The performance of six multi-stage hybrid wetland systems, which were designed and constructed for treating high-content wastewater, was evaluated in the cold climate of Hokkaido, northern Japan. The systems were designed to treat four kinds of wastewater: dairy wastewater (three systems, average inflow 4.9-46.6 m 3 .d -1 , average inflow content 2,400-5,000mg.COD.l -1 , 2-5 years of operation); wastewater from a pig farm, including liquid food washing wastewater (one system, 4.1 m Keywords hybrid constructed wetland; cold climate; dairy milking parlor;potato starch processing wastewater; swine urine; oxygen transfer rate 4 IntroductionJapan's food self-sufficiency ratio is only about 40%, which is remarkably low compared to other industrialized countries. A large surplus of nitrogen and phosphorus is given to animals as feed, which is mostly imported from abroad (Yano et al. 1999).Moreover, the nitrogen and phosphorus from animal manure exceed fertilizer needs because of the small area of farmland in Japan. Treatment of dairy milking parlor wastewater, potato starch processing wastewater, and swine urine wastewater has been a big problem in Hokkaido, northern Japan, where such wastewaters are polluting rivers and groundwater. Conventional mechanical wastewater treatments are expensive. Thus, there is an urgent need for a low-cost technology for treating such wastewater.Constructed wetlands for pollution control have progressed greatly over the past 20 years (Cooper 2009;Vymazal 2009; Kadlec et al. 2000). Moreover, hybrid systems have been in use since the 1980s (Vymazal 2011). Several important studies about the selection of filter media, the treatment performance at low temperature, and the recirculation effect have been conducted in cold climates (Poldvere et al. 2009(Poldvere et al. , 2010Jenssen et al. 2005Jenssen et al. , 2010Speer et al. 2012). However, issues of clogging and freezing for treating high-content wastewater in cold climates still remain. To overcome these issues, we designed and constructed multi-stage hybrid reed bed systems in 2005 with a safety bypass and a floating cover to treat high-content wastewater in the cold climate of Hokkaido, Japan (Kato et al. 2006(Kato et al. , 2009 Mean temperature, rainfall, assessment period, and assessment number are shown in ** All systems operated throughout the year except the system for the potato starch factory P, which was designed to work from May to November. *** P1 was preserved wastewater from May to August and P2 was fresh wastewater from September to November. Schematic diagramA schematic diagram of the hybrid wetland system is shown in Fig. 1. Our systems are composed of three to four V flow beds with a self-priming siphon and no or one horizontal (H) flow bed with a total of three to five beds. A French-type self-priming siphon was applied for the V flow bed with minor modifications; each dosing pipe was single for simplicity and easy maintenance (Kato et al. 2006(Kato et al. , 2009. Some effluents were recirculated (Vr) to the inlet...

A real scale hybrid constructed wetland (CW) system (656 m2), with a configuration of VFA-VFB-HF beds constructed in series is operating since November 2006 in northern Hokkaido, Japan. The system was experimented to assess its capability in purifying 4.5 m3d−1 of high strength milking parlor wastewater under colder climate. Annual mean air temperature at site was recorded as 6.4 oC (extremes vary as -22.8 oC at lowest and 30.6 oC at highest). From the monthly sampling from November 2006 to January 2010, average loading and removal rates of TSS, CODcr, BOD5, TN and TP were 5.4 g m−2 d−1 (98%), 30.3 g m−2 d−1 (88%), 11.5 g m 2 d−1 (89%), 1.2 g m−2 d−1 (76.4%) and 0.2 g m−2 d−1 (76%). System did not stop for a single day, efficiently worked even during snow covered periods and was tolerant to the load fluctuations.

No abstract

A hybrid sub-surface constructed wetland (CW) system consisting of 2 gravel filled vertical sub-surface (VFa & VFb) beds, each 160 m(2) in size planted with Phragmites australis and a sand filled horizontal sub-surface (HF) bed, 336 m(2) in size, planted with rice was operated from 2007 to 2010 for treating milking parlor waste water in Hokkaido, Japan. Hybrid CW system received huge fluctuations in average yearly inlet loads for TSS (526.0-1259 mg L-1 & 2.7-9.0 g m(-2) d(-1)), BOD5 (1,080-2,114 mg L-1 & 8.4-14.4 g m(-2) d(-1)), COD(1,962-7,085 mg L-1 & 14.5-50.0 g m(-2) d(-1)), TN (116.0-243.0 mg L-1 & 0.8-1.6 g m(-2) d(-1)), NH4-N (54.0-90.0 mg L-1 & 0.40-0.64 g m(-2) d(-1)), TC (1,022-2,215 mg L-1 & 6.0-15.1 g m(-2) d(-1)), TP (15.3-41.7 mg L-1 & 0.11-0.28 g m(-2) d(-1)) during study period. Average yearly purification and removal rates were least fluctuated for TSS (95.7-99.4%); moderately for BOD5 (86.1-95.7%), COD(87.5-96.1%) and TC (79.5-91.3%); highly for TN (72.6-90.6%), NH4-N (62.9-85.3%) and TP (64.8-87.2%). A sharp decrease in TP purification and removal rates were observed in 2008 due to sharp decrease in influent TP concentration in 2008 compared to 2007. OTR values for VF(a), VF(b), HF bed and total system were observed as 21.7, 19.3, 4.8 and 12.3 g O-2 m(-2) d(-1) respectively. Average k value of hybrid CW system for BOD5, TN, NH4-N and TP during study period were 7.0 +/- 1.8, 7.4 +/- 3.3, 5.6 +/- 4.1 and 4.9 +/- 2.0 m yr(-1) respectively. Average concentration of TSS, TP, TN and NH4-N in the final effluent for all years were below the discharge limit value of: 150 mg L-1 for TSS; 8 mg L-1 for TP, 60 mg L-1 for TN and NH4-N. However, average BOD5 and COD concentrations could not meet the discharge limit value of 120 mg L-1 during 2007 and 2008. (C) 2013 Elsevier B. V. All rights reserved

The performance of six multistage hybrid constructed wetland systems was evaluated. The systems were designed to treat four kinds of high-content wastewater: dairy wastewater (three systems, average inflow content 2,400-5,000 mg·COD l(-1), 3-6 years of operation); pig farm wastewater, including liquid food washing wastewater (one system, 9,500 mg·COD l(-1), 3 years); potato starch processing wastewater (one system, 20,000-60,000 mg·COD l(-1), 3 years); and wastewater containing pig farm swine urine (one system, 6,600 mg·COD l(-1), 2.8 years) (COD = chemical oxygen demand). The systems contained three or four vertical (V) flow beds with self-priming siphons and surface partitions and no or one horizontal (H) flow bed (three to five beds). In some V flow beds, treated effluents were recirculated (Vr) through the inlet to improve performance. Mean annual temperature was 5-8 °C at all locations. To overcome clogging due to the high load in a cold climate, we applied a safety bypass structure and floating cover material to the V flow beds. Calculated average oxygen transfer rates (OTRs) increased proportionally with the influent load, and the OTR value was Vr > V> H. The relations of load-OTR, COD-ammonium, and a Arrhenius temperature-dependent equation enable the basic design of a reed bed system.