Desalination 202 (2006) 2430

A wind-driven reverse osmosis system for aquaculture wastewater reuse and nutrient recovery
C.C.K. Liua*, W. Xiaa, J.W. Parkb
a

Department of Civil and Environmental Engineering and Water Resources Research Center, University of Hawaii at Manoa, 2540 Dole Street, Honolulu, HI 96822, USA Tel. +1 808 956-7658; Fax +1 808 956-5014; email: clarkliu@hawaii.edu b Department of Civil Engineering, Hanyang University, Seoul 133-791, South Korea
Received 31 July 2005; accepted 23 December 2005

Abstract A wind-driven reverse osmosis system for aquaculture wastewater treatment, developed at the University of Hawaii at Manoa, was tested at the experimental facilities on Coconut Island, Oahu, Hawaii. With this technology, a fish tank becomes a closed aquaculture production system with zero waste discharge. The permeate (freshwater) from the system can be used as the freshwater supply for fish culture, while the brine (concentrated wastewater) can be further processed into fish feed by a duckweed-covered reactor. Keywords: Aquaculture wastewater; Reuse; Reverse osmosis

1. Introduction A prototype wind-driven reverse osmosis (RO) system was constructed in 19971998 by University of Hawaii researchers. The system was successfully tested for brackish water desalination during 19982001 [1]. The system can be operated at a moderate wind speed. At an average wind speed of 5 m/s, brackish feedwater at a total dissolved solids concentration of 3000 mg/L
*Corresponding author.

and at a flow rate of 13 L/min can be processed. The average rejection rate was 97%, and the average recovery ratio was 20%. The energy efficiency of 35% is comparable to the typical energy efficiency of well-operated multivaned windmills. Later, research was conducted on the application of this system for the removal of nitrogenous wastes from the culture water of Oreochromis niloticus (tilapia). Nitrogenous wastes can cause many environmental problems to the receiving water. To protect its pristine coastal water, the

Presented at the conference on Wastewater Reclamation and Reuse for Sustainability (WWRS2005), November 811, 2005, Jeju, Korea. Organized by the International Water Association (IWA) and the Gwangju Institute of Science and Technology (GIST).
0011-9164/06/$ See front matter 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.desal.0000.00.000

C.C.K. Liu et al. / Desalination 202 (2006) 2430

25

state of Hawaii established stringent water quality limits for nitrogen in freshwater aquaculture wastewater effluent. The ammonia nitrogen (NH3-N) concentration in effluent discharged into the coastal water must be less than 20 mg/L, and the concentration of nitrite and nitrate nitrogen (NO -N + NO2 -N) must be less than 35 g/L. 3 At present, effluent discharged from most aquaculture facilities in Hawaii exceeds these limits. Thus the development of cost-effective nitrogen-removal technology is essential for establishing a sustainable and profitable aquaculture industry in Hawaii. Test results of the wind-driven system, which separates the aquaculture wastewater passing through the RO membrane into permeate (freshwater) and brine (concentrated wastewater), indicated that freshwater can be processed and recycled [2]. As for the brine, it must be further processed before being discharged into the environment or, preferably, before being reused. In

this study, a duckweed-based tank system was developed for further brine processing. The use of aquatic macrophytes such as duckweed for wastewater treatment has drawn great attention worldwide in recent years [35]. Duckweed-based treatment of municipal wastewater has been studied at laboratory-, pilot-, and full-scale levels [69]. These studies indicated that, in addition to direct duckweed uptake, nitrogenous wastes are removed in a duckweed treatment system by the biological activities of bacteria and other microorganisms suspended in the water column [10]. Although the ability of duckweed treatment systems for nitrogen removal has been well documented, none of the previous studies investigated the reaction kinetics in a duckweed reactor with ammonia nitrogen of less than 1.0 mg/L. The performance of a duckweed-covered reactor at a low nutrient level is therefore the focus of this study.

Windmill

RO Module Permeate Brine

Pressure Tank

Prefilter

Water Supply Ground Surface Post-treatment

Piston Pump

Solar Energy

Data Logger Photovoltaic System Computer Center Solenoid Valve Data Transportation Line

Brackish Water Tank Flow/Pressure Sensor Check Valve

Electric Power

Fig. 1. Wind-driven reverse osmosis system.

26

C.C.K. Liu et al. / Desalination 202 (2006) 2430

2. Development of a wind-driven reverse osmosis system in Hawaii In 1998, construction of the prototype winddriven RO brackish water desalination system was completed and the system successfully tested on Coconut Island. The system was part of a research effort aimed at desalinating coastal brackish water for use as an alternative freshwater

supply by Pacific island communities [1]. It combines fully developed windmill and membrane technology and modern control theory. Fig. 1 is a schematic of the prototype system. Feedwater, which is pressurized by a wind pump, flows into a pressure stabilizer that reduces large fluctuations of pressure and the flow rate. A relatively stable flow of feedwater from the stabilizer

(a)

(b)

(c)

(d)

(e)

Fig. 2. The experimental site and major system components: (a) Coconut Island, Oahu, Hawaii, (b) windmill, (c) stabilizer, (d) RO module, and (e) control device.

C.C.K. Liu et al. / Desalination 202 (2006) 2430

27

then passes through a cartridge filter, which is a pre-treatment unit to remove contaminants, before entering the RO module. The stabilizer, developed at the University of Hawaii, is a kind of hydropneumatic pressure tank with a 0.3-m3 inner volume; it yields a mean hydraulic detention time of about 30 min under design conditions. A 4.3-m (14-ft) diameter multi-blade windmill installed on a 9-m-tall tower drives a piston pump with a 275-mm (11-in.) stroke and 980-cm3 effective displacement. Both the windmill and piston pump were manufactured by Dempster Inc. (Beatrice, Nebraska, USA). An ultra-low-pressure RO membrane, M-T4040ULP, manufactured by Applied Membrane Inc. (Vista, CA, USA) was used. The effective surface area of a single RO unit is 7.40 m2 (80 ft2), and the design operating pressure range is 3501200 kPa (50175 psi). The feedback control is accomplished by a feedback flow/pressure control device, which is made of three parallel sets of solenoid/throttle valves (Fig. 1). Water pressure signals in the stabilizer are sent through the pressure sensor to the data logger. The data logger evaluates these signals and then sends a command to open one or more sets of solenoid/throttle valves. This control mechanism allows the system to operate continuously and efficiently. The feedback control mechanisms are powered by solar energy, so no commercial electricity is needed. Coconut Island and four system components (windmill/pump, pressure stabilizer, RO module, and data acquisition and control mechanisms) are shown in Fig. 2. 3. Aquaculture wastewater treatment and freshwater recovery Application of the wind-driven RO system was later extended into aquaculture wastewater treatment, for which pilot-scale experiments have been conducted on Coconut Island since 2000 [2]. Aquaculture wastewater passing through the

RO membrane is separated into permeate (freshwater) and brine (concentrated wastewater). Test results indicate that the prototype system can process aquaculture wastewater at flow rates ranging from 230 to 370 L/h. The permeate, at a quality suitable for fish production, is recirculated to the fish tank (Fig. 3). The brine is sent back to the storage tank, where it mixes with the wastewater from the fish tank (Fig. 3). Occasional discharge of brine is necessary, as the nitrogen concentration in the storage tank builds up over time. 3.1. Nitrogen concentration in permeate and in brine Experimental results showed that the ammonia nitrogen concentrations in the permeate remained below 0.02 mg/L (Fig. 4), making the permeate suitable for use in fish culture. On the other hand, the ammonia nitrogen concentration in the brine ranged from 0.40 to 1.20 mg/L, so further brine processing is required. 3.2. Freshwater recovery The rate of freshwater recovery depends on wind speed and the frequency of brine discharge. Experiments were conducted to study the rate of freshwater recovery under brine recirculation periods of 2, 4, and 6 h. Without brine recirculation, the recovery rate would not increase appreciably with wind speed, and vice versa (Fig. 5). 4. Brine processing and nutrient recovery Various options to process occasionally discharged brine were evaluated. Use of the duckweed Spirodela spp. for nitrogen removal from brine, at a relative low concentration of less than 2 mg/L, was selected (Fig. 3). Compared with water hyacinths, duckweed provides a smaller surface attachment area for microbial growth. Duckweed usually forms a dense surface mat

28

C.C.K. Liu et al. / Desalination 202 (2006) 2430

Fig. 3. Aquaculture wastewater treatment with water reuse and nutrient recovery.

Fig. 4. Nitrogen concentration in permeate and in brine, with a 6-h brine discharge frequency. Fig. 5. Freshwater recovery at varying wind speeds and brine discharge frequencies.

covering the entire water surface. The mat provides special treatment system characteristics. In past studies of nitrogen removal by duckweeds, three design factors were considered, namely, nutrient concentration, duckweed reactor

(tank) depth, and surface area of reactor. This study included duckweed density as an additional design factor. Four stocking densities were evaluated to identify the optimal duckweed

C.C.K. Liu et al. / Desalination 202 (2006) 2430

29

Fig. 6. Duckweed-covered reactors.

stocking density for nitrogen removal at low concentrations (Fig. 6). In a well-designed reactor, uptake by duckweed was found to be the major nitrogen-removal mechanism; other mechanisms such as nitrification were relatively unimportant. Experimental results are shown in Fig. 7. The surface loading rate in this treatment system depends on the effective surface area, which is the area covered by duckweed. Desirable duckweed density, or the percentage of surface area covered by duckweed, must be determined experimentally for individual systems. 5. Concluding remarks The wind-driven RO process was applied successfully for nitrogen removal from aquacultural wastewater. The freshwater produced by this treatment process can be used by recirculating it directly back to the fish tanks. The brine produced by this process can be further treated

by duckweed-covered reactors. The duckweed can be manufactured into fish feed. It was demonstrated that duckweed plays an important role in the N-absorption process under conditions of both high and low nutrient levels. The duckweed density should be considered in the design of the treatment for wastewater with a low nutrient level. The reaction rates in a duckweed-covered reactor can be correlated with

Fig. 7. First-order nitrogen-removal coefficients as a function of surface loading.

30

C.C.K. Liu et al. / Desalination 202 (2006) 2430

wastewater for effluent renovation and protein production, Biotechnol. Bioeng., 29 (2) (1987) 258268. [4] S.C. Reed, R.W. Crite and E.J. Middlebrooks, Natural Systems for Waste Management and Treatment, 2nd edn., McGraw-Hill, New York, 1995. [5] P. Van der Steen, A. Brenner, J. Van Buuren and G. Oron, Post-treatment of UASB reactor effluent in an integrated duckweed and stabilization pond system, Water Res., 33 (3) (1999) 615620. [6] L. Mandi, Marrakesh wastewater purification experiments using vascular aquatic plants Eichhornia crassipes and Lemna gibba, Water Sci. Technol., 29 (1994) 283287. [7] O. Hammouda, A. Gaber and M.S. Abdel-Hameed, Assessment of the effectiveness of treatment of wastewater-contaminated aquatic systems with Lemna gibba, Enzyme Microbial Technol., 17 (1995) 317323. [8] G.J. Alaerts, M.D. Rahman Mahbubar and P. Kelderman, Performance analysis of a full-scale duckweed-covered sewage lagoon, Water Res., 30 (4) (1996) 843852. [9] E. Awuah, H.J. Lubberding, K. Asante and H.J. Gijzen, The effect of pH on enterococci removal in Pistia-, duckweed- and alga-based stabilization ponds for domestic wastewater treatment, Water Sci. Technol., 45 (1) (2002) 6774. [10] P.G. Bonomo and N. Zambon, Advantages and limitations of duckweed-based wastewater treatment systems, Water Sci. Technol., 35 (1997) 239246.

modified surface loading rates, which include consideration of the duckweed density. Acknowledgements This work was supported in part by the U.S. National Science Foundation (award no. INT0090193) and by the U.S. Geological Survey (grant no. 01HQGR0079). This is contributed paper CP-2006-04 of the Water Resources Research Center, University of Hawaii at Manoa, Honolulu. Any opinions, findings, and conclusions in this publication are those of the writers and do not necessarily reflect the views and policies of the U.S. National Science Foundation or the U.S. Geological Survey. References
[1] C.C.K. Liu, J. Park, R. Migita and G. Qin, Experiments of a prototype wind-driven reverse osmosis desalination system with feedback control, Desalination, 150 (2002) 277287. G. Qin, C.C.K. Liu, N. Richman and J. Moncur, Aquaculture wastewater treatment and reuse by wind-driven reverse osmosis membrane technology: a pilot study on Coconut Island, Hawaii, Aquacult. Eng., 32 (2005) 365378. G. Oron, D. Porath and H. Jansen, Performance of the duckweed species Lemna gibba on municipal

[2]

[3]