Water quality in recirculating aquaculture system using woodchip denitrification and slow sand filtration

Abstract

In recirculating aquaculture system (RAS), ammonium excreted by the fish is typically transformed to less toxic nitrate by microbial activity in bioreactors. However, nitrate-nitrogen load can be harmful for the receiving water body when released from the RAS facility. A new type of water treatment system for a RAS was designed, including a passive woodchip denitrification followed by a sand filtration introduced into a side-loop of an experimental RAS, rearing rainbow trout (Oncorhynchus mykiss). In the process, woodchips acted as a carbon source for the denitrification, aiming at a simultaneous nitrogen removal and reduction of water consumption while sand filtration was used to remove organic matter and recondition the circulating water. A variety of chemical analyses and toxicological tests were performed to study the suitability of the process and to ensure the absence of harmful or toxic substances in the system. The results did not show increased toxicity, and no increased mortality was reported for the raised species. After the start-up of the system, the concentrations of fatty acids (e.g., hexadecanoic acid <LOD-1.21 mg L^-1) and heavy metals (e.g., Cd < LOD-0.45 μg L^-1, Pb < LOD-14 μg L^-1) remained at very low levels and below those of known toxic effects. In the beginning of the experiment, good denitrification efficiency was achieved, but it declined after 1 month, showing the need for improved stability and dimensioning of the application.

Keywords: Gas chromatography (GC); Heavy metals; Inductively coupled plasma mass chromatography (ICP-MS); Inductively coupled plasma optical emission spectrometry (ICP-OES); Ion chromatography (IC); Rainbow trout; Recirculating aquaculture system (RAS).

Fig. 1

A flowchart of the experimental setup, showing a fish tank (FT), swirl separator, drum filter, fixed-bed reactor (FBBR), moving bed reactor (MBBR), trickling filter (TF), and a side-loop with a woodchip bioreactor (WCBR), and a sand filter (SF)

Fig. 2

Nitrate removal (%, ± SD, n = 24) in the woodchip bioreactor (a) and in the sand filter (b) after 3, 6, and 9 weeks of the experiment

Fig. 3

Concentrations of chloride (Cl⁻, a), nitrate-N (NO₃-N, b), sulfate (SO₄²⁻, c), and phosphate (PO₄³⁻, d) (mg L⁻¹, ± SD, n = 4) in circulating water after the woodchip bioreactor during the 10 weeks of the experiment

Fig. 4

Concentrations (mg L⁻¹, ± SD, n = 4) of benzoic acid, hexadecanoic acid, and octadecanoic acid in the circulation water, after the woodchip bioreactor (a) and the sand filter (b), in small and large side-loops during the 10 weeks of the experiment

Fig. 5

Concentrations of calcium (Ca), potassium (K), magnesium (Mg), phosphorous (P), and sulfur (S) (mg L⁻¹ ± SD, n = 4) in circulating water after the woodchip bioreactor (a small side-loop, b large side-loop) and after the sand filter (c small side-loop, d large side-loop) during the 10 weeks of the experiment