| This study examined micro-particles in a commercial aquaculture recirculating water system that produces Atlantic salmon smolts. The goal was to determine their abundance, size and physical properties of these particles with a view to eventually removing them.; Over a period of months, samples of the system waters were taken along the flow path after each major equipment in the water system. Using a field titrator, a spectrophotometer and meters, the samples were tested for water quality parameters: pH, temperature, ORP, TDS, TSS, alkalinity, hardness, nitrites, nitrates, ammonia, TKN and phosphorus. The last three parameters were digested to isolate the amount of the nutrient associated with the solids. Feed input and water flows were monitored. Samples were also studied for particle size distributions (PSD) and TSS obtained gravimetrically. The PSDs were processed to determine their proximity to a hyperbolic function. Selected samples were processed with density gradient material (DGM) (Percoll™) to determine particle densities. Microscopy was employed to assist in learning the nature and source of the particles. Where required, feed-inwater samples were produced and examined gravimetrically, for PSD and for DGM density determination.; Flow metering determined that the plant was not operating as a recirculation plant in the classic sense. Whereas the make-up water was only 10% of the circulation flow rate, between 100 and 200% of the total system water capacity was replaced per day. Thus the expected build-up of micro particles did not occur.; As expected, the feed was the main source of the solids in the system. DGM studies showed that there was a larger, heavier fraction at about 1160 kg/m3 based on the heavy cellulose fraction from whole wheat portion of the feed, which was not digested. Also there was a light fraction, about 1050 kg/m3 consisting of small particles (3–5 μm) that tended to gather into loose flocs with a binding of viscid material, probably mucopolysaccharide based. There were also a few flocs above the upper band and between the two density bands. These appeared to be of the same floc make-up as the upper band particles but with lighter or heavier components.; The hyperbolic nature of the PSD was confirmed. Gravimetric testing proved not to be useful in this environment due to the introduction and production of the viscid material clogging the meshes before sufficient weight of solids can be collected. This was a mucus-based material which also clogs drum filters in active systems.; It was concluded that most of the heavier band particles can be removed by screening. For the lighter band, of the systems reviewed, flocculation appears to offer the greatest chance of success. The implementation of this is an area for future work. |
