Hydrogen peroxide decomposition kinetics in ...

Aquacultural Engineering 64 (2015) 1–7

Contents lists available at ScienceDirect

Aquacultural Engineering journal homepage: www.elsevier.com/locate/aqua-online

Hydrogen peroxide decomposition kinetics in aquaculture water Erik Arvin a,∗ , Lars-Flemming Pedersen b a b

Technical University of Denmark, Department of Environmental Engineering, DTU Environment, Miljoevej, Build. 113, DK-2800 Kgs. Lyngby, Denmark Technical University of Denmark, DTU Aqua, Section for Aquaculture, North Sea Research Centre, PO Box 101, DK-9850 Hirtshals, Denmark

a r t i c l e

i n f o

Article history: Received 3 July 2014 Accepted 21 December 2014 Available online 30 December 2014 Keywords: Aquaculture Hydrogen peroxide Degradation Kinetics Enzyme inactivation Microbial water quality

a b s t r a c t Hydrogen peroxide (HP) is used in aquaculture systems where preventive or curative water treatments occasionally are required. Use of chemical agents can be challenging in recirculating aquaculture systems (RAS) due to extended water retention time and because the agents must not damage the fish reared or the nitrifying bacteria in the biofilters at concentrations required to eliminating pathogens. This calls for quantitative insight into the fate of the disinfectant residuals during water treatment. This paper presents a kinetic model that describes the HP decomposition in aquaculture water facilitated by microbial enzyme activity. The model describes how the hydrogen peroxide removal declines and eventually stops at relatively low chemical oxygen demand (COD) concentrations. It is hypothesized that this is due to an enzyme deficit because it is destructed due to the reactive radicals created during the HP decomposition. The model assumes that the enzyme decay is controlled by an inactivation stoichiometry related to the HP decomposition. In order to make the model easily applicable, it is furthermore assumed that the COD is a proxy of the active biomass concentration of the water and thereby the enzyme activity. This was, however, not measured. The model developed successfully described the removal of HP in aquaculture water from three types of RAS and model parameters are estimated. The model and the model parameters provide new information and are valuable tools to improve HP application in RAS by addressing disinfection demand and identify efficient and safe water treatment routines. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Hydrogen peroxide (HP) has shown promising results in the treatment of a number of different protozoan and fungal infections ˜ on fish (Schreier et al., 1996; Lilley and Inglis, 1997; AvendanoHerrera et al., 2006; Heinecke and Buchmann, 2009; Giménez Papiol and Roque, 2013). An advantage of using HP over formalin (e.g. Masters, 2004) is that HP workwise is less hazardous. HP decomposes relatively fast to oxygen and water (Block, 1991) and the break down does not include harmful disinfection by-products as is the case for other chemicals applied in aquaculture (Dawson et al., 2003), hence having the potential to be sufficiently eliminated prior to discharge and to comply with discharge regulations for land based aquaculture systems (Schmidt et al., 2006). The studies conducted with HP have mostly concentrated on the parasite treatment efficiency and the tolerance of different fish species to HP in bath treatments (Rach et al., 2000). The treatments conducted in flow-through systems or in fish tanks have included

∗ Corresponding author. Tel.: +45 40628153. E-mail address: [email protected] (E. Arvin). http://dx.doi.org/10.1016/j.aquaeng.2014.12.004 0144-8609/© 2014 Elsevier B.V. All rights reserved.

a restricted treatment period, typically carried out using hydrogen peroxide doses of around 50–100 mg/L for 30 min to 2 h (Arndt and Wagner, 1997; Gaikowski et al., 1999; Bowker et al., 2012). This intermittent treatment procedure in freshwater systems markedly differs from short term dip baths to control sea lice where HP concentrations above 1000 mg/L are applied for a few minutes (Adams et al., 2012). Use of HP in RAS is associated with prolonged contact time and precautions regarding potential impact on biofilter performance, calling for low dose applications. Knowledge on low dose ( 0) is called “catalase enzyme surplus”. When catalase enzyme deficit prevails (k1 < 0), the decomposition of HP gradually stops: rHP = − dCHP /dt → 0 and a constant HP concentration, CHP,c , is approached: From Eq. (4), CHP,c , is calculated, CHP,c = −

k1 ked

(7)

Based on this constant concentration, the solution for enzyme deficit, k1 < 0, is −1 −1 −1 −1 CHP = CHP,c + (CHP0 − CHP,c ) exp(kk1 t)

(8)

If k1 is positive, complete HP decomposition takes place. During the initial decomposition, when time t is “small” (ek·k1·t ∼ 1 + k·k1 ·t), CHP −1 is proportional with time, −1 −1 −1 −1 CHP = CHP0 + kk1 (CHP0 − CHP,c )t

(9)

After a certain time, when the second term in Eq. (6) is dominating, the HP reduction is exponential, −1 −1 CHP = (CHP0 − CHP,c )

−1 −kk t 1

e

(10)

For the special state/case where k1 = 0, i.e. the initial HP concentration balances the initial catalase enzyme concentration, then according to Eq. (6), the HP decomposition is governed by rHP = −

dCHP 2 = kked CHP dt

(11)

and the solution is −1 CHP

−1 = CHP0 + kked t

(12)

Measuring the catalase enzyme concentration in practice may be difficult or unpractical. Alternatively, a simpler, but less exact

E. Arvin, L.-F. Pedersen / Aquacultural Engineering 64 (2015) 1–7

approach may be used. We assume that the enzyme concentration is proportional with the active biomass concentration and that the initial biomass concentration is proportional with the initial COD of the aquaculture water, Eq. (13). CE0 = fe COD0

(13)

where fe is a coefficient. It appears from the analytical equation for the HP removal, Eq. (6), that the HP decomposition is controlled by the product, k k1 , and the ratio, ked /k1 : k k1 = k fe COD0 − k ked CHP0 = a COD0 − b CHP0 ,

(14)

where a = k fe

(15)

and b = k ked ked = k1

(16)

f COD0 e ked

− CHP0

−1

=

a b

COD0 − CHP0

−1 .

(17)

Since the initial concentrations of HP and COD are known (measured), the two parameters a and b are the essential parameters determining the HP degradation.

3

submerged biofilter consisted of a single BIO-BLOK® 150 HD module (Expo-Net, Hjørring, Denmark) with a volume of 0.165 m3 and a surface area of ∼25 m2 . The water volume of the entire system was 0.36 m3 , and 90 L system water was replaced by non-chlorinated tap water (Hirtshals, Denmark, a ground water based water supply without chlorination) twice a week, which gave an average daily water replacement of 7%. The fish were fed daily with 50 g commercial fish feed (DAN-EX 1754; 4 mm pellets), using belt feeders. Water temperature in the system was 15.5–16 ◦ C, oxygen concentration 80–90% saturation and the pH was 7.9–8.1. RAS-B This RAS system was located at a research facility for a commercial fish feed manufacturer holding portion sized trout. The system had a total volume of 35 m3 and was operated with a daily water exchange at approximately 6%. RAS B was equipped with a central Hydrotech® drum filter (mesh size 60-␮m), a submerged fixed bed biofilter and a swirl-separator at each tank for organic material removal. Additionally, RAS-B was equipped with an in-line UV system. Water temperature was 18 ◦ C, pH approximately 7.3 and the oxygen levels were above 80% saturation. RAS-C.

3. Materials and methods

All experiments were conducted at DTU Aqua, Section for Aquaculture in Hirtshals, Denmark. The experimental data on HP decomposition was obtained in batch experiments by spiking HP to aquaculture water from three different freshwater small scale recirculating aquaculture systems. The construction and operation of these RAS are explained in Section 3.2. Dilution series of aquaculture water were made in 1 L Pyrex glass beakers by diluting with tap water (Hirtshals, Denmark, a ground water based water supply without chlorination) whereby a predefined range of COD values were obtained. No biofilter elements were present. Table 1 summarizes the experimental conditions, including initial concentrations for HP and COD, and the number of data points for four sets of experimental runs. The temperature was 20–21 ◦ C in all experiments except in the central composite batch experiment (RAS-C-2) where temperatures were fixed at 5 ◦ C, 7.2 ◦ C, 12.5 ◦ C, 17.8 ◦ C and 20 ◦ C. The temperature in the experimental batches (20–21 ◦ C) was close to the temperature in the RAS systems from where the aquaculture water was obtained (16–18 ◦ C), see Section 3.2. Therefore, acclimatization to experimental temperature is not considered a problem. Blanks without microorganisms were prepared with MilliQ water.

This RAS had a total volume of 1.7 m3 and included a 0.5 m3 rearing unit holding 20–22 kg rainbow trout averaging 150 g/individual. The fish tank was connected to a swirl separator, followed by a pump sump delivering water to a submerged up-flow biofilter with a trickling filter in series (Pedersen et al., 2012). Water temperature was 18 ◦ C, pH 7.3–7.5 and 100–250 g feed/day was delivered with a daily make-up water of 80 L. Water was sampled at two independent experimental runs. The first experiment (RAS-C1) included water sampled from the fish tank during a period with low feeding load (100 g/d; 80 L make-up water/day) and the sample was divided into two fractions; untreated and pasteurized (75 ◦ C for 35 min) with a COD level of 45 mg O2 /L. The second experiment, RAS-C-2, a central composite design, included water sampling over 5 consecutive days at a higher feeding load (250 g/day; 80 L makeup water/day) during which 12 combinations of experimentally predefined temperature and COD levels/combinations were established similar to the experimental setup described by Pedersen et al. (2013). COD in untreated/undiluted samples ranged from 67 to 71 mg O2 /L, and water temperature was between 5.0 and 20.0 ◦ C, adjusted by use of icy bath and a refrigerated cabinet (Medilow® ; Hans Buch, Denmark). A quantity of hydrogen peroxide corresponding to a nominal concentration of 10 mg/L was added to each temperature/COD combination and sampled regularly during a period of 60 min after addition.

3.2. Recirculating aquaculture systems (RAS) studied

3.3. Analytical methods

The three different aquaculture systems described below (RASA; RAS-B and RAS-C) from where water samples were withdrawn were each operated for more than three months prior to the experiments.

HP was measured spectrophotometrically using a modified version of the method described in Tanner and Wong (1998) by a four fold increase in the OPDV complex reagent concentration. Standard curves were made with milli-Q water and the HP concentration was adjusted for potential background interference by measuring sampled water without HP with OPDV reagent. The difference between dublicate samples was generally below 0.1 mg HP/L. Chemical oxygen demand (COD) was measured on raw, unfiltered water samples using either a Hach-Lange® (Brønshøj, Denmark) test kit LCK 314 (15–150 mg O2 /L) or Hach-Lange test kit LCK 414 (5–60 mg O2 /L). For all batch experiments, the HP- and COD concentrations were measured in duplicate and the average concentrations were applied.

3.1. Experimental setup and strategy

RAS-A The system consisted of a circular 200 L polyethylene fish tank which contained between 4 and 6 kg rainbow trout (Oncorhyncus mykiss) during the test period. The water was pumped from the bottom of the fish tank to the bottom of the biofilter, which was located just above the fish tank (Pedersen et al., 2006). The water was then led back to the fish tank from the top of the biofilter. The

4


Table 1 Estimated parameters a and b (Eqs. (15) and (16)) and experimental conditions. (20–21 ◦ C). RAS system a

RAS-A RAS-Ba RAS-C-1b RAS-C-2c a b c

a m3 g−1 h−1

b m3 g−1 h−1

# data points

HP0 mg/L

COD0 mg/L

Duration of exp., h

Chi2

0.012 0.0098 0.0065 0.018

0.022 0.017 0.035 0.031

66 66 20 120

19–21 19–21 43 10–11

2–104 2–89 45 0–71

24 24 4.75 1

65 74 0.4 1582

6 COD0 levels, 11 time points. 10 time points. Five temperatures (5.0, 7.2, 12.5, 17.8 and 20.0 ◦ C), five COD0 levels, 10 time points, 12 exp. runs. (Box et al., 2005).

3.4. Parameter estimation using AQUASIM Model parameters were estimated by the simulation software AQUASIM (Reichert, 1994). We used the AQUASIM option to pool all experimental data from each campaign (RAS-A, RAS-B, RAS-C, respectively) in one parameter estimation run and thus make a parameter estimation that is the best possible with regard to all the experimental data. 4. Results and discussion 4.1. Parameter estimation Figs. 1 and 2 exemplify from RAS-A and RAS-B the experimentally observed HP decomposition and the modelled HP degradation calculated on the basis of the estimated parameters a and b (Eqs. (15) and (16)). The parameters a and b from the different RASsystems are shown in Table 1 together with the experimental conditions. The parameters are within the same order of magnitude. The parameter a varies within a factor of 2.8 and the parameter b varies within a factor of 2.1. The parameters estimated from pilot plants RAS-A and RAS-B were very similar despite different operational characteristics, such as particle removal with swirl separators, drum filtration and UV disinfection in RAS B only. There may be different reasons for the variation of the parameters a and b. The proposed decomposition model assumes that the measured COD is proportional with the microbial presence and abundance and so the catalase enzyme concentration. The applied estimator in the present study was total COD, which is a gross approximation not distinguishing between the dissolved and particulate fraction. A more accurate approximation of the microbial activity would be further quantification of particulate matter as the bacterial abundance and activity in the water phase is related to small particles (Hess-Erga et al., 2008; Wold et al., 2014). Another source of variability between the obtained parameters may be that the parameters estimated for RAS-C are based on experiments covering only the first part of the HP decomposition (1–4.5 h) and not the final constant concentration as in RAS-A and RAS-B. Generally, the model captures very well the HP decomposition from the majority of batch experiments with different COD values. At low COD levels, in particular between 2–10 mg/L, the model fitted the experimental data to a less extent. This might be due to the simplicity of the model where the enzyme concentration is assumed proportional with the COD, Eq. (13), and the fact that the COD analysis at highly diluted/oligotrophic samples is associated with significant analytical variability. According to the COD procedure (DS/ISO 15705) the precision of the measurement will be reduced at low COD levels, typically not able to sufficiently differentiate COD below 5 mg O2 /L. It remains to be resolved whether alternative measures of organic matter (e.g. permanganate index based on KMnO4 as standard oxidant, dry matter content, particle abundance and size distribution) in diluted samples can be applied to predict microbial activity. The COD levels vary from system to

system ranging from 10 to >70 mg O2 /L in commercial RAS (pers. obs.), depending on numerous factors such as input (feed loading, feed type, feed composition and digestibility) and solids removal efficiency via mechanical and biological filtration. 4.2. The HP level off and constant concentration phenomenon It appears from Figs. 1 and 2 (RAS-A and RAS-B) that there are two different scenarios taking place, a complete HP decomposition at relatively high COD, 55 mg/L and above, and a partial HP decomposition where the HP concentration levels off and approaches a constant concentration at relatively low COD, 25 mg/L and below. According to our theoretical model, the stagnation of the HP decomposition is due to destruction of the catalase enzyme. The stagnation occurs in the batches with relatively low COD values, i.e. where the initial biomass and thereby catalase concentration is relatively low. The HP decomposition in the blanks with MilliQ water was zero which is expected because there is no catalase enzyme present. The transition between full HP degradation and partial HP degradation, k1 = 0, occurs at an initial HP/COD ratio (Eqs. (5) and (13) and Table 1, average for RAS-A and RAS-B): fe CHP0 a 0.011 = = = = 0.56 COD0 0.0195 ked b The experiments with RAS-A and RAS-B water were conducted at a CHP0 = ca. 20 mg/L, i.e. the transition happens at an initial COD0 = 20/0.56 = 36 mg/L. This is consistent with the experimental observations. 4.3. Microbial facilitated HP decay The pasteurized sample from RAS-C-1 showed constant HP values over time and hence no degradation. Similarly, sampled RAS water was exposed to batch scale UV disinfection and it was found that the HP decay was approximately seven times slower compared to the untreated control (unpubl. data). These findings confirm that the HP degradation is primarily controlled by microbial enzymatic activity. 4.4. Temperature dependency The temperature dependency of the HP decomposition was estimated from the RAS-C-2 experiment where five temperature levels were applied, Fig. 3 and Table 1. The estimated temperature correction coefficient, k T = 0.1 (f(T) = ek T(T−20) ), implies that the HP decomposition rate at 10 ◦ C is 37% of the 20 ◦ C rate. Accordingly, the rate is reduced nearly by a factor of 3 when the temperature is decreased by 10 ◦ C. This is a relatively strong temperature dependency, according to Henze (2002). 4.5. Model structure The model applied in our study is a 1st order model with respect to HP and catalase enzyme concentrations (applying COD as a


5

Fig. 1. Hydrogen peroxide decomposition at different initial COD values (COD0) versus time in RAS-A. The figure shows experimental data points (average of duplicates) and modelled data (smooth line) based on parameters a and b shown in Table 1.

Fig. 2. Hydrogen peroxide decomposition at different initial COD values (COD0) versus time in RAS-B. The figure shows experimental data points (average of duplicates) and modelled data (smooth line) based on parameters a and b shown in Table 1.

Fig. 3. Data from a central composite experimental design, RAS-C-2, where hydrogen peroxide was added to different factorial combinations of water temperature (low: 5.0 ◦ C and 7.2 ◦ C; middle: 12.5 ◦ C and high: 17.8 ◦ C and 20.0 ◦ C) and organic matter levels measured as COD. The COD levels ranged from 0 to 69 mg O2 /L and were divided into 5 levels (0; 10; 35; 59 and 69 mg O2 /L). Data points are average of duplicates.

6


surrogate variable for the microbial biomass/catalase enzyme). In the work by Alvarez-Cohen and McCarty (1991a,b,c), Monodkinetics was applied. We replaced the 1st order model with Monod-kinetics in order to test whether the model fit improved (lower Chi2 ). Since the Chi2 values did not change or even increased, the model structure was not changed. 4.6. Model applications To our knowledge, it is the first time that a kinetic model for HP decomposition in aquaculture water has been presented. The model can become a very useful tool for the aquaculture operator to manage the dosage of HP to recirculating aquaculture systems. From a measurement of the COD of the aquaculture water, the operator is able calculate which initial HP should be applied in order to obtain a desired time course of HP during the disinfection scenario. The HP decomposition characteristics related to COD can be estimated from the parameters a and b in Table 1, or – for more exact predictions – the model parameters should be determined for the specific RAS. The data presented in this study were obtained in freshwater aquaculture systems. The similar reaction behaviour “HP level off and constant concentration phenomenon” are expected to occur in brackish and saltwater aquaculture systems, however, this is not tested. Measurements of HP concentration are relatively easy and can even be done using colorimetric test strips covering 1–25 mg HP/L. The correlation/relationship between COD and decay of HP (Figs. 1 and 2) can be used to assess the microbial water activity in RAS in a simple way. The operator can add HP to a water sample at T = 0 to an initial concentration HP0. After sufficient time of reaction, for example 1 h, the operator determines the HP concentration (HPT ), and from figures like Figs. 1 and 2, even with a more dense net of “COD scenarios”, the operator can determine the COD0 (∼the microbial activity) for the water sample. This COD0 is a surrogate estimate of the microbial activity. Reliable and fast assessment of microbial water quality has become pivotal with the transition from flow through aquacultures towards RAS with the significantly reduced water consumption. Applicable methods to monitor microbial water quality will provide new important operational information and allow corresponding actions to improve system performance (Martins et al., 2010; Blancheton et al., 2013; De Schryver and Vadstein, 2014) 5. Conclusions The mathematical model developed captured well the experimental data on hydrogen peroxide (HP) decomposition in three different aquaculture water systems. HP is only partly decomposed in the aquaculture water at a relatively high initial HP to COD ratio. The parameters determining the HP decomposition in the three experimental RAS plants (Table 1) are quite similar despite very different operational conditions in the plants. Temperature has a strong effect on the HP decomposition rate, emphasizing the importance of microbial activity. The rate was reduced by nearly a factor of 3 by a temperature reduction from 20 ◦ C to 10 ◦ C. Acknowledgements The study was funded by DTU Environment and DTU Aqua. The authors appreciate the technical support by Ole M. Larsen, Erik Poulsen, Brian Møller, Ulla Sproegel and Dorthe Frandsen from DTU

Aqua, Hirtshals. Also thanks to Professor Henrik Spliid at the Technical University of Denmark (DTU), Dept. of Applied Mathematics and Computer Science, for his constructive comments on the mathematical modelling. References Adams, M.B., Crosbie, P.B.B., Nowak, B.F., 2012. Preliminary success using hydrogen peroxide to treat atlantic salmon, Salmo salar L., affected with experimentally induced amoebic gill disease (AGD). J. Fish Dis. 35 (11), 839–848. Arndt, R.E., Wagner, E.J., 1997. The toxicity of hydrogen peroxide to rainbow trout Oncorhynchus mykiss and cutthroat trout Oncorhynchus clarki fry and fingerlings. J. World Aquac. Soc. 28, 150–157. Alvarez-Cohen, L., McCarty, P.L., 1991a. Effects of toxicity, aeration and reductant supply on trichloroethylene transformation by a mixed methanotrophic culture. Appl. Environ. Microbiol. 57 (1), 228–235. Alvarez-Cohen, L., McCarty, P.L., 1991b. Product toxicity and cometabolic competitive inhibition modeling of chloroform and trichloroethylene transformation by resting methanotrophic cells. Appl. Environ. Microbiol. 57 (4), 1031–1037. Alvarez-Cohen, L., McCarty, P.L., 1991c. A cometabolic biotransformation model for halogenated aliphatic compounds exhibiting product toxicity. Environ. Sci. Technol. 25 (8), 1381–1387. ˜ ˜ Avendano-Herrera, R., Magarinos, B., Irgang, R., Toranzo, A.E., 2006. Use of hydrogen peroxide against the fish pathogen Tenacibaculum maritimum and its effect on infected turbot (Scophtalamus maximus). Aquaculture 257, 104–110. Blancheton, J.P., Attramadal, K.J.K., Michaud, L., d’Orbcastel, E.R., Vadstein, O., 2013. Insight into bacterial population in aquaculture systems and its implication. Aquac. Eng. 53, 30–39. Block, S.S., 1991. Peroxygen compounds. In: Block, S.S (Ed.), Disinfection, Sterilization, and Preservation. , 4th ed. Lea & Febiger, Philadelphia, Pa, ISBN 0-683-30740-1. Bowker, J.D., Carty, D., Dotson, M.M., 2012. Efficacy of 35% PEROX-AID (hydrogen peroxide) in reducing an infestation of Gyrodactylus salmonis in freshwater-reared rainbow trout. North Am. J. Aquac. 74, 154–159. Box, G.E.P., Hunter, J.S., Hunter, W.G., 2005. Statistics for Experimenters: Design, Discovery and 405 Innovation, 2nd ed. John Wiley & Sons Inc., Hoboken, New Jersey, pp. 639, ISBN 0-471-71813-0. Dawson, V.K., Meinertz, J.R., Schmidt, L.J., Gingerich, W.H., 2003. A simple analytical procedure to replace HPLC for monitoring treatment concentrations of chloramine-T on fish culture facilities. Aquaculture 217, 61–72. De Schryver, P., Vadstein, O., 2014. Ecological theory as a foundation to control pathogenic invasion in aquaculture. ISME J. 8 (12), 2360–2368. Ely, R.L., Williamson, K.J., Hyman, M.R., Arp, D.J., 1997. Cometabolism of chlorinated solvents by nitrifying bacteria: kinetics, substrate interactions, toxicity effects and bacterial response. Biotechnol. Bioeng. 54, 520–534. Gaikowski, M.P., Rach, J.J., Ramsay, R.T., 1999. Acute toxicity of hydrogen peroxide treatments to selected lifestages of cold-, cool-, and warmwater fish. Aquaculture 178, 191–207. Giménez Papiol, G., Roque, A., 2013. Uses of hydrogen peroxide in aquaculture. In: Aguilar, G., Guzman, R.A. (Eds.), Hydrogen Peroxide: Detection, Applications and Health Implications. Nova Science Publishers, Inc., New York, USA. Heinecke, R.D., Buchmann, K., 2009. Control of Ichthyophthirius multifiliis using a combination of water filtration and sodium percarbonate: dose response studies. Aquaculture 288, 32–35. Henze, M., 2002. Basic biological processes. In: Henze, M., Harremoës, P., Jansen, J., la Cour, Arvin, E. (Eds.), Wastewater Treatment: Biological and Chemical Processes, vol. 3. Springer Edition, pp. 65–129 (Chapter 5). Hess-Erga, O.K., Attramadal, K.J.K., Vadstein, O., 2008. Biotic and abiotic particles protect marine heterotrophic bacteria during UV and ozone disinfection. Aquatic Biol. 4 (2), 147–154. Hosetti, B.B., Frost, S., 1994. Catalase activity in wastewater. Water Res. 28 (2), 497–500. Klimenko, N.A., Polyakova, T.V., Savchina, L.A., Kozyatnik, I.P., 2009. Adsorption and heterogeneous catalytic oxidation of organic matter in aqueous solutions. J. Water Chem. Technol. 31 (5), 275–285. Lilley, J.H., Inglis, V., 1997. Comparative effects of various antibiotics, fungicides and disinfectants on Aphanomyces invaderis and other saprolegniaceous fungi. Aquac. Res. 28, 461–469. Martins, C.I.M., Eding, E.H., Verdegem, M.C.J., Heinsbroek, L.T.N., Schneider, O., Blancheton, J.P., Roque d’Orbcastel, E., Verreth, J.A.J., 2010. New developments in recirculating aquaculture systems in Europe: a perspective on environmental sustainability. Aquac. Eng. 43 (3), 83–93. Masters, A.L., 2004. A review of methods for detoxification and neutralization of formalin in water. North Am. J. Aquac. 66, 325–333. Møller, M.S., Arvin, E., Pedersen, L.-F., 2009. Degradation and effect of hydrogen peroxide in small-scale recirculation aquaculture system biofilters. Aquac. Res. 42, 1113–1122. Møller, M.S., (M. Sc. thesis) 2008. Degradation of Hydrogen Peroxide in Recirculation Aquaculture System Water. Department of Environmental Engineering, Technical University of Denmark, Kgs. Lyngby. Paller, M.S., Patten, M., 1991. Hydrogen peroxide and ischemic renal injury: effect of catalase inhibition. Free Radic. Biol. Med. 10, 29–34.

E. Arvin, L.-F. Pedersen / Aquacultural Engineering 64 (2015) 1–7 Pedersen, L.-F., Sortkjær, O., Pedersen, P.B., 2006. Dose-dependent decomposition rate constants of hydrogen peroxide in small-scale bio filters. Aquac. Eng. 34, 8–15. Pedersen, L.-F., Pedersen, P.B., Nielsen, J.L., Nielsen, P.H., 2009. Peracetic acid degradation and effects on nitrification in recirculating aquaculture systems. Aquaculture 296, 246–254. Pedersen, L.-F., Pedersen, P.B., 2012. Hydrogen peroxide application to a commercial recirculating aquaculture system. Aquac. Eng. 46 (1), 40–46. Pedersen, L.-F., Good, C.M., Pedersen, P.B., 2012. Low-dose hydrogen peroxide application in closed recirculating aquaculture systems. North Am. J. Aquac. 74 (1), 100–106. Pedersen, L.-F., Meinelt, T., Straus, D.L., 2013. Peracetic acid degradation in freshwater aquaculture systems and possible practical implications. Aquac. Eng. 53, 65–71. Rach, J.J., Gaikowski, M.P., Ramsay, R.T., 2000. Efficacy of hydrogen peroxide to control parasitic infestations on hatchery-reared fish. J. Aquatic Anim. Health 12, 267–273. Reichert, P., 1994. AQUASIM – a tool for simulation and data analysis of aquatic systems. Water Sci. Technol. 30 (2), 21–30, AQUASIM can be obtained from EAWAG: Swiss Federal Institute of Aquatic Science and Technology. www.aquasim.eawag.ch Salem, I.A., El-Maazawi, M., Zaki, A.B., 2000. Kinetics and mechanisms of decomposition reaction of hydrogen peroxide in presence of metal complexes. Int. J. Chem. Kinet. 32 (11), 643–666.

7

Schreier, T.M., Rach, J.J., Howe, G.E., 1996. Efficacy of formalin, hydrogen peroxide, and sodium chloride on fungal infected rainbow trout eggs. Aquaculture 140, 323–331. Schmidt, L.J., Gaikowski, M.P., Gingerich, W.H., 2006. Environmental Assessment for the Use of Hydrogen Peroxide in Aquaculture for Treating External Fungal and Bacterial Diseases of Cultured Fish and Fish Eggs. U.S. Geological Survey, Washington, D.C. Report, pp. 180. Schwartz, M.F., Bullock, G.L., Hankins, J.A., Summerfelt, S.T., Mathias, J.A., 2000. Effects of selected chemo-therapeutants on nitrification in fluidizedsand biofilters for coldwater fish production. Int. J. Recirc. Aquac. 1, 61–81. Tanner, P.A., Wong, A.Y.S., 1998. Spectrophotometric determination of hydrogen peroxide in rainwater. Anal. Chim. Acta 370, 279–287. Tort, M.J., Fletcher, C., Wooster, G.A., Bowser, P.R., 2003. Stability of hydrogen peroxide in aquaria as fish disease treatment. J. Appl. Aquac. 14, 37–45. Wahman, D.G., Henry, A.E., Katz, L.E., Speitel Jr., G.E., 2006. Cometabolism of trihalomethanes by mixed culture nitrifiers. Water Res. 40, 3349– 3358. Wold, P.A., Holan, A.B., Øie, G., Attramadal, K., Bakke, I., Vadstein, O., Leiknes, T.O., 2014. Effects of membrane filtration on bacterial number and microbial diversity in marine recirculating aquaculture system (RAS) for Atlantic cod (Gadus morhua L) production. Aquaculture 422, 69–77.