Postprint (published version

The objective of the work is to improve the design rules of rectangular aquaculture tanks in order to achieve better culture conditions and improve water use efficiency. Particle tracking velocimetry techniques (PTV) are used to evaluate the flow pattern in the tanks. PTV is a non-intrusive experimental method for investigating fluid flows using tracer particles and measuring a full velocity field in a slice of flow. It is useful for analysing the effect of tank geometries and water inlet and outlet emplacements. Different water entry configurations were compared, including single and multiple waterfalls and centred and tangential submerged entries.The appearance of dead volumes is especially important in configurations with a single entry. Configuration with a single waterfall entry shows a zone of intense mixing around the inlet occupying a semicircular area with a radius around 2.5 times the water depth. A centred submerged entry generates a poor mixing of entering and remaining water, promoting the existence of short-circuiting streams. When multiple waterfalls are used, the distance between them is shown to have a strong influence on the uniformity of the velocity field, increasing noticeably when the distance between inlets is reduced from 3.8 to 2.5 times the water depth. The average velocities in configurations with multiple waterfalls are very low outside the entrance area, facilitating the sedimentation of biosolids (faeces and non-ingested feed) on the tank bottom. The horizontal tangential inlet allows the achievement of higher and more uniform velocities in the tank, making it easy to prevent the sedimentation of biosolids.

This work analyzes the simplest inlet and outlet configurations that create homogeneous rotating flow cells in rectangular aquaculture tanks, in order to combine the advantages of rectangular and circular tanks. All the configurations analyzed had a single jet discharge per rotating flow cell, with the drain placed in the center of each rotating flow cell. Length/width ratios (L/W) of 0.95, 1.43 and 1.91 were tested. In addition, the effect of placing oblique baffles in the walls to redirect the water currents was assessed. Experiments were conducted in a laboratory-scale tank with a Reynolds number of approximately 6000. Particle tracking velocimetry techniques were used to characterize the flow pattern in a horizontal cross-section at the midpoint of the water depth. A tank resistance coefficient (C t) was defined in order to characterize the resistance offered by each tank configuration to the circulation of water. Results indicated that when L/W was increased from 0.95 to 1.43, the main vortex that was formed occupied most of the rotating cell area and did not create significant dead volumes in the tank. A L/W ratio of 1.91 dramatically reduced flow uniformity and hardly increased C t values. The presence of baffles contributed to high velocities in the area around the center drains and decreased C t values by 30-35%. Higher velocities are critical to the self-cleaning properties of the tank. The calculation of a C t value makes it easier to obtain the desired average velocities in the tank by adjusting the water exchange rate and the water jet discharge velocity.

No abstract

The suitability of using acoustic Doppler velocimetry (ADV) to study fish swimming activity is evaluated in this study. ADV makes it possible to detect and quantify the relationship between fish density and the turbulence generated by fish swimming activity and to show differences in fish swimming patterns during the scotophase (dark period) and photophase (light period), which has been previously described by other authors. Turbulence was evaluated using the root mean square of velocity (RMS) as an indicator of fish swimming activity, and an ADV probe with an internal sampling rate of 100 Hz, which took 25 velocity data per second.Experiments at the laboratory scale using zebra fish showed a positive correlation between turbulence (RMS), caused by fish swimming activity, and density. The relationship between density and RMS was strongly linear (r 2 = 0.964). In an ongrowing farm, daily turbulence patterns caused by fish swimming activity were evaluated with sea bass at two densities: 35.5 kg m À3 (average weight of 48 g), and 11.8 kg m À3 (average weight of 11.7 g). Greater activity was detected during the photophase, indicating that light has a substantial affects sea bass swimming activity. Average RMS at a density of 35.5 kg m À3 was 3.632 and 2.428 cm s À1during photophase and scotophase, respectively, while working at a density of 11.8 kg m À3, average RMS was 1.728 and 1.419 cm s À1 during the photophase and scotophase, respectively. ADV is a rapid and reliable method to evaluate fish swimming activity at laboratory scales as well as at commercial facilities. However, ADV configuration parameters must be properly chosen in order to obtain the highest possible number of good velocity data. Data post-processing was done by filtering velocity data using correlation (COR > 70), signal-to-noise ratio (SNR > 5) and despiking filters. COR provides a measure of quality of each velocity data, ranging from 0 to 100, and SNR indicates the intensity of the reflected acoustic signal expressed in dB. Finally, despiking filter eliminates spikes generated by fish located near the probe or between the probe and point of measurement. Post-processing showed that COR filter eliminated the higher number of velocity data. #

A growth model was developed to optimize the management of multi-trophic aquaculture systems by analyzing the influence of light and biomass stocking density (SD in the productivity of Ulva ohnoi fed with the effluents from Solea senegalensis culture tanks.Experimental growth rates and productivity were determined in three flat bottom algae tanks with different incident photon irradiances ( ) (163, 280 and 886 µmol m -2 s -1 ), photoperiod 12:12h and with stocking densities ranging from 82 to 340 gdw m -2 . The distribution of photon irradiance in the algae tanks was estimated as a function of the and SD.The results obtained showed that, the algae exposed to the highest (886 µmol m -2 s -1 ) and below 170gdw m -2 experimented a strong decrease in their growth rate, together with morphological changes.The model proposed to estimate the specific growth rate ( ), on the basis of and , assumed that photosynthetic activity is dependent on the local photon flux density and, therefore, spatially distributed in the tank. The non-linear regression used to estimate the growth kinetic parameters showed a standard deviation of the distance between measured and fitted data values equal to 0.011 d -1 .In terms of biomass productivity per unit area ( ) the model shows, for each level, a trend to increase with SD, achieving a maximum , where SD can be considered optimal, and decreasing for higher SD values. The optimal SD and the maximum achievable can be also determined as a function of .

The effect of swimming fish on the average velocity and velocity profile of a circular tank was studied. Working with different inlet diameters and flow rates, nine different impulse forces (configurations) were evaluated. Each configuration was tested with and without fish, and the effects of two different fish sizes were compared.\ud \ud The velocity profiles in experiments with fish presented a considerable reduction in velocity in the centre of the tank near the outlet, which was a consequence of the increase in the kinematic eddy viscosity due to the turbulence introduced by fish swimming. A flattening of the angular velocity profile was observed in the central area of the tank, which had a radius of about 0.3 m (18% of the total volume of the tank).\ud \ud A previous model proposed by Oca and Masaló (2013) was modified in order to better describe the distribution of velocities in the central volume of a tank with swimming fish. The proposed modification was based on Burgers¿ proposal for a bathtub vortex, which implies the determination of the parameter (1¿e¿ar2)(1¿e¿ar2), where r is the radius and the a values were experimentally obtained for each tank configuration, in which they increased with the impulse force.\ud \ud The average velocities in the tank were proportional to the square root of the impulse force in experiments with and without fish. Experiments with fish presented lower average velocities, which imply higher tank resistance coefficients. At similar stocking densities (14.6 kg/m3), the increase in the tank resistance coefficients obtained with small fish sizes (154 g) were slightly higher than those obtained with bigger fish sizes (330 g).Postprint (updated version

No abstract