Content uploaded by Yushun Lian
Author content
All content in this area was uploaded by Yushun Lian on Feb 23, 2024
Content may be subject to copyright.
Engineering design and
economic analysis of
offshore seaweed farm
Yushun Lian
1,2,3
*, Samuel Obeng Boamah
1,2
, Zhenghu Pan
1,2
,
Jinhai Zheng
1,2
, Wenxing Chen
3
, Gang Ma
4
and Solomon C. Yim
5
1
Key Laboratory of Ministry of Education for Coastal Disaster and Protection, Hohai University,
Nanjing, China,
2
College of Harbour Coastal and Offshore Engineering, Hohai University,
Nanjing, China,
3
National Engineering Laboratory for Textile Fiber Materials &Processing Technology,
Zhejiang Sci-Tech University, Hangzhou, Zhejiang, China,
4
Yantai Research Institute, Harbin
Engineering University, Yantai, China,
5
School of Civil and Construction Engineering, Oregon State
University, Corvallis, OR, United States
As global demand for sustainable biomass and need to mitigate global warming
begin to rise, cultivation of seaweed has gained significant attention in recent
years due to its potential for carbon recycling. However, limited availability of
suitable coastal areas for large-scale seaweed cultivation has led to exploration
of offshore environments as a viable alternative. The nature of many offshore
environments often exposes seaweed farming systems to harsh environmental
conditions, including strong waves, currents, and wind. These factors can lead to
structural failures, kelp losses, and significant financial losses for seaweed
farmers. The main objective of this study is to present a robust design and
numerical analysis of an economically viable floating offshore kelp farm facility,
and evaluate its stability and mooring system performance. A numerical method
of preliminary designs of the offshore aquaculture systems were developed using
the OrcaFlex software. The models were subjected to a series of dynamic
environmental loading scenarios representing extreme events. These
simulations aimed to forecast the overall dynamic response of an offshore kelp
farm at a depth of 50m and to determine the best possible farm design with
structural integrity for a selected offshore environment. Furthermore, to assess
the economic feasibility of establishing offshore seaweed farms, a
comprehensive capital expenses analysis was conducted. The results revealed
that, in terms of the kelp farms with the same number of the kelp cultivating lines,
the cost of building kelp farms will be strongly affected by the cost of mooring
lines. The present study may help to understand the dynamic response and
economic feasibility of offshore kelp farms.
KEYWORDS
offshore, mooring systems, kelp cultivation, seaweed farm, dynamical response analysis,
economic analysis, capital expenses analysis of offshore seaweed farms
Frontiers in Marine Science frontiersin.org01
OPEN ACCESS
EDITED BY
Fiorenza Micheli,
Stanford University, United States
REVIEWED BY
Antoine De Ramon N’Yeurt,
University of the South Pacific, Fiji
Bob Li,
Tsinghua University, China
Zhipeng Zang,
Tianjin University, China
*CORRESPONDENCE
Yushun Lian
yushunlian@hhu.edu.cn
RECEIVED 12 August 2023
ACCEPTED 23 January 2024
PUBLISHED 21 February 2024
CITATION
Lian Y, Boamah SO, Pan Z, Zheng J, Chen W,
Ma G and Yim SC (2024) Engineering
design and economic analysis of
offshore seaweed farm.
Front. Mar. Sci. 11:1276552.
doi: 10.3389/fmars.2024.1276552
COPYRIGHT
©2024Lian,Boamah,Pan,Zheng,Chen,Ma
and Yim. This is an open-access article
distributed under the terms of the Creative
Commons Attribution License (CC BY). The
use, distribution or reproduction in other
forums is permitted, provided the original
author(s) and the copyright owner(s) are
credited and that the original publication in
this journal is cited, in accordance with
accepted academic practice. No use,
distribution or reproduction is permitted
which does not comply with these terms.
TYPE Original Research
PUBLISHED 21 February 2024
DOI 10.3389/fmars.2024.1276552
1 Introduction
Currently, the world faces numerous environmental challenges,
ranging from climate change to the depletion of natural resources.
As these issues become more pressing, the need for sustainable
solutions becomes increasingly evident. In recent years, offshore
seaweed farming has emerged as a promising alternative that not
only addresses these challenges but also presents significant
opportunities for economic growth, food security, and
environmental restoration (Buschmann et al., 2017;Campbell
et al., 2019;Abhinav et al., 2020;Ahmed et al., 2022). Seaweed,
also known as macroalgae, is a diverse group of marine autotrophs
that thrive in coastal and offshore environments. While seaweed has
traditionally been used in applications such as food, fertilizers, and
pharmaceuticals in many cultures, its potential as a sustainable
resource has garnered renewed attention worldwide (Duarte et al.,
2017;Fernand et al., 2017;Feng et al., 2021;Coleman et al., 2022;
Chen et al., 2023). One of the key advantages of offshore seaweed
farming lies in its potential to address climate change. Ocean
Macroalgal Afforestation (OMA), or the process of cultivating
macroalgae in the ocean, offers the potential to decrease
atmospheric carbon dioxide levels by increasing the natural
growth of macroalgae. These macroalgae naturally absorb carbon
dioxide, and they can be harvested for the production of
biomethane and biocarbon dioxide using anaerobic digestion
(N’Yeurt et al., 2012;Capron et al., 2020). Seaweed has a
remarkable capacity to absorb carbon dioxide (CO2) from the
atmosphere through photosynthesis, making it a powerful tool for
carbon sequestration (Duarte et al., 2017;Krause-Jensen et al.,
2018). Studies have shown that seaweed farming could potentially
offset a significant portion of global carbon emissions, thus
mitigating the impacts of climate change (Duarte et al., 2017;Zhu
et al., 2020).
By expanding offshore seaweed farming operations, humankind
can tap into this emerging market, creating new jobs and driving
economic development, particularly in coastal regions (Buck and
Buchholz, 2004). However, efficient design and operation of
offshore seaweed farms require a comprehensive understanding of
complex hydrodynamic and structural interactions between the
seaweed, the farm infrastructure, and the surrounding marine
environment. Numerical analysis using tools, e.g., finite element
procedures, has become an indispensable method for studying and
optimizing various aspects of offshore engineering systems. By
utilizing such numerical models, researchers and engineers can
simulate and evaluate sets of alternative scenarios and design
parameters, enabling them to make informed decisions and
improve the overall performance and sustainability of offshore
seaweed farms (Wang et al., 2023). Over the past few years,
various farming systems have been developed for offshore
seaweed cultivation. These include floating systems, submerged
line systems, and fixed structures such as longlines or grid
systems. Each system has advantages and challenges, including
ease of installation, maintenance, scalability, and resistance to
environmental forces. Advances in engineering and technology
have led to the development of innovative systems, such as
floating integrated systems. After examining general
representation of developments and designadvancementsof
offshore seaweed farms over the past 50 years, it is found that
only a limited number of systems managed to endure the
challenging environmental conditions encountered in offshore
and nearshore exposed sites. Specifically, the BAL Ring, and
MACR structures demonstrated technical viability, with the
MACR being tested for over eight years based on personal
observations, and the BAL undergoing a three-year testing period
according to Camus et al. (2018). A review conducted by Bak et al.
(2020) concludes that the failures experienced by early offshore kelp
farms were primarily attributed to the lack of sufficient durability in
the equipment, rendering the systems unable to withstand the harsh
environmental conditions and high capital expenses. The challenge
of ensuring survivability in offshore cultivation has led to a tendency
to over-engineer structures, resulting in excessively high installation
costs. This is exemplified by the Marine Biomass Program and the
TLP, which incurred particularly high expenses (Bak et al., 2020). In
contrast, the MACR and BAL adopted a simpler approach by
incorporating ropes, buoys, and anchors into their design. These
structures offer more flexibility, allowing them to move with the
waves and currents instead of resisting their forces. Unlike complex
systems where the cost is driven by small and delicate components,
the primary cost factors for the BAL are anchors and ropes. In the
case of the MACR system in the Faroe Islands, installation costs
were reduced by enhancing spatial efficiency and repurposing
anchors, chains, ropes, and buoys from the fishing industry and
finfish aquaculture, as observed by Bak et al. (2018). In addition, to
the previous developments discussed, many more researchers such
as Sulaiman Olanrewaju et al. (2013);Laurens et al. (2020);Ma et al.
(2022);St-Gelais et al. (2022);Lian et al. (2023a);Schmid et al.
(2023); and Olanrewaju et al. (2017) etc. in recent years have
contributed extensively to the development of modular offshore
seaweed farm. Note Lian et al. (2023b) and Seghetta and Goglio
(2020) have done a comprehensive life cycle analysis of offshore
seaweed farm to identify the impacts of a production system on
the environment.
However, according to the listed reviews above, except for
St-Gelais et al. (2022), none of the studies performed economic
analysis or capital expenses assessment of their kelp farm platform
taking into consideration the potential for profitability for low to
middle income farmers. In addition, except for Moscicki et al.
(2022), most of the studies did not consider both regular
(monochromatic) waves and random waves in the environmental
loading scenarios of their models to prevent overprediction of
expected tensions and overdesign of structure under investigation.
Also, only a few of the studies considered kelp aggregates, it’s
loading and hydrodynamic coefficients in the model. This study
aims to provide a comprehensive numerical analysis of offshore
seaweed farms, using a finite element analysis software, OrcaFlex,
focusing on key aspects such as hydrodynamics, structural
mechanics, and farm optimization. In addition, economic analysis
of the offshore seaweed farm designs will be assessed to ensure an
economically viable engineering option. The findings of this
research will contribute to the development of robust offshore
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org02
seaweed farming system that can help address the global climate
and environmental challenges we face today. Section 2 of this paper
provides a comprehensive overview of the distinctive features of the
particular offshore seaweed farm design being studied. Section 3
outlines the numerical model employed to assess the dynamic
behavior of the farm design. The process of developing and
applying loading scenarios is elucidated in Section 4.
Subsequently in Section 5, the outcomes obtained from the
numerical model are presented and analyzed. Section 6 then
elaborates on the economic assessment of the offshore seaweed
farm facility. Finally, in Section 7, the conclusions drawn from the
study are deliberated.
2 Description of offshore kelp farm
The proposed farm facility consists of a square-shaped structure
measuring 60m on each side, as shown in Figures 1 and 2. It serves
as the main platform for cultivating seaweed. To secure the farm in
place, four buoys are strategically positioned around the perimeter
of the facility, as shown in Figure 1. Each buoy is connected to the
header lines of the farm. This cultivation system employs the mussel
longline approach. The header lines and cultivation lines consist of
8-stranded nylon ropes. Smaller floaters known as buoyant
droppers are attached to the flexible rope at regular intervals. The
buoyant droppers are attached at 8m intervals on the header lines
and 10m intervals on the cultivation line. These colored buoyant
droppers or floaters provide visibility, easy accessibility and
improved buoyancy to the offshore seaweed farm. The longline
provides efficient deployment, retrieval and maintenance. Mooring
lines (100m) are connected to the buoys of the offshore seaweed
farm and anchored on the seabed at 50m depth ensuring a secure
and reliable anchoring system. These mooring lines provide the
necessary tension and support to keep the farm stable and prevent
excessive movement. To prevent excessive deformation, the anchor
lines undergo precise pretension, establishing a semi-taut mooring
system that can effectively withstand significant loads and maintain
its structural integrity. 14 cultivation lines with lengths of 60m are
connected across the square-shaped offshore seaweed farms. The
cultivation lines are arranged in an interval of 4 m. The kelp that is
attached to the cultivation lines is seaweed (Laminaria japonica),
which is a popularly cultured species. The cultivation lines serve as a
substrate for the attachment of kelp holdfasts, which can be likened
to the root system of a plant. On the other hand, the header line
plays a crucial role in consolidating and transferring the load from
the cultivation lines to the anchor lines. The buoys used in this
offshore seaweed farm are spar buoys with four cylinders connected
with a total length of 9 m which is partially submerged at 6.7m of its
length. The seaweed cultivation lines are submerged 2.3m below
sea surface.
To design reasonable kelp farms, the two different types of
offshore seaweed farm models with varying mooring arrangements
are presented and tested, as depicted in Figures 2 and 3. The
objective is to identify the design that exhibits lower tension on
the mooring lines and possesses high structural integrity.
Following a thorough evaluation, the design of Model 1 is
selected, as illustrated in Figure 2. This kelp farm comprises of 8
mooring lines that are securely anchored at a seabed depth of 50m.
These mooring lines are connected to four buoys. The header line of
Model 1 is constructed using a 60m-long 8-stranded nylon rope
with a diameter of 0.06m. The total length of the mooring line is
100m, with the first 60m consisting of nylon rope and the remaining
FIGURE 1
Plan view of offshore seaweed farm.
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org03
40m composed of stud link chain, anchored at the seabed. The
header lines and cultivation line of Model 1 and 2 are attached with
smaller buoyant droppers. The buoyant droppers are attached at 8m
intervals on the header lines and 10m intervals on the cultivation
line for Model 1 and 2. The seaweed is kept 2.3m below sea surface.
Model 2, illustrated in Figure 3, comprises of four mooring lines that
are securely anchored at a seabed depth of 50m. These mooring
lines are connected to four buoys, spread out at an angle of 45
degrees. The header line of Model 2 is constructed using a 60m-long
8-stranded nylon rope with a diameter of 0.06m. The total length of
the mooring line is 100m, with the first 60m consisting of nylon
rope and the remaining 40m composed of stud link chain, anchored
FIGURE 3
Model 2.
FIGURE 2
Model 1.
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org04
at the seabed. These offshore seaweed farm models and their distinct
mooring arrangements serve as a basis for comprehensive analysis
and comparison. By examining the performance and structural
characteristics of each model, valuable insights can be gained
regarding their suitability and effectiveness in supporting offshore
seaweed farming operations. These offshore seaweed farm models
and their distinct mooring arrangements serve as the basis for
comprehensive analysis and comparison. The mass of each buoy
used in Model 1 and 2 was carefully implemented to ensure
sufficient buoyancy and stability for the overall system of the
offshore seaweed farm under different design scenarios.
Figures 1,2show the numerical model of a modular offshore
seaweed farm structure. Table 1 provides an overview of the essential
components of the farm design, outlining the significant parameters
such as material choices and sizes that define these components. The
seaweed cultivation lines are designed at 2.3m below sea surface.
Table 2 provides the detailed structural parameters of kelp farm
components. Table 3 provides the hydrodynamic parameters. Drag
diameter equals normal drag reference area per length, or tangential
drag reference area per length.
3 Method of kelp farm model
Fiber ropes are simulated as line type with flexible linear elements.
Lines are represented by using a lumped mass method (Heffernan,
2017). That is, the line is modeled as a series of ‘lumps’of mass joined
together by massless springs. Ropes are represented using linear elastic
elements, which are assigned specific values for diameter, density,
modulus of elasticity, and length. The stud link chain is modeled by
a general line type, where the axial, bending and torsional stiffness are
set directly. Similarly, mass is given per unit length, rather than
calculated from a material density. This direct approach gives
complete control over the data, allowing the analysis of flexible risers,
umbilical, hoses, mooring chains, ropes, wires, bundles, seismic arrays,
power cables, nets etc. The strands of seaweeds are represented as linear
elastic elements attached to the line. The attachment type is defined by
its weight, diameter, and arc length at which it is attached to the line.
OrcaFlex models the buoyant dropper as a clump line attachment. The
clump attachment can be buoyant or heavy and represents a small body
that experiences forces (weight, buoyancy, drag etc.). But instead of
being free to move, it is constrained to move with the node to which it is
attached. The clump adds to the mass, buoyancy and hydrodynamic
force to the line through its node. The properties of the line attachment
themselves are given separately on the attachment types data form,
allowing the same set of attachment properties to be used for a number
of different attachments. The seaweed was attached at 2m intervals on
the cultivation line. The structural mechanics of offshore seaweed farms
are investigated in the following sections. OrcaFlex (https://
www.orcina.com/orcaflex/), a commercial FEM and multibody
physics software package, specializes in evaluating the loading and
movement of rigid floating bodies moored by flexible anchor lines.
OrcaFlex relies on finite element analysis and multibody dynamics to
simulate the hydrodynamic forces and response of marine structures
subjected to waves, currents, and wind. Marine structures are modeled
as flexible and rigid elements in the form of lines, 6- or 3-degree-of-
freedom buoys, and rigid body elements (Moscicki et al. 2022).
3.1 Static analysis
The hydrodynamic analysis of offshore seaweed farms also
forms a fundamental component of this research. Understanding
the flow characteristics around seaweed structures is essential for
predicting the drag forces acting on the plants and determining
their motion and stability. OrcaFlex uses steady hydrodynamic
forces and catenary equations, an iterative approach using the
multi-dimensional form of Newton’s method to find positions
and orientations for each element in the model such that all
forces and moments are in equilibrium.
Given a system of n equations with unknowns: F(x) = 0, where
F:R
n
≥R
n
is a vector-valued function, and x=(x
1
,x
2
,…,x
n
)
represents the vector of unknowns. The multidimensional form of
Newton’s method iteratively updates an initial guess x
0
for the
solution by using the Jacobian matrix of F, denoted as J(F), and the
current estimate x
k
as shown in Equation (1):
xk+1 =xk−J(F)(xk)−1F(xk) (1)
where J(F)(x
k
) is the Jacobian matrix evaluated at x
k
, and J(F)
(x
k
)
-1
denotes the inverse of the matrix. The Jacobian matrix J(F)is
an n×nmatrix, where each entry J
ij
(F) represents the partial
derivative of the i–th equation with respect to the j–th unknown.
It can be written as shown in Equation (2):
JFðÞ=
∂F1
∂x1
∂F1
∂x2…
∂F1
∂xn
∂F2
∂x1
∂F2
∂x2…
∂F2
∂xn
⋮⋮⋮
∂Fn
∂x1
∂Fn
∂x2
∂Fn
∂xn
2
6
6
6
6
6
6
4
3
7
7
7
7
7
7
5
(2)
Where ∂Fi=∂xjrepresents the partial derivative of the i–th
equation with respect to the j–th unknown. The iteration continues
until a desired level of accuracy is achieved or until a maximum
number of iterations is reached. This is done to provide a starting
configuration or static analysis solution for a dynamic simulation.
TABLE 1 Major parameters defining offshore seaweed farm structure.
Parameter Parameter’s value
Farm dimension 60m x 60m
Water depth 50m
Mooring line 8-stranded nylon rope with 0.06m diameter diameter
Anchor chain Stud link chain with 0.0159m diameter
Cultivation line 8 stranded nylon rope with 0.06m diameter
Header lines 8 stranded nylon rope with 0.06m diameter
Cultivation lines 14 Lines
Kelp weight 60kg per cultivating line
Buoyant droppers Node float
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org05
3.2 Dynamic analysis
The dynamic simulation uses the static analysis as its initial
configuration and time then evolves forward from there. OrcaFlex
uses numerical time-stepping algorithms to solve the fully nonlinear
Equation (3) in the time domain.
M(p,a)+C(p,v)+K(p)=F(p,v,t) (3)
where, M(p,a) is the system of inertia load, which is related to
the mass matrix, added mass matrix and position vector of the
system; C(p,v) is the system damping load, which is related to
damping matrix and velocity vector of the system; K(p) is the
system stiffness load, whichi is related to stiffness matrix and vector,
F(p,v,t) is the external load, which reflects the wind, wave and
current environmental loads of the systems; here, p,v and a are the
position, velocity, acceleration vectors repectively, and tis the
simulation time. Equation (3) is the simply description about
Cummmin’s equation.The more information about the
application of Cummin’s theory and applicaiton can be found
(Qiao et al., 2020;Li, 2021a;Li, 2021b;Huang et al., 2023;Li
et al., 2023,Li and Wang, 2023;Wang et al., 2023). Especitally, Li
and Wang (2023) give the detailed about theory and modeling of
offshore floating complex operations, these will be helpful to
understand the theory of dynamic analysis of mooring systems
for floating structures.
Equation (3) is solved using implicit integration schemes. This
time domain solution re-compute the system geometry at every
time step and so the simulation takes full account of all geometric
nonlinearities, including the spatial variation of both wave loads
and contact loads. The forces and moments acting on each free
body and node are then calculated. Forces and moments considered
include weight, buoyancy, hydrodynamic and aerodynamic drag,
tension and shear, bending and torque, seabed reaction and friction,
contact forces with other objects, and forces applied by links and
winches. The equation of motion (Newton’s law) is then formed for
each free body and each line node as shown in Equation (4).
M(p,a)=F(p,v,t)−C(p,v)−K(p) (4)
Before the main simulation stage(s) there is usually a build-up
stage, during which wave and vessel motions are smoothly ramped
up from zero to their full size. This gives a gentle start to the
simulation and helps reduce the transients that are generated by the
change from the static position to full dynamic motion. OrcaFlex
uses the extended form of Morison equation formulation to
incorporate the relative movement between structural
components and the surrounding fluid to estimate the
hydrodynamic loading on the structures during each time
interval. To account for the penetration of the water surface, the
simulation adjusts the buoyancy, drag, and added mass of the
affected components based on their submerged volume. This can
be described by Equation (5).
dF =1
2rDCdUf−Us
(Uf−Us)+rA(1 + Ca)
_
Uf−rACa
_
Us(5)
where Dis the characteristic drag diameter, Cd is the drag
coefficient, Ais the cross-sectional area, Ca is the added mass
coefficient, U
f
is the transvers directional fluid particle velocity, and
U
s
is the transverse directional structure velocity. OrcaFlex
TABLE 3 Hydrodynamic parameters of the farm components.
Component Normal
drag Coeff.
Tangential
Drag Coeff.
Drag
Diameter (m)
Added mass Coeff.
Mooring line 1.2 0.008 0.059 1
Cultivation line 1.2 0.008 0.029 1
Header line 1.2 0.008 0.059 1
Buoy 1.3 0 1.4 1
Kelp 1.2 1.1 0.6 1
Stud link chain 2.6 1.4 0.0159 1
Buoyant dropper 1.1 0.6 0.005 1
TABLE 2 Structural parameters of farm components.
Component Mass
(Kg)
Total
Length (m)
Diameter (m) Buoyancy
(kN)
Allowable Tension (kN)
Cultivation Line 97.0 60 0.0290 0.40 348.394
8-strand Nylon rope 190.2 60 0.0600 1.68 682.853
Kelp 60.0 2 0.0038 0.35 N/A
Stud Link Chain 2190 40 0.0159 0.28 2740
Buoy 960 9.0 4cylinders 332.56 N/A
Buoyant dropper 0.5 1 0.5 0.01759 N/A
N/A, Not applicable.
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org06
simulations enable the dynamic adjustment of added mass and drag
coefficients according to changing parameters like the Reynolds
number in each time step. The simulation specifies the current and
wave conditions at the beginning and does not modify them based
on their interaction with the structure.
3.3 Fluid model hydrodynamics theory
The hydrodynamic loads on lines, 6D buoys are calculated by
using an expanded version of Morison’s equation. Morison et al.
(1950) initially proposed the equation to calculate wave loads on
fixed vertical cylinders, incorporating two force components: one
linked to water particle acceleration, representing the fluid inertia
force, and another associated with water particle velocity,
representing the drag force (Heffernan, 2017).
The original Morison's equation can be written as Equation (6):
f=CmDaf+1
2rCdA∣vf∣vf(6)
Where, fis the fluid force per unit length on the body. C
m
is the
inertia coefficient for the body. Dis the mass of fluid displaced by the
body. a
f
is the fluid acceleration. ris the density of water. v
f
is
the fluid velocity. ris the density of water. C
d
is the drag coefficient
for the body. A is the drag area. v
f
is the fluid velocity.
This principle is extended to a moving body, where the inertia
term is reduced by CaDab, and the drag term utilizes the body-
relative velocity, resulting in the extended Morison’s equation, as
shown in Equation (7):
f=(CmDaf−CaDab)+1
2rCdA∣vr∣vr(7)
Here: is the added mass coefficient for the body. a
b
is the body
acceleration relative to earth. v
r
is the fluid velocity relative to the
body. Typically, C
m
is assumed to be 1+C
a
. Simplifying the extended
Morison’s equation can be written as (Heffernan, 2017), as shown in
Equation (8):
f=(Daf+CaDar)+1
2rCdA∣vr∣vr (8)
Here, a
r
=a
f
–a
b
represents the fluid acceleration relative to the
body. The term (Daf+CaDar) denotes the inertia force, while the
other term represents the drag force. The inertia force is comprised of
two parts: one proportional to fluid acceleration relative to earth a
f
(the Froude-Krylov component) and another proportional to fluid
acceleration relative to the body a
r
(the added mass component).
4 Environmental loads
The selected area for the model test and deployment is at Storm
Bay, Tasmania, Australia, Latitude: - 43°08’24.00”S Longitude: 147°
31’48.00”E, as shown in Figure 4.
A complete dynamic response analysis of the offshore seaweed
farm design requires consideration of a wide range of wave
conditions, including operational and extreme conditions.
Therefore, in this study, the analyses are reported for
environmental conditions with regular wave and irregular/
random (JONSWAP) waves described by significant wave height
(H
s
), the zero-crossing period (T
z
) and a constant sea current and
wind speed. According to aquaculture industry standards,
Norwegian standard of NS9415 (Norway Stardard 9415, 2021)
typically suggests the application of waves or currents with a 50-
year return period in design load cases for aquaculture structures.
Extreme conditions for the study area were determined through
statistical analysis of publicly available oceanographic data collected
wave dataset ERA5 for the last 6 years (from 2013 to 2018)
downloaded from the European Centre for Medium-Range
Weather Forecasts (ECMWF) (Chu and Wang, 2020). ECMWF
provides significant wave height H
s
= 6.84 m and the zero-crossing
period T
z
=9.4s and current speed of 0.81 m/s. From the wave raw
data, 50-year (EC3) return period sea states can be statistically
estimated using the probability distribution function (PDF) of
Gumbel distribution (Chu and Wang, 2020). Hence, in the
present simulation, for the regular wave and irregular
(JONSWAP) wave cases, both wave heights are 6.84 meter, the
wave direction is set as 45 deg. For the JONSWAP wave case, the
peak period is set as 9.4 second. For the regular wave case,
the period is 9.4 second. In addition, the current speed is 0.81m/s
and the current direction is 270 deg in these two simulation cases.
FIGURE 4
GPS coordinate for the selected site (s://latitude.to/articles-by-country/au/australia/117901/storm-bay).
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org07
5 Numerical results and discussions
Two numerical models of offshore seaweed farm were
developed. Each of the models was subjected to both regular and
irregular waves as described in Section 4. The simulation time was
set to 1200s at times steps of 0.01s with a maximum iteration of 500.
5.1 Numerical results
5.1.1 Buoy heave motion and elevation
The maximum buoy elevation values for the two models
subjected to environmental loads are shown in Figures 5–8.
Model 1 and Model 2 are subjected to random wave loading and
regular monochromatic wave. The heave movement of the buoy for
both models is compared.
Two different wave types are used in case 1 for regular waves
and case 2 for irregular waves, as shown in Section 4. The initial
stack base (z) position of buoy which serves as initial reference point
of the buoy posiition is -2.3m submerged. Comparing the buoy
elevation of Model 1 subjected to random wave with the buoy
elevation of Model 1 subjected to regular wave, as shown in
Figures 5 and 6, it can be observed that the buoy heave
movement and elevation for the random wave of Case 2 is higher
than that in the regular wave of Case 1.
In addition, comparing the maximum buoy elevation of Model
2 subjected to random wave with the buoy elevation of Model 2
subjected to regular wave, as shown in Figures 7 and 8, it can be
observed that, the heave movement of model subjected to random
wave is higher than the one of Model 2 subjected to regular wave.
Comparing Model 1 and Model 2, it can be observed that the buoy
heave movement is higher in Model 2 almost piercing the surface of
the sea as compared to the lower heave movement in Model 1. Note
that the irregular nature of the buoy heave movement of the models
when subjected to random wave and the steady pattern of buoy
heave motion when the models are subjected to regular. In
summary, the irregular and unpredictable nature of random
waves makes it more challenging for the buoyancy system of an
offshore seaweed farm to adapt to the changing wave conditions.
Consequently, the buoy heave movement tends to be higher in
random wave spectra compared to regular wave. Since the mooring
line attached to the buoy has paid out length which allowed the line
to extend and retract during up and down movement. The
sinusoidal motion of the buoy is determined to be caused by the
time-varying wave elevation. This motion occurs because
the mooring line connected to the buoy has a specific length that
allows it to extend and retract as the waves move up and down.
5.1.2 Effective tension of mooring lines
Figure 9 shows the maximum tension in the entire length of the
mooring line when Model 1 is subjected to random wave,
respectively, while Figures 10 shows the maximum tension in the
entire length of the mooring line when Model 2 is subjected to
random waves, respectively. The mooring line is anchored at the
seabed and represented at the position of 100m length of the
mooring line, whilst the fairlead is attached to the buoy and
represented at the position of 0m on the mooring line as shown
in Figures 9 and 10. The tensile forces for the two models are
analyzed and compared to check whether they are over the breaking
load limit or not. From the parameters shown in Table 2, the
breaking load for the mooring lines is given as 682.853 kN. It is
observed that none of these mooring lines has exceeded their
breaking load limit. However, it can be observed in Figures 9
through 10 that, in environmental conditions (random waves)
Model 2 has the higher mooring tension while Model 1 has the
lower mooring tension at both anchor and fairlead. Lower mooring
tension in Model 1 implies reduced cyclic loading on the mooring
system, which can result in decreased fatigue damage. Fatigue
damage accumulation can leadtofailuresinthemooring
components over time. Therefore, the lower tension levels of
Model 1 indicate potentially longer service life and reduced
maintenance requirements. A more stable mooring system in
Model 1 allows for better control and positioning of the seaweed
farms. This stability helps to maintain the desired orientation and
FIGURE 5
The heave motion of buoys in Model 1 for random wave.
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org08
minimizes the risk of dislodgement or drifting. As a result, Model 1
can potentially achieve higher productivity and operational
efficiency compared to Model 2. It can also be observed that the
mooring tension in all models when subjected to random waves are
higher than the tension when subjected to regular waves. This can
be attributed to the nonlinear effects in the response of the mooring
system and the wide range of wave heights, periods, and directions,
which leads to greater variability compared to regular waves.
From Table 4, it can be observed that maximum tension at
fairlead of mooring lines in Model 2, when subjected to random
waves experience higher tensions as compared to mooring line
tension when subjected to regular waves. As shown in Section 4, the
environmental loading conditions including wave height, period,
direction, current speed, and current direction for both random
wave and regular waves were the same and comparable. The
mooring peak tensions and system response at fairlead and
anchor for both random and regular wave were compared.
According to Table 4, it can be noticed that the line tension at
fairlead for random wave experiences an increasing percentage of
18.40% in mooring Line 3 and lowest in Line 2 at 7.4%. This can be
attributed to the differences in wave characteristics, wave energy
distribution, resonance and interference, between random waves
and regular waves. Random waves exhibit a wide spectrum of wave
heights, periods, and directions. They are more realistic
representations of natural wave conditions. The broad spectrum
of random waves can lead to resonance and interference effects,
causing varying tensions in the mooring lines as different wave
components interact with the system. Whereas regular waves have
constant wave characteristics, including wave height, period, and
direction. They are simplified representations for analytical
purposes. Regular waves are less likely to cause resonance or
interference effects due to their single-frequency nature. The
FIGURE 7
The heave motion of buoys in Model 2 for random wave.
FIGURE 6
The heave motion of buoys in Model 1 for regular wave.
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org09
nonlinear effects in the response of the mooring system and wide
rangeofwaveheights,periods,and directions, led to greater
variability of mooring tension in random waves compared to
regular waves.
5.1.3 Tension of planting lines for two models
Figure 11 shows the tensile forces exerted on planting lines for
the two preliminary models under certain period of time for
random waves, respectively. It is observed that the tension forces
on planting lines in the two models subjected to random wave did
not exceed the planting line breaking load limit which is 348.394
kN. The tensile forces in planting lines as shown in Figure 11
subjected to randon waves reveals that, most of cultivation line
tension for Model 1 is lower than the tension force of cultivation
lines in Model 2. Similar phenomenon that the tensile forces in
cultivation lines for Model 1 are lower than that of Model 2, when
the systems are subjected to regular wave loads. In summary, Model
1 has the least tension in planting lines, while Model 2 has the
highest tension. Thus, the planting lines in Model 1 are much more
suitable for seaweed farming than those in the Model 2.
5.1.4 Maximum tension in header lines
Figures 12,13 show the maximum tension in headlines along
the entire length. According to Figures 1,12,13, it can be observed
that tension in header line 1 and header line 3 at rigid ends for both
models are higher than tension in header line 2 and header line 4.
This can be attributed to load distribution. The cultivation lines
carrying the seaweed impose additional loads on the connected
header lines. Since Header Line 1 and Header Line 3 are connected
to the cultivation lines, they are directly subjected to the additional
load induced by the weight and drag forces of the seaweed. This
additional load leads to higher tension in Header Lines 1 and 3
compared to Header Lines 2 and 4, which do not carry any extra
load. However, none of this header lines for either model 1 or 2
FIGURE 8
The heave motion of buoys in Model 2 for regular wave.
FIGURE 9
Mooring line tension of Model 1 for random wave.
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org10
exceeds its breaking load limit, which could potentially cause the
collapse of the offshore seaweed farm. In addition, the tension
values of Header Lines 1 and 3 in Model 1 is lower than the values of
Header Lines 1 and 3 in Model 2 by comparing Figure 12 with
Figure 13. In addition, the tension response of the header lines for
Model 1 and Model 2 under the regular wave are lower compared to
that under the random waves. Based on the above results of Section
5.1, Model 1 is a more suitable design for keeping the structural
integrity of kelp farms because the tension values of mooring lines
and cultivation lines in Model 1 are lower than those in Model 2.
5.2 Model validation
Model validation is a crucial aspect of any simulation study,
ensuring the accuracy and reliability of the computational model.
This approach provides a means to assess the fidelity of the
simulation by examining how well it reproduces the outcomes
reported in a reputable source. Through this comparative
analysis, it is possible to establish the credibility and robustness of
capturing the dynamic behavior of the offshore seaweed farm in the
present study. For validation, the published paper titled Numerical
Modelling of a Mussel Line System by Means of Lumped-Mass
Approach published by Pribadi et al. (2019) will be simulated using
OrcaFlex. The simulated results of the dynamic response of the
Mussel Line System will be compared to the published results.
5.2.1 Description of mussel longline system
The longline system configurations (depicted in Figure 14)
employ a semi-submerged system comprising an extended central
line connecting two spar-type buoys, which serves as the backbone
supporting mussel collector lines. The dry mass, the outer diameter,
length, and volume, are set as 2500kg, 0.790meter, 8.865meter,
4.345m
3
. Key characteristics of the mooring arrangement for these
experimental setups are presented in Tables 5 (Pribadi et al., 2019).
This system, featuring partially submerged crop lines, is
anticipated to undergo reduced mooring line loads due to its
deeper submersion. The chain is discretized into 12 contiguous
segments, each resembling a homogeneous cylinder. Drag
coefficients are suggested based on the chain’s equivalent diameter
utilized in the numerical computations. The seabed is represented
as a flat bottom or constant bathymetry with a depth of 30 meters,
the buoy is set at the depth of 4.43 meter. The mooring radius,
which is the horizon distance between the nearby buoy and anchor,
set as 30 meters (Pribadi et al., 2019).
5.2.2 Environmental load of mussel
longline system
In terms of the environmental loads, the setup experiences a
regular wave with an amplitude of 5 meters and a period of 8.33
seconds (corresponding to a wavelength of 103 meters), as detailed
in the work by Pribadi et al. (2019). The wave propagates along the
positive x-axis.
5.2.3 Model validation results and discussion
The time series of the buoy’s heave motion position calculated
in MoorDyn by the publisher (Pribadi et al., 2019) and the time
series of buoys position in OrcaFlex model are compared and
analyzed. The comparison between minimum and maximum
values of buoy’s heave (z) motion results are listed in Table 6.
Good agreement can be found for heave motion between the
published results of Mordyn, according to the reference (Pribadi
et al., 2019) and the simulated model results in Orcaflex. The
FIGURE 10
Mooring line tension of Model 2 for random wave.
TABLE 4 Comparing maximum tension at fairlead of mooring lines in
Model 2 under environmental load cases.
Mooring
lines
Regular
wave
Random
wave
Percentage
increase
Line 1 60.41 kN 69.38 kN 14.90%
Line 2 59.90 kN 64.35 kN 7.40%
Line 3 65.40 kN 77.40 kN 18.40%
Line 4 56.90 kN 65.24kN 14.65%
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org11
substantial agreement witnessed between the simulated buoy’sz
positions over the entire 150-second simulation duration signifies a
robust validation outcome, providing confidence in the accuracy of
the model. This positive alignment suggests that the simulation
faithfully replicates the expected behavior of the buoy under the
specified conditions. Several contributing factors bolster the success
of this validation.
The model accurately captures and represents the environmental
conditions, such as regular waves or other external forces that
influence the buoy’s dynamics during the simulation. The
simulation incorporates a precise definition of the buoy system’s
dynamics, including the mooring configuration and the buoy’s
response to loading conditions. This ensures that the model
employs realistic and reliable values, contributing to the accuracy of
the simulation.
All conditions during the validation, including initial conditions
and applied forces, were maintained consistently. Consistency in
simulation conditions enhances the reliability of the validation
process. In summary, the observed agreement in the buoy’s
positions underscores the effectiveness of the model validation,
affirming we have the capability to replicate the behavior of the
buoy under the specified conditions. This positive outcome
FIGURE 11
Tension of planting lines for random wave.
FIGURE 12
Model 1 tension in header lines for random wave.
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org12
enhances the overall confidence in the accuracy and reliability of the
present simulation model.
6 Economic analysis
The decision-making process for selecting the most suitable
offshore seaweed farm model involves not only considering the
mooring analysis and structural integrity but also conducting a
comprehensive analysis of capital expenses. The analysis of capital
expenses plays a crucial role in evaluating the financial feasibility
and cost-effectiveness of offshore seaweed farm models. This
analysis focuses on the initial investment required to construct
the farm facilities, including mooring systems, buoys, header lines,
cultivation lines, labor expenses, and installation costs to provide a
comprehensive overview of the capital expenses associated with
each model. A basic capital expenses to gross income and initial rate
of return assessment was conducted. The assessment was limited to
capital expenses required for purchasing farm components for
Model 1, Model 2 and longline cultivation system on the same
sea surface area (60m×60m), as shown in Figure 15.Farm
equipment, seed costs, and farm-gate crop values were estimated
from current market values. By examining the capital expenses for
Model 1, Model 2 and longline cultivation system, we can gain
valuable insights into the cost implications of each design to
determine the most cost-effective option or more economically
viable option for commercial-scale offshore seaweed cultivation.
An integrated approach, considering both technical and economic
aspects, enables stakeholders to make informed decisions and
support the sustainable development of offshore seaweed farming.
6.1 Capital expenses and operational cost
A single grid of Model 1 comprises of eight mooring lines and
four main buoys, providing a robust anchoring system for the
offshore seaweed farm. The mooring line is constructed using a
100m-long 8-stranded nylon rope with a diameter of 0.06m for the
first 60m, and a stud link chain for the remaining 40m, anchored at
the seabed. The cultivation line is constructed using a 60m-long 8-
stranded nylon rope with diameter 0.03m. The construction costs of
Model 1 mainly consist of materials, labor, and installation expenses
FIGURE 13
Model 2 tension in header lines for random wave.
FIGURE 14
Numerical model of the longline cultivating system.
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org13
for a size of 60m×60m. This farm size will incorporate 8 mooring
lines, 8 anchors, 4 buoys, 4 header lines and other components, as
shown in Figure 15A. Additionally, costs associated with
underwater construction and installation might be higher due to
the complexity of the mooring arrangement. Due to 8 mooring lines
and anchors in Model 1, Model 1 is expected to have a higher initial
construction cost compared to Model 2 with 4 moorings. However,
a detailed cost estimation is necessary to determine the exact
difference. Model 2 comprises of 4 mooring lines, arranged at an
angle of 45 degrees. Similar Model 1, it also requires a 100m-long
nylon rope with a diameter of 0.06m for the first 60m, and a stud
link chain for the remaining 40m, anchored at the water area with
50 meters. Model 2 with the size of 60m×60m includes 4 mooring
lines, anchors and buoys, as shown in Figure 15B. The cultivation
lines are constructed using a 60m-long nylon rope with a diameter
of 0.03m. The construction costs of Model 2 will be lower than
Model 1 due to the reduced number of mooring lines. Additionally,
the simpler mooring arrangement might result in easier installation
procedures, further contributing to cost savings.
A single traditional long-line cultivation system comprises two
mooring lines and two buoys, according to Pribadi et al. (2019). The
mooring lines in the longline system require a 100m-long nylon
rope with a diameter of 0.06m for the first 60m and a stud link chain
for the remaining 40m, anchored at the water area with 50 meters.
The longline has only one cultivation line and is constructed using a
60m-long nylon rope with diameter of 0.06m. The construction
costs of the longline cultivation on a farm with the size of 60m×60m
will incorporate 14 cultivation lines, 28 buoys and 28 anchors, as
shown in Figure 15C. The longline system will be more expensive in
construction cost than models 1 and 2 due to the increase of
mooring lines, buoys, and anchors.
6.1.1 Material costs
The material costs include the expenses for ropes, chains, buoys,
cultivation lines, and other necessary components. Model 1, Model
2 and the traditional longline utilize 8-stranded nylon ropes and
stud-link chains for their mooring lines, and all use nylon ropes for
their header lines and cultivation lines. The costs of these materials
depend on market prices and the required quantities. In the
description of the models, it was mentioned that the mass of each
buoy used in Model 1 and Model 2 was carefully implemented to
ensure sufficient buoyancy and stability. The cost of designing and
manufacturing buoys with specific mass and stability characteristics
will contribute to the overall capital expenses. According to
Figure 15, the main difference of these three models is the
number of mooring lines. Hence, the material cost of mooring
components of kelp farms is listed in Table 7.
6.1.2 Labor expenses
The labor expenses encompass the costs associated with the
installation of mooring lines, buoys, header lines, and cultivation
lines. Model 1, with its more extensive mooring system, might
require additional labor hours for installation, compared to Model 2
and the longline cultivation system. The labor costs depend on local
labor rates and the expertise required for underwater construction.
The man hours for the installation are estimated, as listed in Table 8
(St-Gelais et al., 2022).
Following the Fair Work Commission (FWC) Annual Wage
Review 2022-2023, the Australian national minimum wage has now
increased to $23.23 per hour (The 2023 Australian Minimum Wage
Increase, 2023). We estimate an hourly labor rate in the Australian
marine industry to be about $45. In that regard, the labor cost per
person of construction and deployment of Model 1 costs $7200,
Model 2 costs $3600, whereas the 14 longline cultivation systems
costs $25200.
6.1.3 Equipment costs
Specialized equipment, such as winches, cranes, and underwater
construction tools, may be necessary for the installation of the
offshore seaweed farm. The cost of renting or purchasing this
equipment adds to the initial capital expenses. It is assumed that
the cost of renting this equipment for these three models can be
regarded as the same, because these types of equipment perform the
same functions.
6.1.4 Installation costs
The installation costs are influenced by the complexity of the
mooring system and the underwater construction project. Due to
the complex mooring systems of Model 1,the installation costs of
Model 1 is higher than those of Model 2. The transporting materials
and equipment to the offshore location may also impact installation
expenses which can also be estimated as similarly comparable. As
required, a more spacious van is required to transport equipment to
the loading boat/vessel for Model 1 as compared to Model 2. The
traditional longline cultivation system will incur the highest
installation cost, because the construction costs of the longline
cultivation will incorporate 14 cultivation lines, 28 buoys
and anchors.
TABLE 5 Line componnet properties of the longline system.
Line Type Dry Mass per
Length [kg/m]
Nominal
Diameter
[m]
Line
Length
[m]
Chain (Grade
3 steel)
10.910 0.022 108
Backbone
(Movline Plus
8 strands)
2.1 0.068 57
Mussel sock
(fully
grown mussels)
21.8 0.15 145
TABLE 6 Comparison between the buoy heave motion of (Pribadi et al.,
2019) and the present simulated model.
Model Maximum
heave motion
Minimum
Heave Motion
Mean
Heave
Motion
Model (Pribadi
et al., 2019)
0.8m above
sea surface
7.7m below
sea surface
4.24m
Simulated
Model
0.8m above
sea surface
7.4m below
sea surface
4.10m
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org14
6.2 Operational and maintenance costs
All three model systems have 14 cultivation lines with the length
of 60m, supporting the growth of Laminaria japonica. These
operational costs typically include seeding, harvesting, monitoring,
and labor for cultivation management. Hence, the operation cost of
these three cultivating systems may be almost the same.
In terms of the maintenance costs of the mooring systems, Model 1
might incur higher maintenance costs compared to Model 2, which has
a simpler mooring arrangement. The longline cultivation system will
incur the highest maintenance cost due to its increased number of
buoys and mooring lines. The traditional longline cultivation system
will incur the highest maintenance cost as compared to Models 1 and 2,
while Model 2 incurs less maintenance cost than others.
B
C
A
FIGURE 15
Kelp farms with three models with the same size. (A) Model 1 of the farm size(60m×60m). (B) Model 2 of the farm size(60m×60m). (C) Longline
systems of the farm size(60m×60m).
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org15
6.3 Cost comparison and discussions
To assess the economic feasibility of establishing offshore
seaweed farms, a cost comparison between Model 1, Model 2,
and the longline cultivation system is essential.
It is crucial to consider both direct costs (e.g., material and
labor) and indirect costs (e.g., equipment and transportation) for a
comprehensive analysis. Assuming a 5-member team is considered
for the operation and deployment of the offshore seaweed farm, the
labor cost for Model 1 is $36000, while the labor cost for Model 2 is
$18000, and the longline cultivation system will be $12600
according to Table 8. According to Table 7,thecostof
constructing seaweed farm for Model 1 is approximately
$27857.6, in addition to the labor cost of $36000, the total cost of
Model 1 equals to $63857.6. The cost of constructing Model 2 is
approximately $16494.4, in addition to the labor cost, the total cost
of Model 2 equals to $34494.4. Note that, the cost of constructing
the longline cultivation system will be $91873.6. If the labor cost is
also included, the total cost of longline systems is $217873.6.
Comparing the total cost of Models 1 and 2 with longline
systems, we can find that the number of anchors and mooring
lines leads to an increase in the total cost of building kelp farms.
Hence, developing the optimal mooring systems will help to reduce
the costs of building kelp farms.
Based on the situation of the same (14 lines) lines of all three
kelp cultivating systems, we can assume that the potential revenue
generated from seaweed products is the same. These three models
are made up of 14 cultivating lines of 60 meters length. According to
the research of (St-Gelais et al., 2022), the peak biomass of kelp in
the cultivating lines was about 12.67kg/m (±0.4kg). Hence, the kelp
yield of these kelp cultivation systems produced 10642.8kg
(±336kg) wet weight total over 840(60m×14) meters length of
cultivation lines. According to (https://www.selinawamucii.com/),
the estimation wholesale price of seaweed is $10.5/wet kg in
Australia. Considering the 7-month growth season of seaweed, we
can estimate annual revenue of $111749.4 ($3528) for these three
model cultivation systems with (60m×14) lines.
The (Return on Investment) ROI is a critical financial metric
used to evaluate the profitability of an investment. It is calculated as
the ratio of net profit to the total investment made over a specific
period. A positive ROI indicates a profitable investment, while a
negative ROI suggests a loss. If we assumed an 8% return on
investment over 3 years for the type of longline cultivating system
(St-Gelais et al., 2022), then the total investment of long cultivating
systems is about $1396867.5. Hence, the common expense of the
kelp seed, fuel, and infrastructure (vessel, truck, trailer, etc.) is about
$1178993.9. If we assumed that the common expense of the seed,
fuel, and infrastructure (vessel, truck, trailer, etc.) for Models 1 and
2 is equal to $1178993.9. The total investment of Model 1, which
includes the investment of mooring part ($63857.6) and the
common expense ($1178993.9), equals to $1242851.5. Similarly,
the total investment of Model 2 is $1215168.3 ($1178993.9 +
36174.4). We can calculate that the ROI of Model 1 is 8.99%
($111749.4/$1242851.5), while the ROI of Model 2 is 9.20%
($111749.4/$1215168.3).
Note that ROI Factors, which include kelp yield, seaweed
quality, expense of infrastructure, kelp farm size and pattern of
mooring etc., can impact the ROI. Especially, we can see that the
reducing costs of mooring systems for kelp farms can strongly affect
the ROI of building kelp farms. While Model 1 appears to have
better mooring analysis results and higher structural integrity, the
economic analysis reveals that its capital expense is higher. Here,
Model 2 presents a more economically viable option for seaweed
farmers, while model 1 is the most expensive to build. However, to
design the most economically viable and optimal of offshore kelp
farms, further detailed lifecycle analysis of kelp farms should be
performed in the future.
TABLE 7 List of main components of Kelp Farms (Source from St-Gelais et al., 2022).
Main components Price Quantity
Of M1
Total cost
of M1
Quantity
Of M2
Total cost
of M2
Quantity
Of LS
Total cost
of LS
Anchors $520 8 $4160 4 $2080 28 $14560
Chains with
0.0159m diameter
$51/meter 40×8 $16320 40×4 $8160 40×28 $57120
Nylon rope with
0.06m diameter
$4.68/
meter
60×8 $2246.4 60×4 $1123.2 60×28 $7862.4
Buoy with 332.56
kN Buoyancy
$300 4 $1200 4 $1200 28 $8400
Cultivation lines of
Nylon rope
$4.68/
meter
60×14 $3931.2 60×14 $3931.2 60×14 $3931.2
Total of mooring parts $27857.6 $16494.4 $91873.6
Here, Note that M1 means Model 1, M2 means Model 2; LS means Longline systems.
TABLE 8 Man-hours required to construct offshore seaweed farm
(St-Gelais et al., 2022).
Parameter Model
1
Model
2
Longline cultiva-
tion system
Preparation 8h×4 8h×2 8h×14
Deployment 26h×4 26h×2 26h×14
Post
Deployment
6h×4 6h×2 6h×14
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org16
7 Conclusion
This paper presented a numerical analysis and simulation of an
offshore seaweed aquaculture facility. This study also involved an
economic analysis of the proposed offshore seaweed farm facility to
enable stakeholders to make informed decisions and to support the
sustainable development of the offshore seaweed farm facility. The
study presented model simulations of offshore seaweed farm
facilities, subjecting them to extreme environmental loads such as
waves and currents. Among the seaweed platform models, Models 1
and 2 demonstrated the structural feasibility design, featuring a
stable mooring arrangement with minimal tension exerted on each
line within the model. The analysis encompassed estimating the
tension on the mooring lines, planting lines, and header lines,
comparing them to the line-breaking limit values. However,
economic analysis were performed, the result showed that Model
2 is more economical than Model 1. The traditional longline
cultivation system, which requires more lines than Modes 1 and
2, is the least economically viable option. The present numerical
simulation would help to understand the dynamic response of
offshore seaweed farms and help to design the optimal mooring
system of kelp farms, which can strongly affect the economic
feasibility of offshore seaweed farms.
Data availability statement
The original contributions presented in the study are included
in the article/supplementary material. Further inquiries can be
directed to the corresponding author.
Author contributions
YL: Conceptualization, Methodology, Project administration,
Supervision, Validation, Writing –review & editing, Writing –
original draft. SB: Writing –original draft. ZP: Investigation, Formal
analysis, Validation, Writing –review & editing. JZ: Conceptualization,
Funding acquisition, Writing –review & editing. WC: Funding
acquisition, Supervision, Formal analysis, Writing –review & editing.
GM: Project administration, Resources, Visualization, Writing –review
& editing. SY: Supervision, Writing –review & editing.
Funding
The author(s) declare financial support was received for the
research, authorship, and/or publication of this article. The research
work is supported by National Key R&D Program of China
(SQ2022YFB4200183), China Postdoctoral Science Foundation
(Grant No. 2022M722820), the National Key R&D Program of
China (2023YFC3007900), the National Natural Science Foundation
of China (Grant Nos. 51979050), the Fund of State Key Laboratory of
Coastal and Offshore Engineering (Grant No. LP2213), the Natural
Science Foundation of Jiang-Su Province (Grant No. BK20201314), the
Fund of State Key Laboratory of Hydraulic Engineering Simulation and
Safety of Tianjin University (Grant No. HESS-1910),the Key
R&D Program of Shandong Province, China (Grant Nos.
2023CXGC010407, 2020CXGC010702) and XPRIZE Carbon
Removal Student Award (KelpFarmCareer Team).
Conflict of interest
The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could be
construed as a potential conflict of interest.
Publisher’s note
All claims expressed in this article are solely those of the authors
and do not necessarily represent those of their affiliated organizations,
or those of the publisher, the editors and the reviewers. Any product
that may be evaluated in this article, or claim that may be made by its
manufacturer, is not guaranteed or endorsed by the publisher.
References
Abhinav, K., Collu, M., Benjamins, S., Cai, H., Hughes, A., Jiang, B., et al. (2020).
Offshore multi-purpose platforms for a Blue Growth: A technological, environmental
and socio-economic review. Sci. Total Environ. 734, 138256. doi: 10.1016/
j.scitotenv.2020.138256
Ahmed, Z. U., Hasan, O., Rahman, M. M., Akter, M., Rahman, M. S., and Sarker, S.
(2022). Seaweeds for the sustainable blue economy development: A study from the south
east coast of Bangladesh. Heliyon 8 (3), e09079. doi: 10.1016/j.heliyon.2022.e09079
Bak, U. G., Gregersen, O
., and Infante, J. (2020). Technical challenges for offshore
cultivation of kelp species: Lessons learned and future directions. Bot. Mar. 63 (4), 341–
353. doi: 10.1515/bot-2019-0005
Bak, U. G., Mols-Mortensen, A., and Gregersen, O. (2018). Production method and
cost of commercial-scale offshore cultivation of kelp in the Faroe Islands using multiple
partial harvesting. Algal Res. 33, 36–47. doi: 10.1016/j.algal.2018.05.001
Buck, B. H., and Buchholz, C. M. (2004). The offshore-ring: a new system design for
the open ocean aquaculture of macroalgae. J. Appl. Phycol. 16, 355–368. doi: 10.1023/B:
JAPH.0000047947.96231.ea
Buschmann, A. H., Camus, C., Infante, J., Neori, A., Israel, A
., Hernandez-Gonzalez,
M. C., et al. (2017). Seaweed production: overview of the global state of exploitation,
farming and emerging research activity. Eur. J. Pshycology 52 (4), 391–406.
doi: 10.1080/09670262.2017.1365175
Campbell, I., Macleod, A., Sahlmann, C., Neves, L., Funderud, J., Øverland, M., et al.
(2019). The environmental risks associated with the development of seaweed farming
in europe - prioritizing key knowledge ga ps. Front. Mar. Sci. 6. doi: 10.3389/
fmars.2019.00107
Camus,C.,Infante,J.,andBuschmann,A.H.(2018).Overviewof3year
precommercial seafarming of Macrocystis pyrifera along the Chilean coast. Rev.
Aquaculture 10 (3), 543–559. doi: 10.1111/raq.12185
Capron, M. E., Stewart, J. R., de Ramon N’Yeurt, A., Chambers, M. D., Kim, J. K.,
Yarish, C., et al. (2020). Restoring pre-industrial CO2 levels while achieving sustainable
development goals. Energies 13 (18), 4972. doi: 10.3390/en13184972
Chen, M., Yim, S. C., Cox, D. T., Yang, Z., Huesemann, M. H., Mumford, T. F., et al
(2023). Modeling and analysis of a novel offshore binary species free-floating longline
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org17
macroalgal farming system. J. Offshore Mech. Arct. Eng. 145 (2), 021301. doi: 10.1115/
1.4055803
Chu, Y. I., and Wang, C. M. (2020). Hydrodynamic response analysis of combined
spar wind turbine and fish cage for offshore fish farms. Int. J. Struct. Stability Dynamics
20 (9), 2050104. doi: 10.1142/S0219455420501047
Coleman, S., Gelais, A. T. S., Fredriksson, D. W., Dewhurst, T., and Brady, D. C.
(2022). Identifying scaling pathways and research priorities for kelp aquaculture
nurseries using a techno-economic modeling approach. Front. Mar. Sci. 9.
doi: 10.3389/fmars.2022.894461
Duarte, C. M., Wu, J., Xiao, X., Bruhn, A., and Krause-Jensen, D. (2017). Can
seaweed farming play a role in climate change mitigation and adaptation? Front. Mar.
Sci. 4. doi: 10.3389/fmars.2017.00100
Feng, D., Meng, A., Wang, P., Yao, Y., and Gui, F. (2021). Effect of design
configuration on structural response of longline aquaculture in waves. Appl. Ocean
Res. 107, 102489. doi: 10.1016/j.apor.2020.102489
Fernand, F., Israel, A., Skjermo, J., Wichard, T., Timmermans, K. R., and Golberg, A.
(2017). Offshore macroalgae biomass for bioenergy production: Environmental aspects,
technological achievements and challenges. Renewable Sustain. Energy Rev. 75, 35–45.
doi: 10.1016/j.rser.2016.10.046
Heffernan, D. (2017). News -upcoming in OrcaFlex 10.2: line statics policy (Orcina Ltd.
Company). Available at: https://www.orcina.com/news/upcoming-in-orcaflex-10-2-line-
statics-policy/.
Huang, W., Li, B., and Chen, X. (2023). A new quasi-dynamic method for the
prediction of tendon tension of TLP platform. Ocean Eng. 270, 113590. doi: 10.1016/
j.oceaneng.2022.113590
Krause-Jensen, D., Lavery, P., Serrano, O., Marbà, N., Masque, P., and Duarte, C. M.
(2018). Sequestration of macroalgal carbon: the elephant in the Blue Carbon room. Biol.
Lett. 14 (6), 20180236. doi: 10.1098/rsbl.2018.0236
Laurens,L.M.,Lane,M.,andNelson,R.S.(2020).Sustainableseaweed
biotechnology solutions for carbon capture, composition, and deconstruction. Trends
Biotechnol. 38 (11), 1232–1244. doi: 10.1016/j.tibtech.2020.03.015
Li, B. (2021a). Operability study of walk-to-work for floating wind turbine and
service operation vessel in the time domain. Ocean Eng. 220, 108397. doi: 10.1016/
j.oceaneng.2020.108397
Li, B. (2021b). Effect of hydrodynamic coupling of floating offshore wind turbine and
offshore support vessel. Appl. Ocean Res. 114, 102707. doi: 10.1016/j.apor.2021.102707
Li, B., Liu, Z., Liang, H., Zheng, M., and Qiao, D. (2023). BEM modeling for the
hydrodynamic analysis of the perforated fish farming vessel. Ocean Eng. 285, 115225.
doi: 10.1016/j.oceaneng.2023.115225
Li, B., and Wang, C. (2023). The absolute nodal coordinate formulation in the
analysis of offshore floating operations Part I: Theory and modeling. Ocean Eng. 281,
114645. doi: 10.1016/j.oceaneng.2023.114645
Lian, Y., Shen, S., Zheng, J., Boamah, S., and Yim Solomon, C. (2023a). “A
design and numerical study on new kelp culture facility,”in Proceedings of the
ASME 2023 42nd International Conference on Ocean, Offshore and Arctic
Engineering. Melbourne, Australia: Publisher of American Society of
Mechanical Engineers.
Lian, Y., Wang, R., Zheng, J., Chen, W., Chang, L., Li, C., et al. (2023b). Carbon
sequestration assessment and analysis in the whole life cycle of seaweed. Environ. Res.
Lett. 18 (7), 074013. doi: 10.1088/1748-9326/acdae9
Ma, C., Bi, C. W., Xu, Z., and Zhao, Y. P. (2022). Dynamic behaviors of a hinged
multi-body floating aquaculture platform under regular waves. Ocean Eng. 243, 110278.
doi: 10.1016/j.oceaneng.2021.110278
Morison, J. R., Johnson, J. W., and Schaaf, S. A. (1950). The force exerted by surface
waves on piles. J. Pet. Technol. 2(05), 149–154. doi: 10.2118/950149-G
Moscicki, Z., Dewhurst, T., Macnicoll, M., Lynn, P., Sullivan, C., and Chambers, M.
(2022). “Using finite element analysis for the design of a modular offshore macroalgae
farm,”in MARINE 2021.Availableat:https://arpa-e.energy.gov/technologies/programs/.
Norway Stardard 9415. (2021). Floating aqu aculture farms-site survey, design,
execution and use. Norway: NS 9415, Stardards Norway.
N’Yeurt, A. D. R., Chynoweth, D. P., Capron, M. E., Stewart, J. R., and Hasan, M. A.
(2012). Negative carbon via Ocean Afforestation. Process Saf. Environ. Prot. 90 (6),
467–474. doi: 10.1016/j.psep.2012.10.008
Olanrewaju, S. O., Magee, A., Kader, A. S. A., and Tee, K. F. (2017). Simulation of
offshore aquaculture system for macro algae (seaweed) oceanic farming. Ships Offshore
Structures 12 (4), 553–562. doi: 10.1080/17445302.2016.1186861
Pribadi, A. B. K., Donatini, L., and Lataire, E. (2019). Numerical modelling of a
mussel line system by means of lumped-mass approach. J. Mar. Sci. Eng. 7 (9), 1–24.
doi: 10.3390/jmse7090309
Qiao, D., Feng, C., Yan, J., Liang, H., Ning, D., and Li, B. (2020). Numerical
simulation and experimental analysis of wave interaction with a porous plate. Ocean
Eng. 218, 108106. doi: 10.1016/j.oceaneng.2020.108106
Schmid, M., Biancacci, C., Leal, P. P., and Fernandez, P. A. (2023). Editorial:
Sustainable seaweed aquaculture: Current advances and its environmental
implications. Front. Mar. Sci. 10, 1160656. doi: 10.3389/fmars.2023.1160656
Seghetta, M., and Goglio, P. (2020). “Life cycle assessment of seaweed cultivation
systems,”in Biofuels from Algae: Methods and Protocols, vol. pp. Ed. K. Spilling
(Springer), 103–119. doi: 10.1007/7651_2018_203
St-Gelais, A. T., Fredriksson, D. W., Dewhurst, T., Miller-Hope, Z. S., Costa-Pierce,
B. A., and Johndrow, K. (2022). Engineering A Low-Cost Kelp Aquaculture System for
Community-Scale Seaweed Farming at Nearshore Exposed Sites via User-Focused
Design Process. Front. Sustain. Food Syst. 6. doi: 10.3389/fsufs.2022.848035
Sulaiman Olanrewaju, O., Kader, A. S. A., and Wan Shamsuri, W. N. (2013). Study of
macro algae for marine biotechnology material from large scale offshore cultivation
from multiple mooring system of large aquaculture ocean floating structure. Biosci.
Biotechnol. Res. Asia 10 (2), 621–628. doi: 10.13005/bbra/1173
The 2023 Australian Minimum Wage Increase. (2023). Available at: https://www.
studyaustralia.gov.au/news/2023-australian-minimum-wage-increase.
Wang, C., Liu, J., Li, B., and Huang, W. (2023). The absolute nodal coordinate
formulation in the analysis of offshore floating operations, Part II: Code validation and
case study. Ocean Eng. 281, 114650. doi: 10.1016/j.oceaneng.2023.114650
Zhu, L., Huguenard, K., Zou, Q.P., Fredriksson, D. W., and Xie, D. (2020).Aquaculture
farms as nature-based coastal protection: Random wave attenuation by suspended and
submerged canopies. Coast. Eng. 160, 103737. doi: 10.1016/j.coastaleng.2020.103737
Lian et al. 10.3389/fmars.2024.1276552
Frontiers in Marine Science frontiersin.org18