This article was downloaded by: [University of Connecticut]On: 11 October 2014, At: 07:49Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

International Journal ofPhytoremediationPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/bijp20

The Macro Nutrient Removal Efficienciesof a Vertical Flow Constructed WetlandFed with Demineralized Cheese WheyPowder SolutionArda Yalcuk aa Abant Izzet Baysal University, Faculty of Engineering andArchitecture, Department of Environmental Engineering , Bolu,TurkeyPublished online: 16 Dec 2011.

To cite this article: Arda Yalcuk (2012) The Macro Nutrient Removal Efficiencies of a Vertical FlowConstructed Wetland Fed with Demineralized Cheese Whey Powder Solution, International Journal ofPhytoremediation, 14:2, 114-127, DOI: 10.1080/15226514.2011.582381

To link to this article: http://dx.doi.org/10.1080/15226514.2011.582381

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

International Journal of Phytoremediation, 14:114–127, 2012Copyright C© Taylor & Francis Group, LLCISSN: 1522-6514 print / 1549-7879 onlineDOI: 10.1080/15226514.2011.582381

THE MACRO NUTRIENT REMOVAL EFFICIENCIES OF AVERTICAL FLOW CONSTRUCTED WETLAND FED WITHDEMINERALIZED CHEESE WHEY POWDER SOLUTION

Arda YalcukAbant Izzet Baysal University, Faculty of Engineering and Architecture,Department of Environmental Engineering, Bolu, Turkey

This study aims to remove the macro-sized nutrients that are present in the cheese wheypowder solution through the use of constructed wetland systems. For this purpose, 70% and40% demineralized solutions of cheese whey powder were used. For both concentrations,control reactors are run in parallel with Typha angustifolia planted reactors for the durationof a 92 day period. Zeolite and gravel were used as the filling material. The planted reactor,which was fed with the 70% solution, was named as Cheese Whey Powder Solution (CWPS) 1and its unplanted control was named CWPS 2 while the reactor, which was fed with the 40%solution, was named as CWPS 3 and its unplanted control was named CWPS 4. The removalof COD, PO4-P and NH4-N were obtained as 37.47%, 45.62%, and 68.88% in CWPS 1;24.89%, 35.74%, and 63.15% in CWPS 2; 51.15%, 54.96%, and 64.13% in CWPS 3; and28.35%, 23.99%, and 65.92% in CWPS 4, respectively.

KEY WORDS cheese whey powder, constructed wetland, demineralization, Typha angusti-folia, Zeolite

INTRODUCTION

The worldwide annual cheese whey production is estimated to be over 108 tons.Cheese whey is an important source of environmental pollution since 10 L of cheesewhey with high carbohydrate, protein, and lipid content remains from the productionof 1 kg of cheese (Ozmıhcı and Kargı 2008). Due to the presence of these valuablecompounds, the whey stream can be used in the food industry. However its high salt contentrenders demineralization necessary (Uribe et al. 2006). This is a two-step process; first thedemineralization of the whey solution that was obtained from different kinds of cheeseis carried out and then the whey is dried through pulverization (Astosan Dairy Products).Demineralized cheese whey is preferably used mainly in infant nutritional formulations,but it is also used in many other products such as ice cream, bakery products, confectionary,and animal feed. The major ions that would be removed from whey are Na+, K+, Ca+2,Mg+2, Cl−, HPO4

−, citrate, and lactate. Ion exchange demineralization of cheese whey

Address correspondence to Arda Yalcuk, Abant Izzet Baysal University, Faculty of Engineering and Ar-chitecture, Department of Environmental Engineering, Golkoy Campus 14280, Bolu, Turkey. E-mail: ayalcuk@gmail.com

114

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4

NUTRIENT REMOVAL FROM CWPS BY CONSTRUCTED WETLAND 115

generally requires a strong cation exchanger. This can attain more than 90% reduction insalt content, which is necessary for baby food formulae. Lower levels of demineralization,obtained using a bypass system, may be adequate for other applications due to the high saltcontent of whey, however, the system must be regenerated following the treatment of 10–15bed volumes of whey. This is achieved by the treatment of cation and anion exchangersseparately with strong acids and alkali respectively. Typically a cycle is about 6 h, of which4 h are required for regeneration, therefore two or three systems working in parallel may berequired. The use of counter current regeneration reduces the consumption of regenerationchemicals (Brennan et al. 2006).

Dairy wastewater liberated during the production of cheese contains whey, whichmay cause major deterioration in the quality of the water due to its high production volumeand the high organic matter content with a BOD5 ranging from 30,000 to 50,000 ppm.In this type of effluent, proteins, carbohydrates and to a lesser extent, lipids are largelyresponsible for the high COD and BOD (Farnet et al. 2009).

Successful treatment of the polluted water would reduce the pollution loading sothat the land application could remain a viable treatment option for dairy farmers. Breweret al. (1999) identified the possible polluted water treatment systems to include aeration,mechanical separation and anaeorobic digestion. Constructed wetlands have also been usedfor the treatment of agriculturally polluted water (Sun et al. 1998).

Phytoremediation is an emerging technology, which uses plant and microbial commu-nities from the rhizosphere to eliminate various organic or inorganic chemical contaminants.The enzymatic activities that are involved in this transformation are mainly performed bymicroorganisms from the rhizosphere. Plant roots support microbial growth by aerating thesoil and plant leaves allow the thermoregulation of the soil. Thus, the purifying potential ofsuch a system is promising since it favors the microbial activities in several ways (Gasuinas2005).

In a study that was conducted by McMillan, the salt tolerance of Typha angustifoliawas reported to be more pronounced than that of Typha latifoli (McMillan 1959). Sincethe wastewater that was used in the study had a salinity of 2.6–4%, the subsurface verticalflow wetlands were planted with Typha angustifolia in this study. On the other hand, Typhaangustifolia is widely used and is known to be highly tolerant for being used with varioustypes of wastewater (Kantawanichkul 2009).

The main objectives in this study were (1) to determine the removal of macro nutrients,(2) to compare the wetland efficiency for 70% and 40% demineralized cheese whey powdersolution.

MATERIALS AND METHODS

Cheese Whey Powder

The cheese whey powder that was used in the study was provided by Astosan DairyProducts (Gonen-Balıkesir/Turkey). Demineralized whey powder is manufactured fromfresh whey of the finest quality, which is filtered and defatted prior to demineralization byat least 40% and 70% demineralization (UF), followed by evaporation and spray drying.

The cheese whey powder solution (CWPS) was obtained by 1:20 dilution of the 2%solution of 70% and 40% cheese whey powder in the study. Table 1 shows the characteristicsof the 40% and the 70% demineralized cheese whey powder.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4

116 A. YALCUK

Table 1 Typical chemical & physical properties of cheese whey powder

Typical Values

70% Demineralized 40% DemineralizedProperties Cheese Whey Powder Cheese Whey Powder

Moisture, m/m 3.5% 3.5%Fat, m/m 0.5% 0.5%Protein, m/m (Nx6,38) 7% 7%Ash, m/m 3.2% 5.5%Lactose, m/m 82% 77%Titrable acidity (L.A.) 0.13% 0.14%pH (in 10% solution) 6.20 6.20Salt, m/m 2.6% 4%Density, g/cm3 0.65 0.65Solubility index, mL 0.5 0.5

Reactors

In this study, cheese whey powder solution (CWPS) was treated by using the con-structed wetland systems operated in subsurface flow mode for the whole of the bed mediaalong which the waste water flowed. The experiments lasted 100 days including the plan-ning and the acclimation periods. At the start up, the reactor systems were filled with dilutedCWPS and the continuous feeding started after a few days. The CWPS was diluted withtap water in order to maintain a safe influent COD concentration for the prevention of toxiceffects on the plants.

Four cylindrical PVC reactors with internal diameters of 21 cm and heights of 23 cmwere used in the experimental study. For the 70% cheese whey powder solution, one ofthe reactors (CWPS1) was planted with Typha angustifolia and the other reactor was leftunplanted as control (CWPS2). The same protocol was applied for the 40% cheese wheypowder solution and the planted reactor was CWPS3 while CWPS4 was left barren. Afiltering mechanism was present at 8 cm above the baseline. The treated outlet water wascollected at the bottom of the filter where a global valve was placed 6 cm above the baselevel.

Both cheese whey dust solutions were fed to the reactor systems continuously (forthe duration of 5 min feeding/hour) by a peristaltic pump (Ismatec VC 280), which mighthave enabled air trapping within the porous filling material in between the wastewaterfeeding periods. With this feeding arrangement 5 l/d of OMW passed through the systemproviding hydraulic retention times of 4.2 days for the systems CWPS1, CWPS2, CWPS3,and CWPS4 (Figure 1). The effluent was not put through recirculation in this study.

Zeolite and gravel were used as the filling materials in the reactors. The characteristicsof the reactor systems are summarized in Table 2. Zeolite (clinoptilolite) was used inaddition to the commonly used sand and gravel media in order to improve the performanceof the removal mechanisms (ion exchange, adsorption) in the vertically constructed wetlandsystems. Highly efficient and highly applicable method of “ion exchange” has a significantrole among the recently used methods in terms of the removal of ammonia. Zeolite hasa high selectivity for ammonia. In addition, the biofilm layer that has been formed overthe filling material would aid in the nitrification-denitrification processes and the uptake ofammonia (Mayo and Mutamba 2004).

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4

NUTRIENT REMOVAL FROM CWPS BY CONSTRUCTED WETLAND 117

Figure 1 Reactors were used in experimental studies.

Table 2 The characteristics of the reactors

Vol Layer Thickness Medium Size Q HRT OLRSystem (l) (cm) (mm) (l/day) (day) (kgCOD/m3.day)

0–8 FilterCWPS1 5 8–15.5 Fine gravel (7–15) 1.2 4.2 0.345

15.5–23 Zeolite (0.8-2)0–8 Filter

CWPS2 5 8–15.5 Fine gravel (7–15) 1.2 4.2 0.34515.5–23 Zeolite (0.8-2)

0–8 FilterCWPS3 5 8–15.5 Fine gravel (7–15) 1.2 4.2 0.374

15.5–23 Zeolite (0.8–2)0–8 Filter

CWPS4 5 8–15.5 Fine gravel (7-15) 1.2 4.2 0.37415.5–23 Zeolite (0.8-2)

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4

118 A. YALCUK

Analytical Methods

The efficiencies of the reactor systems were evaluated by measuring the organic andthe inorganic parameters. The influent and the effluent samples were analyzed according toU.S. Standard Methods called the COD (APHA 5220-D), the ammonia-N (NH4-N) (APHA4500-NH3-D), the orthophosphate (PO4-P) (APHA 4500-PC) protocols, and using a MerckPharo 100 spectrophotometer (APHA 2005). The EC and pH measurements were carriedout using the Thermo Scientific Orion 5 Star series EC meter and pH meter.

Statistical Analysis

In order to evaluate the wastewater treatment performance of the CWPS 1-2 andCWPS 3-4 reactors, the independent sample t-test analysis was performed. Independentsample t-test analysis (at a significance level of 0.05) was carried out on the removalefficiencies for the 92 day monitoring period for the removal quality parameters. Thisstatistical analysis was conducted using SPSS (PASW 18) software. The results of thestatistical analysis are presented in the following form: (t, df, p) where t = t value; df =degrees of freedom, and p > 0.05, as described in the Results and Discussion section.

RESULTS AND DISCUSSION

The performance data regarding the experimental vertical flow constructed wetlandsystems are presented in the following sections, investigation of the removal of COD,PO4-P, NH4-N, and the changes in EC and pH.

Removal of COD

The main routes for the removal of organic carbon in wetlands include volatilization,photochemical oxidation, sedimentation, sorption, and biodegradation. Organic contami-nants are absorbed onto particles flowing into the wetlands which settle out in the quiescentwater and are then broken down by the microbiota in the sediment layer. Organic moleculesare broken down by the microbiota through fermentation and aerobic/anaerobic respirationand they are later utilized as a source of energy or are assimilated into biomass. The effi-ciency and the rate of organic carbon degradation by microorganisms are highly variableand they depend on the type and the amount of the organic compounds present in theinfluent. Volatilization may also be considered as a significant removal mechanism in themicrobial breakdown of organic products (ITRC 2002).

The 70% and the 40% demineralized CWPS were fed to the reactor systems ata concentration of 1437.79 ± 300.63 mg/L, and 1558.67 ± 71.61 mg/L, respectively.The exit concentrations for the CWPS1 and the CWPS2 reactors that were fed with 70%CWPS were 898.3 ± 269.88 mg/L and 1089.43 ± 363.56 mg/L, respectively. The effluentconcentrations for the CWPS3 and the CWPS4 reactors that were fed with 40% CWPSwere 767.44 ± 322.54 mg/L and 1120.86 ± 331.92 mg/L, respectively. Lactose is largelyresponsible for the presence of high COD and BOD (Farnet 2008). The demineralizedcheese whey powder that was used in the study had a high level of 82–77% lactose in itscontent. This aspect leads to the thinking that the source of the COD in the cheese wheypowder was mainly lactose.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4

NUTRIENT REMOVAL FROM CWPS BY CONSTRUCTED WETLAND 119

The COD removal for the 70% demineralized CWPS was calculated as 37.47 ±14.29% and 24.89 ± 15.85% for the planted CWPS1 and the barren CWPS2, respectively(Figure 2a). The COD removal for the 40% demineralized CWPS was calculated as 51.15 ±19.44% and 28.35 ± 19.73% for the planted CWPS3 and the barren CWPS4, respectively(Figure 2b.).

The presence of stable organic compounds in the environment, which could not bebiologically degraded, has caused the COD removal values to be rendered low. Anotherreason for the low removal efficiencies would be lower organic loads associated with thestudy when compared to other studies. Farnet et.al. (2008) have obtained 84% BOD re-moval efficiency using cheese dairy farm effluents in a constructed soil reed bed. They haveobtained 90.75% COD removal efficiency in another study that they have conducted. Woodet al. (2007) have obtained BOD removal efficiency results ranging between 58–99.6% inconstructed soil reed beds in various studies. A statistically significant difference betweenCWPS1 and CWPS2 exists in terms of the analysis that was carried out for 70% demineral-ized CWPS (t = 2.194, sd = 26, p = 0.037, p < α, α = 0.05). Also a statistically significantdifference between CWPS3 and CWPS4 exists in terms of the analysis that was carriedout for 40% demineralized CWPS (t = 3.081, sd = 26, p = 0.005, p < α, α = 0.05). Theinvestigation of the COD removal from the demineralized cheese whey powder solutionyielded that better removal was obtained when 40% solutions were used (Figure 2c). Anaverage of 32.76% removal was obtained for the 40% CWPS in the first 43 days of the studywhereas an average removal of 69.54% was obtained between days 50–92. An average of30.79% removal was obtained for the 70% CWPS in the first 43 days of the study whereasan average removal of 44.00% was obtained between days 50–92. The plant was observedto grow better in the 40% solution and the investigation of both planted reactors yieldedthat at the end of the 43rd day, an observable growth difference existed between the plantsin the reactors. The systems have been operated under laboratory conditions. The ambienttemperature of the laboratory has been kept at 16◦C for the first 43 days whereas it hasbeen raised to 22◦C for the remaining time being. Consequently, based on this increase inthe ambient temperature, the number of microorganisms increased and formed a biofilmlayer. This condition is thought to result in an increase in the COD removal. The presenceof several species provides a more propitious habitat which encourages the developmentof a great diversity of microbial communities (Zurita et al. 2009). This situation was alsostatistically confirmed with a statistically significant difference existing between CWPS3and CWPS1 (t = –2.133, sd = 26, p = 0.043, p < α, α = 0.05). A statistically significantdifference could not be observed between the CWPS% values of the plant-free reactors(CWPS4-CWPS2) (t = 0.511, sd = 26, p = 0.613, p > α, α = 0.05).

Removal of PO4-P

The removal of phosphate from wetland systems should be carried out throughadsorption by the sediments or by the direct uptake of phosphate into a plant (Bonomoet al. 1997). The long term phosphorous removal mechanisms in wetlands depend on theamount and the type of the plant residues, the amount and the type of the filling materialand the presence of Al/Fe compounds in the medium (Sakadevan and Bavor 1998). Insystems using gravel as the filling material, the main phosphorous removal mechanismis reported to be carried out through filtration, utilization via bacteria and the uptake bythe plants (Korkusuz et al. 2005). Plant residues may also dissipate PO4-P back into theenvironment. Phosphorous is usually only stored for a short period in plants such as Typha

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4

120 A. YALCUK

Figure 2 The variations in COD removal. (a) For 70% CWPS solution; (b) For 40% CWPS solution; (c) For theplanted reactors.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4

NUTRIENT REMOVAL FROM CWPS BY CONSTRUCTED WETLAND 121

latifolia and most of it is liberated in plant residues (Vymazol 2007). The 70% and the40% demineralized CWPS was fed into the systems with a PO4-P concentration of 15.05 ±1.80 mg/L and 29.13 ± 3.56 mg/L, respectively. The outlet concentrations for the 70%CWPS were 8.04 ± 0.77 mg/L for CWPS1 and 9.51 ± 1.51 mg/L for CWPS2. The outletconcentrations for the 40% CWPS were 12.86 ± 1.35 mg/L and 22.16 ± 2.47 mg/L forCWPS3 and CWPS4, respectively. The PO4-P removal for the 70% demineralized CWPSwas calculated as 45.62 ± 9.32% and 35.74 ± 13.46% for the planted CWPS1 and thebarren CWPS2, respectively (Figure 3a). The PO4-P removal for the 40% demineralizedCWPS was calculated as 54.96 ± 8.61% and 23.39 ± 7.87% for the planted CWPS3 andthe barren CWPS4, respectively (Figure 3b). A statistically significant difference betweenCWPS1 and CWPS2 exists in terms of the 70% demineralized CWPS (t = 2.256, sd = 26,p = 0.033, p < α, α = 0.05). Also a statistically significant difference between CWPS3 andCWPS4 exists in terms of the 40% demineralized CWPS (t = 10.119, sd = 26, p = 0.0,p < α, α = 0.05). The investigation of the PO4-P removal from the demineralized cheesewhey powder solution yielded that better removal was obtained in 40% solutions than in70% solutions.

The PO4-P removal in 70% and 40% cheese whey powder solution in planted andbarren reactors was approximately 13% and 31%, respectively. The plant heights weredetermined as 62.13 ± 10.25 cm in 70% solution and as 75.68 ± 10.14 cm in 40% solution.Based on this result, plant growth was determined as more effective in 40% solution. Inaddition, investigation of Table 1 yields that the ash and salt content of 40% solution wasapproximately twice the content of 70% solution. The biofilm layer was also observedto be thicker in 40% solution. Taking all this information together with PO4-P removalmechanisms into consideration, the difference between the reactors CWPS3 and CWPS4was thought to result from these differences.

A statistically significant difference existed between the planted CWPS3 and CWPS1reactors (t = –2,753, sd = 26, p = 0.011, p < α,α = 0.05). A similar statistically significantdifference could also be observed between the CWPS% values of the plant-free reactors(CWPS4-CWPS2) (t = 2,963, sd = 26, p = 0.006, p < α,α = 0.05) (Figure 3c).

Removal of NH4-N

Nitrogen in the form of nitric oxide, nitrite, or ammonia/ammonium is soluble inwater and can reach water ways from a variety of sources, including non-point sources(agricultural runoff), and point sources (wastewater treatment plant discharges). N removalin a constructed wetland system includes uptake by plants and other living organisms,nitrification, denitrification, ammonia volatilization, and cation exchange for ammonium(Yalcuk and Ugurlu 2009).

The NH4-N concentration of the 70% CWPS influent was observed to fluctuate be-tween 1.2–2.55 mg/L with an average value of 2.28 ± 0.41 mg/L. The average effluentconcentration was 0.73 ± 0.39 mg/L for CWPS1, and 0.84 ± 0.43 mg/L for CWPS2.The NH4-N concentration of the 40% CWPS influent was observed to fluctuate between2.15–3.14 mg/L with an average value of 2.63 ± 0.41 mg/L The average effluent concen-tration was 0.85 ± 0.59 mg/L for CWPS3, and 0.82 ± 0.42 mg/L for CWPS4. During the92 days period, the average percent removals were 68.88.06 ± 14.27%, 63.15 ± 16.41%,64.13.02 ± 29.42%, and 65.92 ± 22.52%, for CWPS1, CWPS2, CWPS3, and CWPS4, re-spectively (Figure 4). Farnet et al. (2009) have reported 75.65% removal of the total Kjedahl

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4

122 A. YALCUK

Figure 3 The variations in PO4-P removal (a) For 70% CWPS solution; (b) For 40% CWPS solution; (c) For theplanted reactors.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4

NUTRIENT REMOVAL FROM CWPS BY CONSTRUCTED WETLAND 123

Figure 4 The variations in NH4-N removal (a) For 70% CWPS solution; (b) For 40% CWPS solution.

nitrogen from the cheese-dairy farm effluents using subsurface wetland treatment mech-anisms. A statistically meaningful difference could not be observed between the plantedand the plant-free reactors for either 40% CWPS or the 70% CWPS [CWPS1-CWPS2 (t =0,916, sd = 26, p = 0.368, p > α, α = 0.05), CWPS3-CWPS4 (t = –0.181, sd = 26, p =0.591, p > α, α = 0.05)] using Typha angustifolia wastewater treatment systems.

Nitrification of NH4-N through microorganisms is relatively a slower process due tothe low respiratory rate of the autotrophic nitrification bacteria in comparison to other Nremoval mechanisms (Conolly et al. 2003). Adsorption is a fast reversible process on thecontrary to the slow nitrification process. Based on this, NH4-N removal was thought to beinitially conducted through adsorption and later as the capacity of the filling material andthe plant was full, the removal efficiency decreased until days 36–78 in CWPS 3–4 anduntil the 71st day in CWPS1-2, recovering through the aid of the microorganisms as thebiofilm layer matured in the environment.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4

124 A. YALCUK

Figure 5 pH changes (a) For 70% CWPS solution; (b) For 40% CWPS solution.

Zeolite, one of the filling materials, was expected to be effective in the removal ofammonia. Approximately 32% of the filling material mixture in each reactor is composedof zeolite and therefore the similarity in percentage removal values could be attributed tothis situation.

It is known that the Na+, K+, and Mg+2 cations interfere with the NH4+ ions and that

zeolite decreases the amount of NH4+ ions through ion selectivity. The organic materials

in the waste water and the solid particles in suspension may fill in the channel gaps of thezeolite and may cover its surface thereby reducing the ammonia removal capacity (Pınar andUgurlu 2003). Minerals such as Na+, K+, Ca++, and Mg++ are present in their salt forms inCWPS. Since the CWPS with 70% solution was more demineralized than the CWPS with40% solution, the presence of the mentioned ions affect the ammonium removal efficiencyof the zeolite less. The removal in the 70% demineralized cheese whey powder was 4%higher although the difference was statistically insignificant.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4

NUTRIENT REMOVAL FROM CWPS BY CONSTRUCTED WETLAND 125

Figure 6 EC changes (a) For 70% CWPS solution; (b) For 40% CWPS solution.

The average influent pH value for the CWPS with 70% solution was 6.71 ± 0.71 andfor the CWPS with 40% solution it was 6.57 ± 0.67. The effluent pH values for CWPS1,CWPS2, CWPS3, and CWPS4 were 7.27 ± 0.38, 7.36 ± 0.34, 7.24 ± 0.45, and 7.29 ±0.41, respectively. The standard deviation in the pH values were thought to be resultingfrom the quality differences during the preparation of the solutions using tap water.

Vertical flow reed-beds are usually effective for the nitrification of ammonia (Zuritaet al. 2009). Specifically, the decreasing pH value around the root area due to the presenceof the dissolved oxygen is an indicator of nitrification (Bezbaruh and Zhang 2004). Theinfluent pH values for both concentrations of CWPS are generally lower than the effluentvalues and following the 71st day, a decrease in the effluent pH values was observed. Thisphenomenon leads to the thinking of the presence of nitrification.

Nitrification of ammonia nitrogen can occur in the oxidized root zone of the Typhalatifolia and denitrification occurs into reduced environments in water columns or in thesediments (Bonomo et al. 1997). Besides these mechanisms, the presence of zeolite medium

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4

126 A. YALCUK

makes nitrogen removal mechanism in the studied systems more complex. It is suggestedthat ammonium volatilization plays a negligible role since pH was lower than 8–9 and alsothe average temperatures were below 25◦C (Figure 5).

The Changes in EC

The average inlet EC values in the CWPS with 70% solution were 636.15 ±53.67 µS/cm and at the outlet from the CWPS1, and the CWPS2 reactors, the valueswere 1465.36 ± 294.52 µS/cm, and 1472.36 ± 238.43 µS/cm, respectively and the aver-age inlet EC values in the CWPS with 40% solution were 623.52 ± 36.32 µS/cm and atthe outlet from the CWPS3, and the CWPS4 reactors, the values were 1541.10 ± 253.43µS/cm, and 1512.63 ± 242.81 µS/cm, respectively (Figure 6a–6b). The effluent EC val-ues for all the reactors were greater than that of the influent EC value. This leads to theidentification of an ion release into the environment in all systems. Mashauri et al. (2000)have identified lower influent EC values than the effluent EC values in their study on theartificial wetland systems for the treatment of domestic wastewaters. They have reasonedthis situation to stem from the vegetation deaths and the release of the nutritional elementsinto the water or due to the increase in the amount of the dissolved ions in the medium. Alsohigher EC values were obtained from the systems and this fact is probably to the higher ionconcentration in the effluent. The increase in the plant free control reactor was thought tobe caused by the desorption of the filling materials in time, releasing ions to the medium(Yalcuk et al. 2010).

CONCLUSION

The removal efficiencies of COD, PO4-P, and NH4-N were obtained as 37.47%,45.62%, and 68.88% in CWPS 1; 24.89%, 35.74%, and 63.15% in CWPS 2; 51.15%,54.96%, and 64.13% in CWPS 3; and 28.35%, 23.99%, and 65.92% in CWPS 4, respec-tively. The results that were obtained showed that vertical flow constructed wetland systemsdid not remove COD, PO4-P, or NH4-N from CWPS completely. However the best resultswere obtained for the 40% demineralized cheese whey powder solution for COD and PO4-P removal. This result also indicates that the selected plant was resistant to saline stress.NH4-N removal was more effective at CWPS with 70% solution. The zeolite in the fillingmaterial mixture has contributed to the removal of ammonia.

REFERENCES

American Public Health Association (APHA). 2005. Standard methods for the examination of wasteand wastewater. 21st ed. Washington (DC): APHA.

Astosan Dairy Products. 2003. Specification of 40% demineralized whey powder; specificationof 70% demineralized whey powder. Available from: www.astosan.com.tr/eng/40eng.htm,www.astosan.com.tr/eng/70eng.htm.

Bezbaruah AN, Zhang T. 2004. pH, redox, and oxygen microprofiles in rhizosphere of Bulrush in aconstructed wetland treating municipal wastewater. Biotechnol Bioeng. 88(1):60–70.

Bonomo L, Pastorelli G, Zambon N.1997. Advantages and limitations of duckweed-based wastewatertreatment systems. Water Sci Technol. 35: 236.

Brennan JG, Grandison AS, Lewis MJ. 2006. Seperations in food processing. In: Brennan JG, editor.Food Processing Handbook.Weinheim (Germany): Wiley-VCH. p. 500–501.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4

NUTRIENT REMOVAL FROM CWPS BY CONSTRUCTED WETLAND 127

Brewer AJ, Cumby TR, Dimmock SJ. 1999. Dirty water from dairy farms II: Treatment and disposaloptions. Bioresour Technol. 67: 161–169.

Conolly R, Zhao Y, Sun G, Allen S. 2003. Removal of ammoniacal-nitrogen from an artifical landfillleachate in downflow reed beds. Process Biochem. 38(7):1003–1018.

Farnet AM, Prudent P, Cigna M, Gros R. 2008. Soil microbial activities in a constructed soil reedbed under cheese dairy farm effluent. Bioresour Technol. 99: 6198–6206.

Farnet AM, Prudent P, Ziarelli PF, Domeizel M, Gros R. 2009. Solid state 13C NMR to assess organicmatter transformation in a subsurface wetland under cheese- dairy farm effluents. BioresourTechnol. 10: 4899–4902.

Gasuinas V, Strusevicius Z, Strusevicius MC. 2005. Pollutant removal by horizontal subsurface flowconstructed wetlands in Lithuania. J Environ Sci Health A. 40: 1467–1478.

ITRC. 2002. Tecnical and regulatory guidance document for constructed wetlands; [cited December2003]. Available from: www.itrcweb.org/Documents/WTLND-1.pdf.

Kantawanichkul S, Kladprasert S, Brix H. 2009. Treatment of high strenght wastewater in tropicalvertical flow constructed wetlands planted with Typha angustifolia and Cyperus involucratus.Ecol Eng. 35: 238–247.

Korkusuz A, Beklioglu M, Demirer G. 2005. Comparison of the treatment performances of blastfurnace slag- based and gravel- based vertical flow wetlands operated identically, for domesticwastewater treatment in Turkey. Ecol Eng. 24: 187–200.

Mashauri DA, Mulungu DMM, Abdulhussein BS. 2000. Constructed wetland at the University ofDar es Salaam. Water Res. 34(4):1135–1144.

Mayo AW, Mutamba J. 2004. Effect of HRT on nitrogen removal in a coupled HRP and unplantedsubsurface flow gravel bed constructed wetland. Phys Chem Earth. 29: 1253–1257.

McMillan C. 1959. Salt tolerance with in an Typha population. Am J Bot. 46: 521–52.Ozmıhcı S, Kargı F. 2008. Ethanol production from cheese whey powder solution in a packed column

bioreactor at different hydrolic residance times. Biochem Eng J. 42: 180–185.Pınar A, Ugurlu A. 2003. Ammonium removal from leachate by Turkish clinoptilolite: Effect of com-

peting ions. 9th International Waste Management and Landfill Symposium. Cagliari, Sardinia,Italy.

Sakadevan S, Bavor H. 1998. Phosphate adsorption characteristics of soils, slags and zeolite to beused as substrates in constructed wetland system. Water Res. 32: 393.

Sun G, Gray KR, Biddlestone AJ. 1998. Treatment of agricultural waste water in downflow reedbeds:Experimental trials and mathematical model. J Agric Eng Res. 69: 63–71.

Uribe BC, Miranda MIA, Costa ES, Pia, AB. 2006. Comparison of two nanofiltration membranesNF200 and Ds-5 DL to demineralize whey. Desalination. 199: 43–45.

Vymazol J. 2007. Removal of nutrients in various types of constructed wetlands. Sci Total Environ.380: 65–78.

Wood J, Fernandez JG, Barker A, Gregory J, Cumby T. 2007. Efficiency of reed beds in treatingdairy farms. Biosyst Eng. 98: 455–469.

Yalcuk A, Pakdil NB, Turan S. 2010. Performance evaluation on the treatment of olive mill wastewater in vertical subsurface flow constructed wetlands. Desalination. 262(1-3):209–214.

Yalcuk A, Ugurlu A. 2009. Comparison of horizontal and vertical constructed wetland systems forlandfill leachate treatment. Bioresour Technol. 100(9):2521–2526.

Zurita F, Anda DJ, Belmont MA. 2009. Treatment of domestic wastewater and production of com-mercial flowers in vertical and horizontal subsurface flow constructed wetlands. Ecol Eng. 35:861–869.

Dow

nloa

ded

by [

Uni

vers

ity o

f C

onne

ctic

ut]

at 0

7:49

11

Oct

ober

201

4