thereby making the
product use more intensive. The roof of the hall can be used to place solar panels that
generate energy for either the operation of the dock or for other purposes at the
shipyard, or additional workplaces and offices can be placed within the dock hall.
- Reduce (R2) means to increase the efficiency in product manufacture or use by
minimizing the amount of natural resources and materials. This strategy is closely
linked to the Life Cycle Analysis, which will be treated later on in this chapter. Increasing
the efficiency of the manufacturing process of products is more of a task for ‘higher
level’ instances and lies outside the scope of this thesis but aiming to reduce the
amount of natural resources offers a lot of possibilities for the design of the graving
dock. Designing for Disassembly is one of them and allows structures to be easily
disassembled at the end of their lifetime so that the elements which make up the
structure can easily be removed and used for other purposes. The main issue here is
ensure the structural performance of the dock. For example, the loads acting on a
graving dock are usually significant and the inclusion of a hinge in the retaining wall
which might make disassembly of the retaining wall easier could prove to be structurally
impossible. Similarly, the use of a gravity wall consisting of separate interlocking
concrete blocks might ease the deconstruction efforts for the wall but on the other hand
it would increase the amount of concrete that is necessary. The use of sustainable
building materials such as recycled steel or sustainable concrete solutions serve as
another possibility that reduces the overall climate impact of the design. A permanent,
steel fibre reinforced underwater concrete floor would reduce the need for multiple
concrete floors and consequently reduce the amount of concrete necessary in the
design. Choosing local manufacturers would reduce the emissions during the transport
of the products and CO2-neutral machinery are other possibilities to implement the R2-
strategy in the design of the graving dock.
For Reuse (R3), the element is still in good condition after disassembly and can directly be
applied in a different project to fulfil its original function. For Repair (R4), Refurbish (R5) and
Remanufacture (R6) an increasing amount of maintenance is required to bring the element
back to the appropriate quality to fulfil the same function. Repurpose (R7) uses the discarded
element or parts of the element in a new product with a different function. These strategies
could also be applied to the design of the graving dock, for example by removing structural
elements at the end of their lifetime and depending on the state of these elements, applying
them in other projects.
For Recycle (R8) this includes process the material into obtain a similar or lower quality. For
example, the concrete which will be used in the design of the graving dock can be downcycled
into construction rubble or concrete bricks and create value in a different purpose than
originally in the graving dock. Recovery (R9) includes the incineration of the building materials
175
and recovering energy as a result. These two strategies are common practice in the building
sector but they retain the lowest value of the material.
The way that the environmental impact of a design is typically determined is by using software
such as DuboCalc to perform an Life Cycle Assessment. The following paragraph summarizes
the environmental impact categories and life cycle phases that the software takes into account.
DuboCalc takes 11 different impact categories into account which describe the various
negative effects that the environment might experience during the lifetime of certain product.
These Environmental Impact Categories are expressed in equivalent units. For example, not
every greenhouse that is emitted during the manufacturing of a certain product is equally
powerful and are therefore expressed in CO2-equivalent units, meaning that the amount of
CH4 and N2O need to normalized to equivalent amounts of CO2 and therefore multiplied by a
factor of 21 and 310 respectively (Jonkers, 2018).
DuboCalc takes the following 11 categories into account and assigns a so-called ‘shadow
price’ to each of these categories to express the impact in terms of monetary value per unit
equivalent (van 't Wout, Groot, Sminia, & Haas, 2010):
Table N 1: Environmental Impact Categories (van 't Wout, Groot, Sminia, & Haas, 2010)
176
of plants and algae, limiting the
growth of other types of
organisms.
10 Abiotic Depletion Potential non-Fuel The depletion of ‘rare earth Kg Sb eq 0.16
(ADP non-fuel) elements’ such as minerals and
ores.
11 Abiotic Depletion Potential Fuel (ADP- The depletion of fossil fuels. Kg Sb eq 0.16
fuel)
DuboCalc scores the performance of the various construction elements and processes for
each of these Environmental Impact Categories during each individual stage of the element’s
life cycle so that an insight can be gained which phase contributes most to the ECI of the
element and the structure as a whole. In this way, insight can be gained in which stage of the
dock’s lifetime the largest amount of environmental impact can be saved by implementing more
sustainable alternatives. The stages that are identified in DuboCalc are the following:
- A1 – A3 Production stage:
o A1: Raw material supply
o A2: Transportation from excavation site to factory
o A3: Manufacturing of the product
- A4 – A5 Construction stage:
o A4: Transportation from factory to construction site
o A5: Construction installation process
- B1 – B5: Use stage:
o B1: Use
o B2: Maintenance
o B3: Repair
o B4: Replacement
o B5: Refurbishment
- B6 – B7 Energy use during life cycle:
o B6: Energy use
o B7: Water use
- C1 – C4: End-of-life stage C1-C4:
o C1: Demolition Process
o C2: Transport from construction site to waste plant
o C3: Waste processing
o C4: Disposal
- D: Remaining value: Re-use, recovery, recycling potential
177
Appendix O: Dock design requirements
This appendix elaborates on the specific dock design aspects that form the basis of the object
chapter in chapter 4.1.
The dimensions of the dock are determined by the design vessel that the dock aims to handle.
Eyeing the development in vessel dimensions over the past years, Damen Harlingen aims to
be able to dock vessels with a capacity of 900-1250 TEU. The design vessel that is chosen for
the design of the dock is based on the Endeavor, which is a container vessel that is built by
Volharding Shipyard and is currently owned by JR Shipping BV. Exact information on this ship
can be found in Appendix A, but this ship is characterized by a L.o.a. (Length over all) of 134.65
m, a beam of 21.5 meters and a draught (which is the distance from the bottom of the keel to
the waterline) of 7 meters (Confeeder Shipping & Chartering, 2022). For the dimensions of the
dock, extra space of approximately 5 meters is necessary on either side of the vessel when
docked in order to make repair and maintenance works possible. The blocks on which the
vessel will be placed need to be around 1,80 meters high so that personnel can access the
bottom of the ship and a freeboard of 2 meters is applied for the height of the dock. This comes
down to the following dimensions for the dock: a length of 150 meters, a width of 30 meters
and a depth of 11.8 meters. As far as the facilities within the dock are concerned, an
intermediate gate is required in order to split up the dock into multiple compartments so that
multiple vessels can be handled simultaneously. An entrance ramp has to be constructed in
order to make de dock accessible for equipment, machines and personnel. Damen Harlingen
currently mainly uses aerial lifts to perform the maintenance works and so the slope of the
entrance ramp and thus the space that is required for construction of the ramp will depend on
the maximum slope that an aerial lift can handle, which is called the gradeability of the aerial
lift. In practice, mainly a scissor lift or a telescopic boom lift are used to perform the
maintenance works and after consulting the technical specifics of various types of aerial lift, it
is found that the gradeability of the scissorlift is typically lower than that of a telescope boom
lift. The governing gradeability is found to be 25% after consulting the current aerial lift supplier
Verno. (PB Lifttechnik GmbH, 2022). For safety, a slope of 20 % should be applied in the
design. When maintaining a dock chamber depth of 11.8 meters, this comes down to a
entrance ramp length of 60 meters.
In case insufficient space is available for the construction of the entrance ramp, an alternative
with overhead rail cranes can be chosen. The minimal requirements for this crane is a minimum
of 2 cranes, each consisting of 2 trolley with a capacity of 20 t per trolley. Within the hall,
sufficient space needs to be present on either side of the dock chamber so that either a forklift
and/or a truck can access the dock for the delivery and transport of equipment and materials.
6 workspaces need to be present within the dock to facilitate the maintenance works, since
electricians, carpenters, bank workers/fitters, painters and installers all need a workspace each
and another spare workspace is required. The required dimensions of these workspaces are
a length of 10 meters, a width of 8 meters and a height of 3.5 meters. An office area, from
which the dock master can operate the dock needs to be present as well and is often build
above the workplaces in practice.
The total height of the hall depends on the required lifting height for the cranes which is
determined by the total height of the design vessel. For safety reasons, an exaggerated vessel
height is chosen, namely by taking the height of the next largest JR Shipping vessel, the MV
Elysee. Based on the dimensions of this ship, which can be seen in Appendix A, the total height
of this vessel can be estimated to be 46 meters. When maintaining a clearance between the
vessel and the crane of 1 meter and 3 meters between the crane and the top of the hall (space
for the overhead crane beam and entrance path to perform maintenance), a total hall height
between the dock chamber floor and the hall roof of 50 meters is necessary. A clearance of 40
meters above the hall floor is chosen. Thus, the hall has the following dimensions: A length of
155 m, a width of 50 meters and a height of 40 meters.
The requirements for the dock doors are mainly determined by the ability to keep the dock
completely watertight, so minimization of the leakage discharge is crucial, and the possibility
178
to perform maintenance works to the door which is required every 5 years on average. For
both the dock door and the filling/emptying system, speed of operation is a governing factor
for this type of dock since it will mainly be used to perform works on damaged vessels that
require a quick docking. The frequency of which the gate doors and filling/emptying systems
need to be operated is considerably larger than for a dock that performs construction works
and thus speed of operation plays a large role in choosing the optimal dock door.
Another requirement focuses on removing the inflowing sediment in a fast and effective way,
since cleaning the dock floor is nowadays done manually, which delays the repair and
maintenance works by a couple of days. This is unwanted since it leads to an increase in costs
and reduces the capacity of the dock and so the requirement for the new dock is that a solution
is found for removal of the inflowing sediment.
Finally, a solution must be found that enables the removal of the long propeller shaft of the
vessel, which will be necessary for some construction works.
179
Appendix P: Stakeholder analysis
This appendix explains the decisions behind the scores for the stakeholder analysis, starting
off with a definition of the terms ‘Interest’ and ‘Power’. The level of interest described how
closely each stakeholder would want to monitor the progress of the project and this is mainly
determined by the way in which the stakeholder is influenced, both in a positive and negative
way. Examples of this influence are the effects of the dock to the environment, the job
opportunities that are created and the potential boost to the local economy.
The amount of power that each stakeholder has is determined by its ability to make decisions
or alter conditions that directly influence the project. For example, governmental institutes will
have a large amount of power due to their right to hand out or withdraw permits. Local
businesses might be interested in the increased opportunities for employment but they do not
have the right to make any decisions. While the values for both of these parameters is
estimated and thus leave room for discussion, the main aim of this stakeholder analysis is gain
insights into the relative importance of each stakeholder and to summarize their role.
Damen Harlingen naturally is the main primary stakeholder in this project, as it acts as the
client in the design and construction of the new graving dock. It serves as the investing party
and has the major say in what the dock will look like, which facilities should be present and
what the functions of the dock should be. Their interest in the project can be described as the
need for an increased capacity that will be created by the construction of the dock. At the same
time, Damen wants to minimize the costs as much as possible while complying with all the
laws and regulations that are set out by other (governmental) stakeholders and it wants to
create a certain support base within the group of stakeholders that is influenced by the
presence of the new dock such as the neighboring companies within the port of Harlingen, the
small and medium sized businesses in and around the Harlingen municipality, contractors and
local residents. The interest in the project is naturally very high, this stakeholder’s power lies
in determining the lay-out, facilities and overall design of the dock.
There are several governmental institutions that will have a say in the projects, each on a
different scale, which all share the fact that they will act as a condition-creating party in the
project. They will make sure that during the full design and construction process, all relevant
laws are obeyed such as the ‘Wet milieubeheer’, which is the main environmental law on the
protection of the existing nature (Rijkswaterstaat, 2022). The ‘Waterwet’ describes the main
regulations regarding the quality of surface and ground waters, which is a very relevant topic
for a graving dock due to the fact that waste waters are often present and need to be handled
in an appropriate way. Besides monitoring whether all regulations all followed appropriately,
the Harlingen municipality, wants to make sure that the new project fits into the zoning plan
(bestemmingsplan) and the structural visions (structuurvisie) that exist. The most recent
versions of these documents describes that the activities within the port should ‘be combined
with the creation of jobs and value’ and that the acquisition of new assets should focus on
‘creating additional value to the port of Harlingen’. The municipality wants to develop the port
into a versatile, safe, sustainable and smart port and the design of the new dock should fit
these ideas (Arcadis, 2012). Furthermore, this stakeholder is interested in the boost to the local
economy that the construction of the new dock might bring through the opportunities that are
created for the small to medium sized businesses (the middle class) in and around the
municipality. On the other hand, this party has some sort of a mediating function as the
municipality would want to keep other stakeholders in the vicinity of the municipality, like the
citizens, neighboring companies and environmental parties satisfied. It is therefore also
important for this stakeholder that the nuisance to the environment that might be caused during
the construction and use of the new dock are all obeying the regulations that exist. This
stakeholder can be identified as both powerful, due to the rights to hand out and withdraw
permits, and highly interested in the proceedings of the project.
180
The second governmental party that plays a role in this project is ‘Provinsje Fryslân’, together
with ‘Wetterskip Fryslân’ which are the provincial government institution and provincial water
board respectively. Many of the interests that this party has match those of the municipal party,
in the sense that this party has interest in the boost of the local economy through increased
employment opportunities that are created and the increase of the capacity of the Port of
Harlingen. The role of this stakeholder with regards to the maintenance of environment is more
extensive than for the municipality. Making sure the all environmental laws and the
corresponding requirements with regards to emissions are followed is one of its main tasks, as
well as handing out environmental permits that will play an essential role in the construction of
the new dock due to its vicinity to the Wadden Sea. These tasks are carried out by the FUMO
(Fryske Utfieringstsjinst Miljeu en Omjouwing), which is an executive organization created by
the two aforementioned provincial institutes, that is responsible for the compliance of all
environmental laws with regards to emissions that affect the air-, soil- and water quality as well
as the acceptable noise emissions. It can be concluded that this stakeholder is powerful
through its ability to withdraw permits and has high interest in the project.
The Port of Harlingen is responsible for maintaining the order in and regulating the use of the
port, by imposing the so-called ‘Havenverordening’, which states rules that ensure the safety
within the port itself. This party is therefore interested in the effect that the new dock will have
on the use of the port, for example through the increased vessel intensity that will be created
and the effects this has on the capacity of the port. The new dock should fit well into the existing
port infrastructure and it should not intervene with the current regulations regarding the spatial
distribution of the port, as is determined in the ‘Harbour Plan’ (Port of Harlingen, 2022). Similar
to the earlier mentioned government institution, the role of the Port of Harlingen can be
described to be of high interest and high power due to its condition-setting nature.
One of the secondary stakeholder are the inhabitants of Harlingen, whose main interest in the
project is the creation of new employment opportunities by the realization of the dock, both
during construction as well as a dock crew that handles the use of the dock during its lifetime.
On the other hand, this stakeholder does not want to experience any nuisance in the form of
noise, light or odor. While the port of Harlingen is relatively remote, there are no residential
181
areas in the vicinity of the new dock, it is still important to minimize the negative impact that
the new dock might have on the inhabitants. In relative sense, both the interest in the project
as well as the power that this stakeholder has can be described to be low.
Small and medium sized businesses in the area of Harlingen will be interested in the new
employment opportunities and increase in economic development that the new dock might
bring. For instance, the fact that additional vessel and operational crews will (temporarily)
reside in Harlingen can bring an increase in revenue for these companies. The construction
works can bring additional employment and revenue to the local contractors and construction
companies. This stakeholder is considered to have a high interest in the project but a relatively
low power.
The final stakeholder is the neighboring companies in the port of Harlingen, whose main
interests is the fact that they don’t want to experience nuisance from the presence of the new
dock. The realization and operation of the new dock should not interfere with their own
activities. This stakeholder will also be interested in the potential increase in revenue that is
created. For example, a nearby crane rental company can be hired for the construction of the
dock, which would both create a boost in revenues for this company and it would reduce the
environmental impact of the construction phase by reducing the amount of transport that is
needed. This stakeholder can described to be of low power but of medium to high interest in
the project.
182
Appendix Q: Dock head design
This appendix gives the design calculations behind the design of the dock head structure. The
dock head structure will form the seat for the dock gate and has to be completely watertight.
The structure will have a length of 12 meters, where the 2 meter wide gate is placed in the
middle and so 5 meters of dock structure is placed on each side for stability purposes. Anchor
piles are placed in 5 rows with a centre to centre distance of 3 meters, each row contains
tension piles at a centre to centre distance of 2.4 meters and an edge distance of 1.8 meters
which is in accordance with the tension pile plan of the dock chamber. The floor of the dock
head is 1.5 meters thick and the walls have a thickness of 1 meter. A side view of the dock
structure can be seen in figure Q1 below, including the governing water level on the outer side
since this will be the design load situation.
The load on each tension pile row is determined by the stability of the dock head structure
under the design load conditions, where the loads are the hydrostatic pressures acting on the
gate, the upward water pressure underneath the structure, the self-weight of the gate and the
self-weight of the dock head. The loading situation showing the hydrostatic loads on the
structure can be seen in the figure underneath.
183
Figure Q 3: Hydrostatic loads on the dock head
Calculating the moments around the centre of the gate at the dock floor, the only forces causing
a bending moment are the hydrostatic pressures acting on the outer side of the dock head.
Load Value [kN/m’] Distance causing Moment [kNm/m’] Direction Load type
moment [m]
Hydrostatic, left 0.5*10*12.62=794 4.2 3334 Right/clockwise Variable
Hydrostatic left 5*10*12.6 = 630 3.5 2205 Down/counter Variable
clockwise
Upward water 12*10*12.6 = 1512 - - Up Variable
pressure
Gate s.w. 90 - - Down Permanent
Dock head s.w. 425 - - Down Permanent
ΣV 367 (Upward) ΣM 1129 (Clockwise)
The table shows the magnitude of all loads acting on the dock head, from which it can be
concluded that an effective moment on the dock head in clockwise direction is present as well
as an resulting upward load. The 5 rows of tension piles need to ensure both a rotational and
vertical equilibrium, from the resulting forces on each pile row can be computed. The nett
vertical loads is spread equally for each row resulting in a downward directed force of 367/5=73
kN/m’. For the rotational stability, each row is considered to contribute an equal bending
moment, resulting in varying reaction forces per row, which can be seen in Figure Q4 below.
184
Combining these results in the final reaction forces in Figure Q5:
After applying safety factor of 1.5 for variable loads and 1.35 for permanent load and
performing the calculations again, the governing load on the tension pile row amounts to 238
kN/m’ in downward direction meaning that the piles function as tension piles. Per pile the
governing required load therefore amounts to 238*2.4=572 kN. This means that the maintained
installation depth of -33 mNAP is sufficient and could even be reduced by a meter.
The horizontal equilibrium must be ensured by the frictional forces along the perimeter of the
combiwall and the floor. The acting load is the hydrostatic force with a magnitude of roughly
800 kN/m’ over a width of 30 meters. The resisting loads are determined using Rankine’s
model for soil pressures, which resulted in a resisting frictional force per m’ of 780 kN/m’ and
since the length of the chamber is 5 times the width, it is concluded that the resisting force is
larger than the acting force and so the horizontal equilibrium is guaranteed. Figure Q6 shows
a screen shot of the calculation.
These results also form the basis for the strength and stiffness checks of the dock head floor,
which takes the governing row of tension piles and the design load acting on it to calculate the
required longitudinal and shear reinforcement and check the cracks and deformations. The
loading scheme, M-line and V-line for both ULS and SLS conditions are shown in figure Q7
below.
185
Figure Q 7: Loading schemes, M-lines and V-lines for ULS (left) and SLS conditions
For the ULS conditions, it is found that a φ20-100 configuration of longitudinal reinforcement
is required in combination with φ12-200 configuration for the shear reinforcement. The cracking
moment is not reach for SLS conditions and the maximum deflection is found to be 20
millimetres which is within the limits. All in all, this concludes the design of the dock head
structure.
186