Experimental investigation and numerical analysis to develop low-energy large-midwater trawls

Fuel consumption in fisheries is a primary concern because of its effects on the environment and the costs incurred by fishermen. Many studies have been conducted to reduce the fuel consumption in fishing operations. Fuel consumption due to fishing gear during a fishing operation is generally related to the hydrodynamic resistance on the gear. This means that fuel consumption is proportional to the drag created by the towing speed. Based on numerical methods, this study suggests a new approach to reduce fuel consumption in fisheries. The results of the simulation are in good agreement with those of model experiments. The total as well as partial resistance forces on the gear are calculated by simulation. The simulation results suggest improved materials and gear structure for reducing the hydrodynamic forces on the gear while maintaining gear performance. The method for assessing the gear performance involves measuring the height and width of the net mouth. Furthermore, this study investigates the efficiency of a low-energy trawl from an economic point of view. The findings of this study will be useful in reducing greenhouse gas (GHG) emissions in fishing operations, and thereby contribute toward lowering fishing costs by saving fuel.


Introduction
There is growing concern regarding the impact of greenhouse gases (GHGs) on the environment. Therefore, the Kyoto Protocol, which was officially announced in 2005, was created to reduce GHG emissions and minimize environmental degradation caused by climate change. The production of GHG emissions from fisheries was one of the issues on the agenda at the Cancun conference held in Mexico in 1992. This issue was given serious consideration by developed countries, which were obligated by the Kyoto Protocol to reduce such emissions. Korea identified itself as a voluntary exclusion management country at the 16th COP of the UNFCC held in Cancún, Mexico in 2010. Korea aims to provide objectives for reducing emissions after 2013, and is considering to implement by a carbon tax, especially in businesses such as fisheries that emit GHG.
GHG emissions are proportional to the amount of fuel used in fishing. Therefore, the increases in the energy consumed during fishing result in increased GHG emissions, which is an economic burden for the marine product in-dustry with the high price of oil. Factors affecting energy consumption in fisheries include the distance a fishing vessel moves within the fishery ground, the state of the sea, the usage of equipment required for operating fishing gear and storing fish, and the gear drag. In fact, in modern-day fishing, the energy consumption is greater than the nutritional energy production (Tyedmers, 2004;Ellingsen and Aanondsen, 2006). Recent studies about the reduction of energy consumption during the fishing process, which is a major emitter of GHGs within fish production (Tyedmers, 2001;Ziegler and Hansson, 2003;Ziegler, 2007;Thrane, 2004;Schau et al., 2009) suggest that decreasing the hydrodynamic resistance of fishing gear using numerical analysis could result in lower fuel consumption (Priour and Khaled, 2009). In addition, these studies attempted to reduce fuel consumption through improvements to the hull form to decrease its resistance (Ellingsen et al., 2002). Furthermore, to achieve reductions in GHG emissions from fisheries,  quantitatively analyzed the amount of GHGs emitted during the production of marine items. Recent studies have been carried out to reduce the drag force on the trawl with adjusting of mesh orientation or net design in prawn trawl (Balash et al., 2015a(Balash et al., , 2015b.
Studies aimed at reducing the fuel consumption by changing the designs of fishing gear are still in their nascent stages. This study utilizes representative fishing gear selected to reflect similar energy usage as that of current large mid-water trawl fishing gear. The distribution of resistance is analyzed by applying numerical methods to the selected fishing gear. However, the resistance of the gear is proportional to its fuel consumption, which is, in turn, proportional to the GHG emissions of the gear. Therefore, we suggest a design for mid-water trawl fishing gear that decreases energy consumption by reducing the amount of drag through the water. It is achieved by replacing the current net with a new net material that has a smaller diameter due to its higher mechanical resistance. Additionally, the study analyzes manufacturing costs of the proposed gear and the reduction ratios of energy in accordance with operation.

Material and method
2.1 Analysis of energy usage of current large trawl fishing gear and selection of representative fishing gear According to investigation results of listening research, on-site surveys, and literature reviews pertaining to energy usage of domestic offshore large mid-water trawl fishing gear (total 52 vessels, all the 200 tons class excluding 39 trawl vessels in the East Sea), it was determined that there was no significant change in the amount of fuel usage from 2005-2010 except for in 2008 as shown in Fig. 1a (Cha, 2003). The amount of fuel used in 2008 was less, but the expenditure for fuel cost in comparison to fuel usage was higher because of the elevated oil prices. Since 2009, fuel expenditure has experienced a steep rise, further increasing the economic burden on fishery management. The annual fuel costs of a large offshore trawl fishery are shown in Fig.  1b. A trawler voyages to work 80 times per year. The average time required for a voyage is 17 h, the average trawling time is 6.88 h, and the average number of fishing operations is 2.92 (Cha, 2003). A trawler uses approximately 80% of its maximum horsepower during general navigation and 100% during trawling. 41% of the total consumed energy is used during towing.
2.2 Modeling of fishing gear resistance analysis by numerical method

The equation of motion
In this study, we divide the fishing gear system by finite mass points and describe them as physical systems connected to one another by an elastic line. For example, in the case of netting, we consider the net knots to be mass points and the lines to be springs connecting the mass points. Furthermore, we include external forces, such as the drag and lift experienced by the sinkers and floaters attached to the knots, in the analysis. Buoyancy and sinking forces act on the mass points and springs connecting these mass points through internal elastic forces as shown in Fig. 2.
The basic form of each mass point's equation of motion can be expressed as follows using Newton's Second Law of motion, where, m is the mass of the mass points, is the added mass, is the acceleration vector, represents the internal force between the mass points, and represents the external forces applied to the mass points. The mass added to the mass points is obtained from the following equation: is the density of seawater, is the volume of the mass points, and is the added mass coefficient. The value of of the floaters and sinkers, regarded as spheres, is set to 1.5 (Takagi et al., 2004;Wakaba and Balachandar, 2007;Lee et al., 2008. Cylindrical structures, such as ropes, are described by the following equation: α where is the angle of attack.

Internal forces
The internal force is comprised of the forces applied to the springs connecting the mass points. This force is gener- Jihoon LEE et al. China Ocean Eng., 2017, Vol. 31, No. 6, P. 700-708 701 ally a function of the fractional extension of the material extending along the line of the spring for each element of the lines and ropes. The change in the length of the spring is assumed to be linearly proportional to the magnitude of the applied force. The internal forces between the mass points are described as: where is the stiffness of the line (N/m), is the unit vector along the spring, is the position vector between adjacent mass points, is the magnitude of the position vector, and is the unstretched length of the spring. The unit vector, , may be obtained by dividing the vector by its magnitude, . The force, , is proportional to the elongation of the spring, and is the displacement in three-dimensional space.
The stiffness of the line, , can be described as: E A where is the modulus of elasticity and is the effective area (i.e., actual cross-sectional area) of the material.
The actual cross-sectional area (i.e., effective cross-sectional area) of a mesh bar and rope is approximately 60% of the apparent cross-sectional area. Under the same tensile load, the elongation of a cable is greater than the elongation of a solid bar made of the same material, and has the same metallic cross-sectional area because the wires in a cable tighten and behave as fibers in a mesh bar and rope. Thus, the effective elasticity modulus of a cable is less than the modulus of the original material (Gere and Goodno, 2009). In the present work, the effective modulus of the mesh bar and rope is assumed to be 65% of the base material's modulus.

External forces
External forces represent the interaction between the points and the environment (Lee et al., 2008). The forces acting on the bar are shown in Fig. 1. External forces acting on the mass points include the drag force ( ), the lift force ( ), and the buoyancy or sinking force ( ). The external forces are the sum of the individual forces and can be written as: (6) The current is assumed to be steady and uniform, and the inertial force is neglected. The drag and lift forces on the mass points affect the shape of the structure under water, and can be described as: (7) is the drag coefficient, is the density of seawater, is the projected area of the structure, and U is the magnitude of the resultant velocity vector, . The magnitude of is obtained by subtracting the velocity vector of the mass point from the current velocity vector. The vector is the unit vector for the drag and acts in the direction opposite to that of the resultant velocity vector. The quantity is the lift coefficient and is the unit vector of the lift force. The drag and lift coefficients used in this study are adopted from a previous investigation (Thresher and Nath, 1973;Lee et al., 2005;Hosseini et al., 2011).
For each element, the direction of lift forces is obtained from the vector product and the unit vector, , for a particular direction: To obtain the coefficient of the drag, the angle of attack between the velocity vector and the position vector is determined. is obtained from the following equation. . (10) The force of the gravity and the buoyancy of the structures ( ) can be described as: where is the density of the material, is the density of seawater, is the volume of the elements, and is the acceleration of gravity.

Methods of calculation
After substituting the internal and external forces in Eq. (1), the governing equation is transformed into the following non-linear second-order differential equation in the time domain: q(t) where M is the total mass and is the acceleration. Eq. (12) is transformed into two first-order differential equations: In this study, the structure is described as a stiff system of equations, which are solved using the Newmark-β (Newmark, 1959) method.
2.3 Resistance analysis of current large mid-water trawl fishing gear and large mid-water trawl fishing gear that uses low energy In this study, large mid-water trawl fishing gear consisting of a 1400 HP (i.e., 1029.7 kW) vessel currently used in large offshore mid-water trawl fisheries is selected as the representative fishing gear model for experiment and simulation. The fishing gear is designed using fishing gear design tools to perform simulations, and the designed numerical model is expressed in a three-dimensional space for simulation through an automatic modeling system (Fig. 3). To analyze the sectional resistance that occurs in the fishing gear, we separate the netting into two parts: knotted netting and knotless netting in the form of sections (i.e., Sections 1 and 2) while maintaining the characteristics of each netting component (Fig. 3).
The resistance generated by fishing gear is one of the important elements affecting fuel consumption during the fishing operation. By improving the design of the elements generating the most resistance, we can theoretically reduce the energy consumption by reducing the entire resistance that occurs in the fishing gear. Consequently, the fuel consumption difference, fuel consumption decrement per voyage, and fishing gear construction costs (depending on the net height, difference in the width of the net, and application of low-energy fishing gear) are assessed to determine the change in performance of the fishing gear and the difference observed in the resistance generated. This is accomplished through comparing current fishing gear with the modified gear having an ultra-high-molecular-weight polyethylene (UHMWPE) net with the half of the thickness of the currently used net in the knotless webbing section (Section 2) where resistance is observed in most fishing gear.

Simulation conditions
Computer simulations are performed at the towing speeds ranging from 3 to 5 knots, in the incremental step of 1 knot. The simulations are used to compare changes in the tension and shape of current large mid-water trawls on the basis of change in fluid velocity. The results are visualized using a three-dimensional graphic program. The total number of mass points generated is 2043, the computation time interval is 3×10 -4 s, the fishing gear's target is 150 m, and the sea area depth of water is 300 m. The hanging ratios of knot and knotless netting are analyzed to determine the sectional resistance and entire resistance that occurs in the fishing gear, which is divided into two sections (Sections 1 and 2) for analysis (Fig. 3).
The same towing speeds used in the analysis of current large mid-water trawl gear are applied to the low-energy large mid-water trawl gear to compare the tension and change in shape. The computation time interval is reduced to 3×10 -5 s, because the Young's modulus of the material used in the large mid-water trawl operated currently (PE) is 1.4 GPa, whereas the valid value of Young's modulus is approximately 25 GPa. The UHMPWE's Young's modulus of 55-172 GPa is used in the low-energy-based large mid-water trawl analysis. Thus, the simulation compensated by reducing the computation time interval by one-tenth of the former computation time interval.
2.5 Fishing gear resistance analysis based on fishing gear model tests The fishing gear model is developed by reducing the 1400-horse power large mid-water trawl fishing gear shown in Fig. 3 by 1/80 using the geometric similarity law. The fishing gear model's design plan is shown in Fig. 4. The net material of the fishing gear model is PE, the buoyancy of float is 16.5 g, the sinking force of the chain on the ground rope is 21 g, and additional weights for the each ground rope of 11 g are placed on both sides.
A vertical circular circulating tank with the observation The fluid velocity in this experiment is increased to 0.17 m/s, 0.22 m/s, and 0.28 m/s by adapting actual velocities of 3 knots, 4 knots, and 5 knots to the similarity rate. The motion of the model at each fluid velocity is described by measuring the tension and change in shape of the fishing gear model from the stable point. For all experimental cases, the sampling time and frequency are 20 s and 100 Hz, respectively. An average value of approximately 2000 sampling data points is used to determine the tension. The net height at each fluid velocity is calculated by measuring the location of the ground rope's core and the float line's core us-ing a digitizer. During installation, the otter board's distance is applied to each fluid velocity. The reason for selecting the installed otter board's distance is that its distance is acquired from the distance established at the design stage and the results of the simulations. A maximum distance of 85 m between otter boards is applied to the geometric similarity rate during simulation, and the tension is measured after fixing it to 1.07 m. The tension value of the mid-water trawl model is converted into a tension value that corresponds with that of the actual mid-water trawl according to the dynamic similarity. The similarity rate is defined by the following equation: where, L 2 /L 1 is the reduction ratio, and V 2 /V 1 is the velocity ratio.

Analysis of fuel consumption and fishing gear construction costs
The fuel consumption is proportional to the power used by the fishing vessel; thus, the fuel consumption is determined by the previously determined fluctuation in resistance of fishing gear as a function of towing speed. This study computes the change in amount of fuel consumption using Eq. (16) from that consumed by fishing gear (Prado, 1990), provided by the FAO, after calculating the change in resistance caused by the change in material and net thickness of the fishing gear through model and simulation method. For the comparison of construction costs between currently used gear and the proposed gear, the netting part, otter boards, and riggings' section are analyzed. Construction costs of the fishing gear are subsequently considered.
Ψ where, C is the fuel consumption, P max is the engine's maximum output (HP), and S is the regulated fuel consumption (g/HP/h), d is the fuel's specific gravity, and is the engine's motion time (h). Hyundai Company's marine engine HIMSEN, which belongs to the 1400-HP power class, is selected for computation. As per regulations, fuel consumption while towing should be 188 g/kWh and that during general navigation should be 15.04 g/kWh. Using Eq. (15), the trawl-fishing gear's netting and navigating fuel consumptions are calculated as 176.25 L/h and 141 L/h, respectively.

Results and discussions
3.1 Resistance analysis of presently operated trawl fishing gear The resistance of the fishing gear was analyzed using a model experiment and simulation at three different towing speeds. In the case of the large mid-water trawl fishing gear, the towing speed used was 4 knots. This towing speed was  taken into consideration in the interpretation of the results. At a towing speed of 4 knots, the shape and resistance of the fishing gear were found to be relatively consistent between the model experiment and the simulation. Upon analyzing the results of the numerical calculation, it was found that the resistance in Section 2 accounted for approximately 60%-70% of the netting part (i.e., Sections 1 and 2) resistance, as described in Table 1. Fig. 6 shows the underwater shape of the fishing gear, at a towing speed of 4 knots, of both the simulation and model experiment. As listed in Table 1, the sum of the resistances in Sections 1 and 2 in the case of the simulation was lower than that of the model experiment. In the simulation, we could measure the drags of Sections 1 and 2 respectively excluding the bridle. However, it was impossible to measure the drag of Sections 1, 2 and bridle part separately in the model experiment. Accordingly, this resulted from measuring the tension between the bridle of the net and the codend. Therefore, the total simulation resistance appropriated for comparison with the resistance of the fishing gear model was represented by the resistance from the bridle to the codend, and the sum of resistances of Sections 1 and 2 was relatively lower than the total resistance. Table 2 shows the changes in the net height and width in accordance with changes in fluid velocity. As the fluid velocity increased, the net height decreased and the net width increased in the simulation. However, the distance between bridles remained constant as the velocity increased in the model experiment because the bridle was fixed when the fishing gear was placed in the circulating tank. The fixed position of the bridle was acquired from the simulation results. Therefore, the simulations and model experiments were carried out under the same condition.
3.2 Resistance analysis of low-energy large-trawl fishing gear Resistance analysis of low-energy fishing gears at fluid velocities of 3, 4 and 5 knots was conducted using a model   Jihoon LEE et al. China Ocean Eng., 2017, Vol. 31, No. 6, P. 700-708 experiment and simulation. In the case of the model experiment, the fishing gear was found to have a net height and tension of 41.8 m and 7630 kgf, respectively, at the velocity of 4 knots, and in the case of the simulation, the net height and tension were found to be 35.8 m and 8435 kgf, respectively. The results showed better performance in the case of the model experiment than that in the case of the simulation. However, numerical calculation was required for interpretation because the shape and tension of the fishing gear in the simulation were similar to those of the model experiment. Table 3 lists the similar net heights and tensions. Fig. 7 shows the underwater shapes of the fishing gear at a towing speed of 4 knots. Table 4 lists the net heights and resistances of low-energy large-trawl fishing gear. For the model, these values were calculated by applying the mechanical similarity law.
3.3 Comparison of the resistance in the current and low-energy large mid-water trawl fishing gear The resistance of the current large mid-water trawl fishing gear at the towing speed of 4 knots was calculated, and the total values of the tension from the bridle to the codend were found to be 14998 kgf and 12800 kgf, for the model and simulation, respectively. The resistance of each section of the fishing gear of the simulation was acquired using a small mesh size, and the resistance accounted for 60%-70% of the entire resistance of Section 2, which was the projected area of the netting; this netting area was much larger than the other parts. In this study, the simulation was conducted using UHMWPE as the net material, which had a thickness equal to half of that of the original net. This reduced the overall resistance in the fishing gear by decreasing the resistance in Section 2, where the high resistance   previously occurred. Table 5 indicates that in the simulation, the resistance of the fishing gear at the velocity of 4 knots decreased by 44% and the net height increased by 36%. This form of the fishing gear demonstrated the expanded net height and width, which improved its shape. It was found that in Section 1, where the net thickness remained unchanged despite changes in the distribution of tension, the resistance decreased. This was because the coefficient of the resistance decreased with changes in the shape of the flexible netting, which became smooth since the power experienced at the front part of the net reduced with the decreased resistance at the back part of the net. Thus, it was concluded that flexible bodies like nets achieve stable forms by changing their shapes in the presence of the decreased resistance.
The floats on the float lines of the currently operated and low-energy trawls exhibited an unusual behavior during the experiments. When the experiments were performed, the floats were oscillating considerably with the increasing velocity, which induced an additional drag force on the gear, deforming the net mouth area. To eliminate such phenomena, the floats were replaced with others that maintained the buoyancy of the float line at the mouth of the net. We assume that a kite placed on the float line instead of floats could eliminate the extra resistance induced by oscillation of the floats with the increasing velocity.
3.4 Comparative analysis of fuel consumption and construction costs Fig. 8 shows the increase of the construction costs of the fishing gear caused by the implementation of the expensive UHMWPE material to the fishing gear. The construction cost of the trawl with UHMWPE netting is approximately 1.9 times higher than that of the currently used trawl. With the survey data pertaining to the energy usage of existing large mid-water trawl fishing gear, the fuel consumption over 10.12 h of navigation and 6.88 h of fishing operation was found to be 1426.9 L and 1212.6 L, respectively. Thus, the total fuel consumption for the entire voyage was found to be 2639.5 L as shown in Fig. 9a. Therefore, it was concluded that the fuel consumption during a fishing voyage can be reduced by 20% using the proposed design, instead of existing fishing gear, as shown in Fig. 9b.
This paper presents a study on developing a low-energy, large mid-water trawl. The resistance in the existing fishing gear was calculated and analyzed by dividing the gear into two parts. Simulations and model experiments were used in   the analysis. More resistance occurred in Section 2 of the gear. From the results of the analysis, it was found that the resistance of the fishing gear using UHMWPE was 44% less than that of the currently used fishing gear, at a towing speed of 4 knots. This can be attributed to the reduced thickness of UHMWPE to an half of that of the existing gear. Furthermore, the construction costs of gear containing the UHMWPE nets increased to 1.9 times those of the existing fishing gear. It was also verified that the fuel consumption per voyage was reduced by 20%. This study focuses on replacing the material and reducing its thickness. Further researches will be conducted to determine the effect of the mesh shape (i.e., hexagonal mesh), orientation of the mesh (i.e., T90), a kite on the float line, and efficient trawl doors on the resistance of the gear.

Conclusions
In this study, a numerical model was used to analyze the resistance gear both in sections and as a single entity, at different towing speeds. The resistance force reflected the usage of the fuel consumption during fishing operations. The cost of gear containing the low-energy netting was higher than that of the existing gear. However, the construction costs would be partially compensated by fuel cost reduction. According to the results, the following conclusions can be drawn.
(1) The resistance force of the low-energy trawl at a speed of 4 knots decreased 44% compared with the resistance force of the currently used trawls. This is due to implementation of UHMWPE that creates low resistance due to the thin thickness of the twine.
(2) Fuel consumption of the low-energy trawl was reduced 20% per a fishing voyage compared with the current type of trawl. However, the construction costs of implementing UHMWPE were 1.9 times that of the existing fishing gear.