Aquaculture America 2024

February 18 - 21, 2024

San Antonio, Texas

EVALUATING THE FEASIBILITY OF INTEGRATING MULTI-VESSEL MEMBRANE BIOLOGICAL REACTORS WITHIN RAS: EFFECTS ON WATER QUALITY, WATER USE, AND RAINBOW TROUT Oncorhynchus mykiss PERFORMANCE

John Davidson *, Curtis Crouse,  Brian Vinci,  Christine Lepine,  Megan Murray,  and Christopher Good

 

  The Conservation Fund Freshwater Institute

 1098 Turner Road

 Shepherdstown, WV 25443

 



Membrane biological reactors (MBRs) utilize  ultrafiltration membrane s that  create a clean  filtrate, while microbial processes  occurring  within mixed liquor  solids  facilitate nitrification and denitrification. Onsite research has shown that MBRs  effectively  reduce waste concentrations in  aquaculture effluents and  have potential for integrat ion within RAS.  For example, a recent  Freshwater Institute study  found that  single-vessel  MBRs  receiving  RAS  backwash  produced acceptable water quality  for  rainbow trout Oncorhynchus mykiss  and led to  significant water savings . However, a range of operational deficiencies were identified including rapid membrane fouling, incomplete denitrification , and challenges to maintain  low oxygen levels and proper mixing of the activated sludge. Th ese systems were redesigned by decoupling the membranes and  the aerobic  and anoxic zone s  for improved access and control of  each treatment process.

Six replicated RAS (9.5 m3 volume)   three of which were  integrated  with  the redesigned systems along with Mempulse® MB R modules (Dupont, Indiana, USA) The other three RAS were operated without MBRs  and  enough flushing to maintain < 75 mg/L nitrate-nitrogen (NO3-N) in the fish culture water .  Equal numbers of 1- kg rainbow trout were stocked in each RAS to begin.  Granulated sugar  was continuously added to the MBRs as a supplemental carbon source at a rate of ~ per kg feed  using repurposed belt feeders.   flow  was returned to RAS .  After 3.5- months of operation,  low-dose ozone was used 

Water use was dramatically reduced in RAS equipped with MBRs. The mean system hydraulic retention time in MBR RAS was 82 days vs. 5 days in typically operated RAS. Water quality concentrations were maintained within acceptable limits. Average NO3-N in RAS with and without MBRs was 78 ± 2 and 68 ± 1 mg/L, respectively. When operating with ozone, MBR RAS maintained comparable color and turbidity despite limited dilution. Moreover, rainbow trout growth was similar between treatments. After 4.7-months of production, trout in RAS with and without MBRs weighed 2.807 ± 0.033 and 2.800 ± 0.129 kg, respectively. Changes to MBR design and operation solved the challenges encountered during the initial study; however, new problems were identified including solids clogging behind restricted valves of small diameter (1.27 cm) pipes. Overall, however, MBRs appear to be a viable water treatment option that substantially reduces RAS water use, which could enable facility siting in water-limited areas and/or expansion of fish production volumes.