Engineering Feasibility Assessment of Cage Aquaculture in Offshore Wind Power Generation Areas in Taiwan

Abstract

This paper investigates the engineering feasibility of cage aquaculture in the offshore wind farm area of Changhua, Taiwan. Two types of net cages commonly used in Taiwan are compared under typhoon wave and monsoon wave conditions to determine the suitable one. The submersible function of the selected cages to reduce the typhoon wave effect is analyzed. The yearly probability of the remaining volumes of the selected cages (deformed volume/undeformed volume) under a wave scatter diagram combined with a tidal current of the study area are analyzed and discussed. The influences of water depth on the selected cages for the site selection are investigated. Regarding the seabed sand waves in the study area, two potential disasters of mooring line failure and anchor sinking on the selected cages are considered and discussed. Both issues are a concern for the wind farm developers, fishery groups, and government authorities. According to the results, the selected cage aquaculture is feasible from an engineering standpoint in the case study area.

1. Introduction

Offshore wind power is one of the green energy policies that the Taiwan government is actively promoting at present. This policy is expected to increase the proportion of green power generation to 20% by 2025, with a target of 5.7 GW for wind power generation, and an additional 10 GW for the following 10 years. However, most of the sea waters scheduled for wind farms are traditional fishing farms. The operation and daily maintenance of wind turbines derived from the installation of offshore wind turbines in these areas will directly affect the fishing activities of local fishers, which in turn will affect their livelihoods. Currently, the offshore wind industry is facing the difficulty of negotiating fishery compensation, and this is the issue with the biggest disagreement. Thus, the way to overcome this disagreement is the key to increasing the development of offshore wind farms.

At present, the marine ranch is expected to be a possible solution for fishery compensation which is commonly accepted by the groups of fishers and wind farm developers. Cage aquaculture is one type of marine ranch, and it is expected to play an important role in the conflict between wind farms and fish farms. If offshore wind power is integrated with cage aquaculture, the fishermen who originally worked in the sea area will turn to cage aquaculture, which will avoid affecting fisher’s livelihood and make full use of the offshore wind farm area [1]. Moreover, cage aquaculture has been developing for more than 20 years. Since 2000, the hydrodynamic properties of the offshore net cage have been extensively investigated through laboratory experiments, numerical simulations, and field measurements. For example, the hydrodynamics of a central-spar fish cage in a field measurement study was simulated and analyzed through transfer function by [2]. A series of experiments to investigate the forces acting on and the deformation of a net cage in a uniform current was conducted in [3]. A numerical study on the dynamic properties of the gravity cage in combined wave–current flow was carried out in [4]. The sensitivities of the mooring line tension and the remaining volume of net cages under different wave heights, wave periods, incident wave angles, current velocities, and water depths were investigated in [5]. The screen type of the Morison force model for the viscous hydrodynamic load on nets was proposed and discussed in [6]. The structural responses of high-solidity net–cage models in uniform flow were studied in [7]. Recently, the literature on the hydrodynamics of net cages has been systemically reviewed in [8,9]. Therefore, there are many tools to help assess the engineering feasibility of cage aquaculture in the offshore wind power generation area.

This study is part of a research project initiated by Taiwan Power Company in 2018 to evaluate the engineering feasibility of marine ranches installed in wind farms, as shown in Figure 1. The mooring line tension and the remaining volume of cages subjected to waves and currents [5] are two major engineering concerns in the industry. In addition, there is a special problem that needs to be overcome in this sea area, which is seabed sand waves [10,11]. For drag-embedded anchors, sand waves can cause mooring systems to fail. For gravity-type anchor blocks, scouring around the anchor can cause it to sink. Both issues are of concern to wind farm developers, government authorities, and fishing groups and require careful investigation.

In this study, the first step is to choose a suitable type of net cage from the conventional cage aquaculture for the sea area. In this manner, the existing experience can be easily transferred to the test site, and the net cage facilities of the local supply chain can reduce installation costs and increase job opportunities. According to that, this study conducts the engineering analysis for two types of net cages commonly used in Taiwan under the local sea states. The submersible function to reduce typhoon wave effects is also examined. To be a reference to the stocking density, the remaining volumes of net cages under the wave scatter diagram and tidal currents are simulated. Finally, due to the special problem of seabed sand waves in the study area, the effects of mooring line failure and anchor point sinking on the net cage are analyzed and discussed.

2. Materials and Methods

2.1. Flowchart of This Study

This study assesses the engineering feasibility of aquaculture cages installed in wind farm areas. To obtain the most concerning issues regarding the coexistence of wind farms and fishing farms, the stakeholders (fishers, wind farm developers, and government authorities) were invited to attend all the review meetings of this project to address their comments. The flowchart of this study is shown in Figure 2.

The first step is to gather background information, including local cage types and nearby sea conditions, water depths, and environmentally sensitive areas. The second step is to determine the appropriate net–cage system based on its performance under typhoon wave and monsoon wave conditions through numerical simulation. The third step is to investigate potential engineering issues that may be encountered in the study area, including remaining cage volume in annual sea conditions, submersible functionality to reduce typhoon strikes, mooring failures, and anchor sinking. Finally, we discuss with stakeholders to reach the conclusion.

2.2. Types of Net Cages Commonly Used in Taiwan

It has been more than 20 years since the development of marine cage aquaculture in Taiwan. Most cage aquaculture operators have accumulated rich experience and have been able to reduce the damage caused by natural disasters, such as typhoons. It should be possible to replicate existing net cages through technical guidance and cooperation and exchanges. The successful experience of the industry reduces the threshold of the learning curve, and at the same time, it can also use the existing supply chains to reduce the cost. Therefore, in the initial stage, Taiwan’s existing cage aquaculture system can be used first. In the future, the advanced net cage system [8] can be introduced.

There are two kinds of cage aquaculture commonly used in Taiwan. One is the submersible round cages, as shown in Figure 3a. Its design can be submerged in the water to reduce the impact of typhoon waves, so it has better wave resistance and is more suitable for open sea areas in Taiwan; the other one is the flexible square cages in Penghu Bay, as shown in Figure 3b, which have poor wave resistance and cannot sink, so they are more suitable for the sheltered inner-bay-type sea area. Both have been developed for more than 20 years. The numerical models for them have been developed and examined. Thus, both numerical models can be used as a reference for engineering feasibility assessment and site selection.

2.3. Numerical Model

Taiwan has developed its own numerical model for cage aquaculture, including the intact and failure of moorings [5,12], which has been validated through field measurements [13]. In general, the numerical method has been developed for many years and is commonly used in other studies [4,6,7]. The advantage of the self-developed model is that we can update it for specific purposes. For example, we use it to study the engineering problems of mooring failure and anchor sinking; both are new topics in the field.

In this model, the net cage is assumed to be deployed in the constant water depth and subjected to waves following a uniform current. To calculate the hydrodynamic forces acting on the net cage, the whole system is divided into small elements according to the lumped-mass method. To simplify the numerical calculations, these elements have been chosen as relatively small compared to the characteristic wavelength. This way, the scattering effect between any element and the flow field can be neglected. Then, it is appropriate to apply the Morison equation to the net cage components. All components of the net cage can be divided into two categories, namely, deformable and rigid. The deformable parts include fish nets and mooring lines, whereas the rigid parts include a non-deformable floating collar and tube sinker. In this study, each of the deformable components comprised several elements and nodes with numbers assigned. For each element, the external forces were calculated, and the resultant force was evenly distributed among the corresponding nodes. Thus, each node contained the lumped-mass of all neighboring elements and their corresponding forces. Its equation of motion can be expressed as follows:

$${\displaystyle \sum }_{j=1}^{N_e}{m}_j{\ddot{\mathit{x}}}_i={\displaystyle \sum }_{j=1}^{N_e}\left({\mathit{F}}_{D_j}+{\mathit{F}}_{I_j}+{\mathit{F}}_{B_j}+{\mathit{F}}_{W_j}+{\mathit{F}}_{T_j}\right)$$

(1)

where subscript $i$ denotes node number; $j=1,\dots ,N_e$ denotes all neighboring elements associated with node $i$; and $N_e$ represents the total number of neighboring elements. $m_j$ is the mass matrix; $\mathit{x}$ is the position vector; and ${\mathit{F}}_{D_j}$, ${\mathit{F}}_{I_j}$, ${\mathit{F}}_{B_j}$, ${\mathit{F}}_{W_j}$, and ${\mathit{F}}_{T_j}$ are the vectors of drag force, inertia force, buoyant force, gravity force, and tension force on the element $j$, respectively. The details regarding the external force modeling for the cage components (netting, floating collar, mooring line, buoy, and tube sinker) can be found in [5,13].

For the rigid components of the net–cage system including the floating collar and tube sinker, their motion can be resolved by the rigid body motion with six degrees of freedom in a three-dimensional space. The three translation motions can be expressed as follows:

$$\{\begin{array}{c}{\ddot{x}}_G={\dot{y}}_G{\omega }_3-{\dot{z}}_G{\omega }_2+\frac{1}{m_G}{\displaystyle {\displaystyle \sum }_{i=1}^N}{F}_{xi}\\ {\ddot{y}}_G={\dot{z}}_G{\omega }_1-{\dot{x}}_G{\omega }_3+\frac{1}{m_G}{\displaystyle {\displaystyle \sum }_{i=1}^N}{F}_{yi}\\ {\ddot{z}}_G={\dot{x}}_G{\omega }_2-{\dot{y}}_G{\omega }_1+\frac{1}{m_G}{\displaystyle {\displaystyle \sum }_{i=1}^N}{F}_{zi}\end{array}$$

(2)

and the three rotational motions can be expressed as follows:

$$\{\begin{array}{c}{\dot{\omega }}_1=-\frac{1}{I_1}\left(I_3-I_2\right){\omega }_2{\omega }_3+\frac{1}{I_1}{\displaystyle {\displaystyle \sum }_{i=1}^N}{M}_{1i}\\ {\dot{\omega }}_2=-\frac{1}{I_2}\left(I_1-I_3\right){\omega }_3{\omega }_1+\frac{1}{I_2}{\displaystyle {\displaystyle \sum }_{i=1}^N}{M}_{2i}\\ {\dot{\omega }}_3=-\frac{1}{I_3}\left(I_2-I_1\right){\omega }_1{\omega }_2+\frac{1}{I_3}{\displaystyle {\displaystyle \sum }_{i=1}^N}{M}_{3i}\end{array}$$

(3)

where subscripts 1, 2, and 3, and x, y, and z correspond to the body coordinate system and global coordinate system, respectively; (x_G, y_G, z_G) are the coordinates of the center of gravity; (ω₁, ω₂, ω₃) are the angular velocities along the principal axes; m_G is the body mass; (F_xi, F_yi, F_zi) and (M_1i, M_2i, M_3i) are the components of the resultant forces and moments acting at the lumped-mass node i; N is the number of nodes in the body; and (I₁, I₂, I₃) are the principal moments of inertia of the rigid body. For the floating collar and tube sinker, I₁ = 0.5m_GR², I₂ = 0.5m_GR², and I₃ = m_GR², where R is the radius of the rigid body.

In addition, the hydrodynamic forces on the netting are calculated using the screen force model [6]. In the algorithm, the whole netting structure is divided into many small screens (or elements) in which each screen contains many mesh twines. Then, the tension, buoyancy, and gravity forces that act on each screen represent the sum of the forces exerted on the twines in the screen. Finally, the total forces on the screen are assumed to be evenly distributed among the nodes. The drag coefficient $C_D\left(\beta \right)$ and lift coefficient $C_L\left(\beta \right)$ depend on flow attack angle $\beta$, and solidity $S_n$ can be determined using the experimental results by Løland [14].

$$C_D\left(\beta \right)=0.04+\left(-0.04+0.33S_n+6.54S_n{}^2-4.88S_n{}^3\right)\cos (\beta )$$

(4)

$$C_L\left(\beta \right)=\left(-0.05S_n+2.3S_n{}^2-1.76S_n{}^3\right)\sin (2\beta )$$

(5)

The solidity ratio is defined as the ratio between the area covered by the twines in the screen and total area of the screen. For a knotless net, it is expressed as follows:

$$S_n=\frac{2D}{\lambda }-{\left(\frac{D}{\lambda }\right)}^2$$

(6)

where $D$ is the twine diameter and $\lambda$ is half of the mesh size. After entering the net cage, the fluid particle velocity is slightly retarded. This phenomenon is known as the shielding effect, and a velocity reduction coefficient of 0.85 is adopted for the rear part of the nets [14]. By applying the fourth-order Runge–Kutta method (RK4) to the equation of motion of the net cage, Equations (1)–(3), we can obtain a solution with a short-time increment (10⁻³ s for this study). Then, the path for each node including the mass center of the rigid body is depicted accordingly. By computing the neighboring node’s distance, the tension force for the flexible part is obtained, and becomes the input tension data for the next time step, until the required time interval is exhausted. Note that the total numbers of nodes and elements for a four-cage system are 1166 and 1285, respectively, and the computation time for a 200-second simulation is about 7 min on a general desktop computer.

In addition, the remaining volume of cages represents the percentage value of the deformed cage volume to the undeformed cage volume, where the volume is calculated from the node positions on the netting according to the divergence theory [5]. For the simulations of mooring failure or anchor sinking, the target anchor was changed to a free node at the failed moment and allowed to move freely according to the equations of motion or to make a prescribed motion, while in the intact condition, all anchors were considered as fixed nodes.

2.4. Flowchart of the Numerical Model

In this section, we briefly describe the flowchart of the numerical model, as shown in Figure 4. The first step is to start the routine program and enter the sea states and cage parameters. The second is to mesh nodes and elements based on the lumped-mass method. The third is to calculate the external forces exerted on elements. The next step is to form the equations of motion of nodes with reference to Equations (1)–(3). Then, the equations of motion are solved by the time-marching scheme (RK4). Next, results such as mooring line tension and remaining cage volume are calculated. If the failure time is reached, the nodes and elements are updated to form new equations of motion and then continue the program. If the simulation time is reached, then the program is stopped.

3. Results

As mentioned earlier, many of the relevant engineering issues in this study are feedback from stakeholders. This section first conducts an engineering analysis of two common cages in Taiwan. Next is the functional analysis of the submersible to reduce the impact of typhoon waves. This is followed by a simulation of the remaining volume of the cage as a reference for stocking density. Finally, the effects of mooring line failure and anchor sinking caused by seabed sand waves on the net cage are discussed.

3.1. Engineering Analysis of Two Common Types of Net Cages

In this section, the engineering analysis of two kinds of net cages commonly used in Taiwan is studied through numerical simulation. In our previous study [5], the effect of different incident wave angles with a uniform current on the net cage was analyzed, and it was shown that the wave with a coplanar current is the worst case. Therefore, it is taken as the input condition. The sea conditions in the study area are shown in

Table 1. Wave and current conditions of the study area.

Scenario	Wave Height (m)	Wave Period (s)	Current (m/s)	Depth (m)
Typhoon waves	7.32	12.53	1.0	26
Monsoon waves	3.0	7.5	0.5	26

Data source: Taiwan Power Company [15].

. The first one is the 50-year return period typhoon wave and current, which is the design condition of the wind turbine structure [15]. Although the typhoon wave condition is an accidental event, the overall impact time is not long, but the force is extreme and must be included in the engineering design reference, such as the minimum breaking loads of the mooring line and the design of the anchor. Moreover, the remaining volume of net cages is also an essential factor, which can help the cage industry operator to make decisions before a typhoon comes, such as submerging the cages or reducing the stocking density. The winter monsoon is an inevitable process every year, and its influence lasts for a long time, so the analysis results are more suitable as a reference for aquaculture evaluation.

According to a preliminary economic evaluation of this project [1], the scale of cage aquaculture in the study sea area must be at least 10,000 m³ in order to be economical and cost effective. Therefore, the numerical model is based on four round cages (circumference of 60 m, net depth of 10 m, and total volume of 11,460 m³) and ten square cages (length, width, and depth are 10 m each, and the total volume is 10,000 m³) for simulation analysis. Figure 5 shows the anchor point number and the cage number description of the two types of cages.

Table 2. Input parameters of the submersible round cages.

Component	Size and Materials	Value
Floating collar	Inner circumference (m)	60
	Outer circumference (m)	64.8
	Tube diameter (m)	0.25
	Material	HDPE
Buoy	Diameter (m)	0.91
	Height (m)	1.3
	Buoyance (kN)	8.5
	Total mass (kg)	41
	Material	HDPE
Mooring line	Diameter (m)	0.05
	Unit mass (m/kg)	304/200
	Minimum breaking load (kN)	413
	Material	Nylon
Netting	Twine diameter (m)	0.0021
	Mesh size (m)	0.0604
	Solidity (m)	0.14
	Net depth (m)	10
	Material	Nylon
Tube sinker	Circumference (m)	64.8
	Tube diameter (m)	0.2
	Inserted chain (kg/m)	13
	Material	HDPE

and

Table 3. Input parameters of the flexible square cages.

Component	Size and Materials	Value
Floater	Length (m)	0.909
	Width (m)	0.606
	Height (m)	0.454
	Material	Styrofoam
Mooring line	Diameter (m)	0.05
	Unit mass (m/kg)	304/200
	Material	Nylon
Netting	Twine diameter (m)	0.0021
	Mesh size (m)	0.0604
	Solidity (m)	0.14
	Net depth (m)	10
	Material	Nylon
Sinker	Mass (kg)	25 × 4
	Material	Stone

list their specifications.

According to the above-mentioned, the following comparison of these two types of cages is based on a similar scale of aquaculture volume. In engineering analysis, it is crucial to first discuss the maximum mooring line tension and minimum remaining volume. The former is related to the safety of cage facilities, and the latter is related to the safety of farmed fish.

Figure 6 shows a comparison of the maximum tension at each anchor point of the round cages and the square cages under the monsoon wave and typhoon wave conditions (see Table 1). The results show that the maximum tension of the round cages is much larger than that of the square cages, but neither reaches the breaking strength. As generally expected, the middle anchor point (#2) has a maximum tension of 170 kN, and due to symmetry, the anchor tensions on both sides are almost the same, about 119 kN. In the winter monsoon condition, the maximum tension still occurs at the middle anchor point, but the maximum value has dropped to about 51 kN, and the anchor tension on both sides has dropped to 32 kN. It can be seen that although the overall impact of typhoon wave conditions is not long, its force is much greater than that of winter monsoon conditions, so it must be incorporated into the design of mooring system to avoid mooring line breakage or anchor-dragging events.

Figure 7 shows a comparison of the minimum value of the remaining volume for the two types of cages under monsoon wave and typhoon wave conditions. For the submersible round cages, the results show that under the typhoon wave condition, the minimum values of the remaining volume (the most serious deformation) which appear in the upstream cages (no. 1 and 3) are both 29%, and those for the rest of the cages (no. 2 and 4) are both 32%, while under monsoon wave conditions, the minimum values still appear in the upstream cages, but they remain up to 73%, while the rest are about 80% for the downstream cages. It is noteworthy that the time domain vibrations of the remaining volume are less than 10% under the monsoon wave condition, but they are up to 30% under the typhoon wave condition. For the square cages, the results show that the minimum values for the upstream cages (no. 1, 2, 3, 6, 7, and 8) are only in the range of 17–25%, and the rest are in the range of 31–35%, while under the monsoon wave condition, the minimum values are in the range of 32–56%.

Through the analysis of the above results, in the case of typhoon waves, the volume deformation of the two types of cages is very serious, and the round cages perform better than the square cages in the monsoon wave situation. Moreover, the round cages have a submersible function, so this type is more suitable for use in this sea area. After all, the typhoon is predictable and a short-time event. The impact of the typhoon can be mitigated by submerging the cages into deeper water. In addition, through aquaculture management, the stocking density can also be reduced during the typhoon season to reduce possible disaster losses. Therefore, it is recommended to use submersible round cages in this project. Moreover, the simulation results of the typhoon wave scenario (mooring line tension) are recommended for the engineering design, while the simulation results of the monsoon wave scenario (remaining volume of net cages) are recommended for evaluating the total amount of farmed fish.

3.2. Evaluation of the Submersible Function

The anti-typhoon measures for cage aquacultures include submerging the cages below the water surface to reduce the impact of typhoon waves, towing the cages to a sheltered area such as a harbor, and reducing the stocking density before a typhoon comes, etc. In this study, we only focus on the submersible function.

In engineering analysis, we usually adopt the 50-year return period typhoon wave with a coplanar uniform current to estimate the worst-case scenario [5]. However, the submersible function of net cages is designed for water waves, not work for the uniform current. Moreover, most ocean currents are classified as surface currents whose velocity profile decreases with water depth [16]. Therefore, in this study, we only consider the 50-year return period wave to analyze the submersible function, and the diving depth of the cages is 10 m below the water surface, where the wave effect is small [5].

Figure 8 shows the maximum tension per anchor in floating and submersible conditions for cages exposed to the 50-year return period typhoon wave (waves only). For the floating state, the maximum tension present at the upstream middle anchor (#2) is about 55 kN, and the remaining upstream anchors (#1 and #3) are 38 kN. On the other hand, #2 has a maximum tension of about 23 kN in the submersible state, and both (#1 and #3) are 27 kN. Overall, the submersible function can reduce the maximum tension by approximately 51% in this study. In addition, in the submersible state, the mooring tensions distributed on each anchor are relatively even.

In terms of remaining volume, the results before and after diving are shown in

Table 4. The range of remaining volume of the cages in floating and submersible states under typhoon waves (without current).

Cage Number	Floating (%)	Submersible (%)
1	78–93	89–98
2	80–93	92–98
3	78–93	89–98
4	80–93	92–98

. The minimum remaining volumes of the front two cages (no. 1 and 3) in the floating condition are about 78%. Those for the rear two cages (no. 2 and 4) are about 80%. In the submersible condition, the minimum remaining volumes of the front two cages settlement are about 89% and those for the rear cages are about 89%. It can be seen that the minimum value can be increased by about 12% after submersion. In addition, in the time domain, the remaining volume of the front two cages in the floating state varies from 78% to 93% (15%), and those of the rear two cages are from 80% to 93% (13%). After the cages are submerged 10 m from the water surface, the range of remaining volume for the front two cages becomes 89–97% (8%), and for the rear two cages is 92–98% (6%). It can be seen that the amplitude in the remaining volume after being submerged is significantly reduced. In summary, after the cages are submerged, the severity of the deformation of the net volume significantly improves, and the amplitude of the remaining volume is also smaller, which provides a better environment for the farmed fish.

3.3. Yearly Probability of Remaining Volumes of Cages

3.3.1. Under Current-Only Condition

Table 5. Comparison of the simulation of the minimum remaining volume per cage in currents and the probability of current velocity of the ADCP (TSWC1) from 1 December 2016 to 30 November 2017.

Numerical Simulation			Probability (%)
Velocity (m/s)	Cage No.	Minimum Remaining Volume (%)	Velocity (m/s)	Spring (March–May)	Summer (June–August)	Fall (September–November)	Winter (December–February)	Yearly
0.25	1	99.6	>0.25	56.8	55.8	60.7	62.0	58.8
	2	100.0
	3	99.6
	4	100.0
0.50	1	84.3	>0.50	18.6	23.5	24.3	24.5	22.7
	2	93.0
	3	84.3
	4	93.0
0.75	1	62.7	>0.75	0.8	2.4	3.3	2.2	2.2
	2	76.5
	3	62.7
	4	76.5
1.00	1	46.0	>1.00	0.0	0.0	0.0	0.0	0.0
	2	59.6
	3	46.0
	4	59.6

Data source: Taiwan Power Company [17].

lists the minimum remaining volume per cage at different current velocities and the probability of averaged velocities measured for a year. The current velocity was measured using an ADCP for a year from 1 December 2016 to 30 November 2017 [17], as shown in Figure 1. Herein, the occurrence probability of velocity is divided into spring, summer, autumn, winter, and one year to understand the seasonal variation and annual average better. The results show that the probability of occurrence of current velocity higher than 0.75 m/s is only 3.3% in Fall (September–November), and in this case, the minimum remaining volume is 62.7%. In addition, the yearly probability that the current velocity is less than 0.5 m/s is about 78.3% (=100%−22.7%), and the remaining volume can be kept above 84.3%. Overall, the higher the current velocity, the smaller the remaining volume of cages, but the smaller the probability of occurrence; this trend also appears in different months.

3.3.2. Under Combined Waves and Currents Condition

Table 6. Wave scatter diagram of the ADCP (TSWC1) from 1 December 2016 to 30 November 2017.

Probability (%)		Wave Height (m)
		0.5	1.0	1.5	2.0	2.5	3.0	3.5	4.0	4.5	5.0	>5.0
Wave period (s)	5.0	0.9	0.6	0.1
	6.0	24	8.7	3.7	2.8	2.2	1.2	0.6	0.3	0.1
	7.0	5.6	6.1	5.5	6.8	7.3	6.1	5.5	3	1.6	0.7	0.4
	8.0		0.2	0.5	0.6	1.1	1.1	0.7	0.4	0.4	0.2	0.3
	9.0			0.3	0.1		0.1	0.1
	>9.0

Data source: Taiwan Power Company [17].

shows the one-year wave scatter diagram measured using ADCP (TWSC1), see Figure 1. The maximum probability of occurrence was 24%, and it occurred in a wave with a wave height of 0.5 m and a wave period of 6.0 s. A total of 92.5% of the waves were less than 3.5 m high.

Since waves and currents often coexist in real oceans, it is important to discuss their combined effects on the hydrodynamics of cages [5]. In order to find out the most possible value of the remaining volumes of cages in the study sea area, two current velocities of 0.5 m/s and 0.75 m/s with a high probability of occurrence of 78.3% and 98%, respectively, are selected. The results are shown in Figure 9 and Figure 10, respectively. As a result, it can be seen that the minimum remaining volume decreases as the current velocity and wave height increase, and it increases with the wave period. For the current velocity of 0.5 m/s, the minimum remaining volume of the first two cages is about 59% (Figure 7), and that is about 43% for the current velocity of 0.75 m/s (Figure 8). From Table 6, the probability of wave heights less than 3.5 m is 92.5%. From Table 5, the probability of current velocities less than 0.75 m/s is 97.8%. Thus, the probability of wave heights less than 3.5 and current velocities less than 0.75 m/s is more than 90%. In summary, the remaining volume of cages can be kept at more than 40% for more than 90% of a year, which is acceptable for the industry. Note that this value is based on the local experience of cage aquaculture in Taiwan.

3.4. Influence of Water Depths on Site Selection

In the following simulation, we use the 50-year return period typhoon wave condition as the input sea state condition as shown in Table 1; the parameters of the net cages are listed in Table 2, and the selected water depths are 26 m, 30 m, 40 m, and 50 m. The first one is the average water depth of the study sea area; the rest are suitable depths regarding the submersible function. The contour plot of water depths is shown in Figure 1.

Figure 11 shows the time domain comparison of the mooring line tension at different water depths of the upstream middle anchor (no. 2) under the typhoon wave condition in Table 1. Note that the anchor selected (no. 2) is the anchor that will bear the most tension in the entire mooring system. Therefore, it is representative of engineering analysis. The results show that the magnitude and amplitude of the mooring tension are inversely proportional to the water depth; that is, the deeper the water depth, the smaller the anchor’s mooring load. In addition, the tension reduction from 30 m to 40 m is significantly higher than that from 40 m to 50 m.

Figure 12 shows the time domain comparison of the remaining volume of the front cage (no. 1) at different water depths under the 50-year return period typhoon wave condition as shown in Table 1. The results show that the deeper the water depth, the smaller the remaining volume and the smaller the response amplitude. It is noteworthy that the minimum value of the remaining volume increases the most when the water depth increases from 30 m to 40 m.

Based on the results of mooring line tension and remaining volume of net cages, we can conclude that the most suitable water depth for cage aquacultures in the study sea area should be around 40 m.

3.5. Influence of the Failure of a Mooring Line on Cages

As previously mentioned, the seabed sand wave is one of the most concerning engineering problems in the study area. In the shallower water of the eastern Taiwan Strait, strong tidal currents and abundant river sediments contributed to the formation of seabed sand waves. Due to the dynamic environment, the shape, wavelength, and height of sand waves can change significantly [10,11]. This specific nature can eventually result in the failure of a mooring system, especially for embedded anchors.

In addition, the operation and maintenance of the wind farm and adjacent areas will become frequent in the future. If cage aquaculture is set up in the wind farm, the ship’s propeller may accidentally cut the mooring line, causing the line to be damaged, or the ship may accidentally hit the cages. In both cases, the mooring system may fail at the time of the accident, or the mooring system may be damaged and unnoticed at the time of the accident, resulting in a mooring line failure event in the next severe sea conditions, such as typhoons or strong currents.

Therefore, in the study area, the failure of mooring must be considered in the engineering feasibility assessment. Generally, the safety factor of the mooring system must be considered in the design stage before construction. For example, the safety factor of the mooring system should be increased to avoid accidental line breaking or anchor dragging. Therefore, the anchor system should not fail under the design conditions of the typhoon waves Table 1. The design standard can also refer to the accidental limit state or the ultimate limit state [16]. Moreover, the mooring design must survive under the one-line failure condition to prevent triggering the domino effect of mooring [18].

3.5.1. Under Current-Only Condition

In the following simulation, the upstream mooring line (#1) is selected to fail at 40 s after the simulation under different current velocities. The results are listed in

Table 7. The maximum tension on each anchor of cages in currents under intact and failed states.

Mooring	Intact State				Failed State (#1)				Tension Ratio
Velocity (m/s)	0.25	0.5	0.75	1.0	0.25	0.5	0.75	1.0	0.25	0.5	0.75	1.0
Anchor No.	Maximum Tension (kN)								Failed Tension/Intact Tension
#1	2.61	9.75	21.01	35.02	-	-	-	-	-	-	-	-
#2	4.52	16.58	33.28	54.46	6.88	25.34	48.99	80.57	1.52	1.53	1.47	1.48
#3	2.61	9.75	21.01	35.02	2.65	10.20	22.91	38.31	1.02	1.05	1.09	1.09
#4	0.80	0.83	0.94	0.69	0.80	0.83	0.94	0.69	1.00	1.00	1.00	1.00
#5	0.65	0.95	0.75	0.83	0.65	0.95	0.75	0.83	1.00	1.00	1.00	1.00
#6	0.68	0.62	0.81	0.74	0.68	0.62	0.81	0.74	1.00	1.00	1.00	1.00
#7	1.29	4.02	7.88	12.82	1.25	3.80	7.58	12.73	0.97	0.95	0.96	0.99
#8	1.29	3.58	6.63	10.33	3.24	9.87	17.42	26.17	2.51	2.76	2.63	2.53
#9	0.70	1.87	3.51	5.67	0.77	2.27	5.68	9.87	1.10	1.21	1.62	1.74
#10	1.29	4.02	7.88	12.82	3.13	8.93	15.05	21.56	2.43	2.22	1.91	1.68
#11	1.29	3.58	6.63	10.33	1.32	3.69	7.56	12.84	1.02	1.03	1.14	1.24
#12	0.70	1.87	3.51	5.67	0.73	1.85	3.46	5.59	1.04	0.99	0.99	0.99

, including the maximum tensions on the remaining lines under intact and failed states and the tension ratio of the maximum tension at the failed state to the intact state. The results show that the higher the velocity, the higher the maximum tension of the remaining lines at the failed state and intact states. The overall maximum tension appears at anchor #2 with a tension ratio of about 1.5. It is noteworthy that two big tension ratios appear in anchors #8 and #10. That means they play an important role after one-line failure, even though the tension values on them are not large. In summary, all the maximum tension after the failure is far from the minimum breaking loads (413 kN). That means the domino effect of the mooring failure will not be triggered after one-line failure. This will not cause massive economic loss if the failure can be detected on time and repaired appropriately [18].

3.5.2. Under Combined Waves and Currents Condition

In the following simulation, two kinds of failures are considered, namely anchor #1 failed and anchor #2 failed. Both typhoon wave and monsoon wave conditions are included. The maximum tensions on each anchor under the conditions of intact, #1 failed, and #2 failed are analyzed and listed in

Table 8. The maximum tension on each anchor of cages in currents under intact and failed states.

Mooring	Intact		#1 Failed		#2 Failed
Sea States	Typhoon	Monsoon	Typhoon	Monsoon	Typhoon	Monsoon
Anchor No.	Maximum Tension (kN)
#1	119	32	-	-	175	46
#2	170	51	221	66	-	-
#3	119	32	127	33	175	46
#4	7	4	7	4	7	4
#5	15	4	15	4	15	4
#6	8	4	8	4	8	4
#7	30	18	38	18	41	20
#8	40	17	64	26	42	18
#9	26	13	38	17	30	13
#10	30	18	39	21	41	20
#11	40	17	43	16	42	18
#12	26	13	24	13	30	14

. The results show that after #1 failed, the maximum tension of the adjacent anchor #2 significantly increased from 170 kN to 221 kN under the typhoon wave situation, and that it increased from 51 to 66 kN under the monsoon wave condition. However, due to the symmetry mooring feature, after #2 failed, the maximum tension of the adjacent anchors #1 and #3 significantly increased from 119 kN to 175 kN under the typhoon wave situation, and it increased from 32 to 46 kN under the monsoon wave condition. The tension ratios of the former (#1 failed) and latter (#2 failed) in typhoon wave cases were 1.3 and 1.47, respectively. Although the tension ratio of the latter one was higher, its maximum tension (175 kN) was only 5 kN higher than that of the intact state (170 kN). In addition, the worst-case scenario was #1 failed with the maximum tension of 221 kN, which can be considered as the design condition to prevent successive failures of moorings. Moreover, it is seen that all the maximum tensions were less than the minimum breaking strength (413 kN). That means the selected net cages can survive in the study sea area even in the one-line failure scenario in a harsh marine environment.

In this study, we do not further discuss the remaining volume of net cages under the failure of mooring. However, this topic can be found in the previous study [13]. It can be described shortly that the response of the remaining volume is significantly increased at the instant of failure and then it will soon reach the steady-state condition.

3.6. Influence of the Anchor Sinking on Cages

Since the study area is a sandy sedimentary bed, under the action of waves and currents, the surrounding of the anchor block (usually a concrete block) will inevitably be subsided due to digging and brushing. Usually, this type of anchor sinking is slow. On the other hand, the seabed sand waves could also force the anchor block to subside. Generally, this type of anchor sinking is very fast. Both scenarios of the anchor sinking potentially affect the net cages. Therefore, this section will discuss the influence of anchor block sinking both slowly and rapidly on the mooring line tension and the remaining volume of net cages.

Assuming that the anchor block is made of concrete with a length of 4.2 m, a width of 4.2 m, and a height of 1.5 m, and the density of concrete is given by 2320 kg/m³, the weight of the anchor block can reach 61.4 tons. In addition, if the scour depth of the anchor block is given by two times its height, the total sinking amount of the anchor block can reach 3 m. Due to anchor #2 always suffering the largest mooring load, it is selected to be the sinking one, as the simulation of the worst-case scenario. In the following simulation, two anchor sinking scenarios are considered, namely (1) sinking rapidly: simulating anchor #2 directly sinking 3 m at 40 s after the beginning of the simulation; and (2) sinking slowly: simulating anchor #2 sinking gradually with a constant speed of 0.05 m/s from 40 s to 100 s of the simulation time.

Figure 13 shows the tension comparison of anchor no. 2 under the conditions of intact, sinking rapidly, and sinking slowly for the net cages exposed to the 50-year return period typhoon condition as shown in Table 1. The differences among them occur after the moment of anchor sinking, namely 40 s. The result shows that the sinking rapidly scenario of an anchor can result in the mooring line breaking event. However, if the anchor is sinking slowly (3 m in a minute) the maximum tension will increase but still far under the minimum breaking load. In summary, if the seabed sand wave can result in a rapid anchor sinking, it may result in the failure of mooring.

Figure 14 is a comparison of the remaining volume of the front cage (no. 1) in the intact, sinking rapidly, and sinking slowly conditions of anchor #2 of the net cages exposed to the typhoon wave condition. The results show that the effect of the anchor sinking on the net deformation is almost negligible.

4. Discussion

As far as the author knows, there are no wind farms and fish farms coexisting in the offshore areas of the world, at least not yet on a commercial scale. However, in recent years in Taiwan, the coexistence of fish pools and solar panels has been very common. Therefore, we believe that this operating model of coexistence of marine aquaculture and renewable energy will be developed in the near future. However, there are still many hurdles to overcome. Here, we discuss some topics for future study.

First, there is currently no offshore standard for the coexistence of aquaculture and wind power. In this study, input from stakeholders including fishing groups, wind farm developers, and government authorities was collected for economic [1] and engineering feasibility assessments. It can serve as an example for a local test site, but is not enough to form a standard.

Second, the current test site is farther than the conventional cage aquaculture site. The operation and maintenance costs will be much higher than the traditional model. Therefore, there is a need to develop AI-technology-based devices for monitoring water quality, disease, feeding, harvesting, etc., to reduce human resources. In addition, monitoring methods for cage remaining volume in the fields [19,20,21], failure prediction methods [22,23], and digital twins for rapid damage detection [24] need to be developed.

Third, wind farms will be built in farther and deeper sea areas in the future. Advanced aquaculture equipment [25,26] is required to survive in the harsh marine environment. The further development of aquaculture engineering is critical.

5. Conclusions

This study aimed to evaluate the engineering feasibility of marine cages installed in the offshore wind farm area using a self-developed numerical model. Based on the consideration of low technical content, low cost, and convenient use, we have selected the appropriate type of cage from traditional cage culture. Through a series of numerical simulations on general engineering topics and stakeholder concerns, we arrive at some conclusions.

Through the analysis of the remaining volume of the cages under the conditions of typhoon waves and monsoon waves, we found that the submersible round cages are better than the flexible square cages when the rearing space is similar. Additionally, the submersible cage can be submerged to a depth of 10 m to reduce the impact of typhoons on mooring loads and the remaining cage volume. The results show that in the study sea area, the probability that the remaining volume of the cages can be maintained above 40% in one year is above 90%. The results also show that a water depth of 40 m is a good choice for the location of such cages. Furthermore, we found that the rapid sinking of the anchors could lead to mooring failure. Finally, if detected in a timely manner and properly repaired, a single line failure event will not result in successive failures of the mooring.

This study can serve as an example of a local testing site, but it is not enough to form a standard. There are still many obstacles to overcome and many innovative technologies to be developed.