Cemile S?eyma Arzum Yap?c?, Dilan Toprak, Müjgan Y?ld?z, Sinan Uyan?k,2, Yakup Karaaslan,Deniz U?ar,2,*

1 Environmental Engineering Department, Engineering Faculty, Harran University, Sanliurfa 63100, Turkey

2 Gap Renewable Energy and Energy Efficiency Center, Harran University, Sanliurfa 63100, Turkey

3 General Directorate of Water Management, Republic of Turkey Ministry of Agriculture and Forestry Bes?tepe Mahallesi, Yenimahalle 06560, Turkey

Keywords:Nitrate removal Bioreactor Bioprocess Anaerobic process Groundwater treatment

ABSTRACT In this study, a sequential process (heterotrophic up-flow column and completely mixed membrane bioreactors) was proposed combining advantages of the both processes.The system was operated for 249 days with simulated and real groundwater for nitrate removal at concentrations varying from 25 to 145 mg·L-1 NO3--N.The contribution of heterotrophic process to total nitrate removal in the system was controlled by dozing the ethanol considering the nitrate concentration.By this way, sulfur based autotrophic denitrification rate was decreased and the effluent sulfate concentrations were controlled.The alkalinity requirement in the autotrophic process was produced in the heterotrophic reactor, and the system was operated without alkalinity supplementation.Throughout the study,the chemical oxygen demand in the heterotrophic reactor effluent was (23.7 ± 22) mg·L-1 and it was further decreased to(7.5±7.2)mg·L-1 in the system effluent,corresponding to a 70%reduction.In the last period of the study,the real groundwater containing 145 mg·L-1 NO3--N was completely removed.Membrane was operated without chemical washing in the first 114 days.Between days 115-249 weekly chemical washing was required.

1.Introduction

Excessive use of nitrogenous fertilizers and the discharge of domestic and industrial wastewater without proper treatment have led to an increase in the nitrate concentration of surface and groundwater supplies.In many countries, the concentration of nitrate in groundwater has exceeded the maximum permissible limits for human consumption[1].In a study carried out in Harran Plain in Turkey,180 mg·L-1-N was detected in some wells and the average in the whole plain was reported as 35 mg·L-1-N[2].In a study conducted in 2017 in the same region, the highest value was reported as (83.2 ± 5.4) mg·L-1-N [3].

Excess nitrate in drinking water can cause methemoglobinemia(‘‘Blue Baby Disease”) in infants [4,5], which is characterized by above-normal levels of methemoglobin in blood streaming [6].Additionally, nitrate causes diseases such as gastriculer, slow working of the thyroid gland, and cancer due to the formation of nitrosamines [4].The maximum nitrate and nitrite concentrations in drinking water determined by the United States Environmental Protection Agency (EPA), the European Economic Community, and the World Health Organization (WHO) are 10 mg·L-1-N and 1.0 mg·L-1-N, respectively [7].

In recent years, biological denitrification has gained popularity as a promising method for nitrate removal from drinking water[1,8].The biological reduction methods have several advantages,such as elimination of chemical demand, low cost, and nonexistence of secondary pollution such as concentrate in membrane applications.Autotrophic (Reaction (1)) and heterotrophic (Reaction (2)) processes have been conducted both simultaneously[9,10]and separately [8,11]to ensure high nitrate removal rates.

Elemental sulfur is non-toxic,low water solubility,stable under normal conditions [12].However, in the sulfur-based denitrification process, 7.56 mg ofis produced while consuming 4.57 mg of CaCO3alkalinity per mg reduced-N (Reaction(1)).

To eliminate these disadvantages, sulfur-based denitrification can be combined with heterotrophic process with the addition of organic carbon.In this way, heterotrophically produced alkalinity can provide the necessary alkalinity in autotrophic process to balance the pH, and the heterotrophic reduction of a part of nitrate could control the sulfate in the system effluent.For this purpose,mixotrophic reactors were used in recent studies, to reduce high nitrate concentrations(75 mg·L-1N)without alkalinity addition [1].There are many reported mixotrophic denitrification processes in single column reactors [7,9,10].

Membrane bioreactors have been actively employed for domestic and industrial wastewater treatment.In recent years, the combination of membrane processes and biological denitrification has received great interest [8].The incorporation of the sulfur-based denitrification process with MBRs allows the use of powdered sulfur particles.In all these processes, both autotrophic microorganisms and heterotrophic microorganisms are working in the same environment.

As an autotrophic process, MBR process has the advantages of keeping both biomass and sulfur particles in the bioreactor.Hence,powdered sulfur can be used instead of granules, which can improve the process efficiency due to increased surface area.Therefore, a combo process consisting of a heterotrophic column reactor and autotrophic MBR would be interesting as the process includes advantages of both processes together with membrane process advantages.Also, autotrophic membrane bioreactor can remove the un-oxidized organic matter remaining in the effluent of heterotrophic reactor effluent.

Apart from electron donor type, plug flow column reactors and completely mixed reactors have different μmaxand KSPvalues due to the difference in nitrate concentrations interacted by microorganisms.Membrane bioreactors, on the other hand, provide high effluent water quality due to membrane filtration.In this study, a column reactor that may have high μmaxand KSPvalues and a membrane bioreactor were sequentially operated to upgrade current sequential systems.To the best of the authors knowledge,this is the first study in a column and a membrane bioreactor was operated sequentially.In this way,system effluent was provided from a membrane and a plug flow reactor provided the high nitrate removal.In this context, denitrification performance of sequential heterotrophic column and autotrophic MBR was assessed in this study for the treatment of synthetic and real groundwater contaminated with high concentration of nitrate.The system’s performance was investigated at varying hydraulic retention times(HRTs) and nitrate loadings.

2.Materials and Method

2.1.Reactors

The heterotrophic and autotrophic sequential system (Fig.1)consisted of an up-flow column reactor and a completely mixed membrane bioreactor (MBR).Sand particles (1-2 mm) were used as column filling material in the heterotrophic column reactor and the bed volume was 500 ml.The autotrophic reactor dimensions were 15 cm × 15 cm × 50 cm, corresponding to total and active volumes of the reactor were 12.5 L and 9 L, respectively.

Reactors were covered with aluminum foil to prevent phototrophic growth.A magnetic stirrer was used to provide complete mixing conditions in the autotrophic reactor.A 9 cm × 4.3 cm double-sided microfiltration membrane module was used for solid,liquid separation in the reactor effluent and the active membrane surface was 0.0077 m2.The membrane used was a Polyethersulfone (PES) microfiltration membrane with a pore size of 0.22 μm.Volatile suspended solid(VSS)in the autotrophic reactor was kept at around 4000 mg·L-1VSS and the sludge retention time was limitless.

Fig.1.The heterotrophic-autotrophic sequential system.

2.2.Operational conditions

The reactors were inoculated with sludge taken from another mixotrophic denitrifying reactor.Synthetically prepared groundwater containing nitrate at concentrations ranging from 25 to 50 mg·L-1-N was fed to the reactors during the first 5 periods.Two different HRT and nitrate loading rates were applied to determine their effect on membrane fouling and nitrate removal rates.In the last period,the reactors were fed with real groundwater containing 145 ± 5 mg·L-1-N.The real groundwater was taken from Ug?urlu village in the Harran Plain, Turkey.The Harran Plain groundwater has high levels of nitrate, and there are several studies in the literature[2].The operational conditions are presented in Table 1.

During the first 5 periods,synthetic groundwater was prepared by adding nitrate and phosphate to the tap water.To ensure anoxic conditions in the reactor, the feed was aerated with N2gas for 10 minutes and kept at 4 °C.The heterotrophic reactor was fed with a peristaltic pump and the heterotrophic reactor effluent was collected in a container where it fed to the autotrophic reactor.Ethanol was used as carbon and electron sources.The ethanol concentration was adjusted to remove a portion of total nitrate(Table 1).According to Reaction (2), 1.77 mg ethanol is required to reduce each mg-N.The autotrophic reactor was fed with elemental sulfur and the reactor effluent was filtered by a 0.22 μm microfiltration membrane.Elemental sulfur was added weekly to the reactor.During the study,-N,-N,,COD,pH and alkalinity analysis were carried out from the influent and the effluent regularly.Also, flow rate and trans-membrane pressures were monitored on a daily basis during the operation of the autotrophic membrane bioreactor.The gas produced by the reactors was also measured and compared with the theoretical value calculated based on influent and effluent nitrate and nitrate concentrations.

2.3.Theory and calculation

Mixotrophic reactors can eliminate the drawbacks of both autotrophic and heterotrophic reactors.Both up flow and completely mixed membrane bioreactors were used for mixotrophic denitrification.The microorganisms in the up flow reactors inlet, may be exposed to high concentrations of nitrate, hence may have higher μmaxand KSPvalues, providing higher treatment efficiencies than completely mixed reactors.The membrane bioreactors, on the other hand,may provide better effluent quality than up flow reactors.Therefore the theory in this study is providing the higher treatment efficiency with the up flow heterotrophic and completely mixed autotrophic membrane bioreactor configuration.The calculations and mass balances were performed according to Reactions(1)and(2).Removal efficiencies were calculated according to the following Equation:

Here, Ciand Ceare the concentrations in the feed and effluent,respectively.

2.4.Analytical methods

3.Results

3.1.Nitrate reduction

3.1.1.Heterotrophic reactor performance

The influent nitrate concentration during the first three periods was 25 mg·L-1-N and ethanol was supplemented to feed stoichiometrically required to reduce half amount of the influent nitrate(12.5 mg·L-1-N)(Table 1).In accordance with the ethanol concentration in the feed, the average effluent nitrate concentration during the first two periods was(11.2±7.2)mg·L-1-N,corresponding to a 55% reduction.This value was theoretically compatible with the feed COD concentration.Although the HRT decrease to 8.6 h in the 3rd period caused a temporary negative effect on nitrate removal, the average nitrate concentration at the end of the period was 12.5 ± 3.83.At period 4, nitrate was increased to 50 mg·L-1-N, however the COD concentration in the feed was kept at (46 ± 4.7) mg·L-1(equal to reduce 12.5 mg·L-1-N).The nitrate removal performance of heterotrophic reactor under these conditions reduced probably due to famine conditions and the effluent average nitrate concentration was (44.9 ± 5.6) mg·L-1-N during this period (Fig.2).To increase the rate of heterotrophic nitrate removal in period 4, the influent COD was increased to(138±8.7)mg·L-1in the 5th period.At this period the effluent nitrate concentration was reduced to 25 mg·L-1-N.

Table 1 The operational conditions

Fig.2.Nitrate and COD variations throughout the study.

In the last period, feed was contained (145 ± 5) mg·L-1-N and COD concentration also increased to a level that is stoichiometrically enough to reduce 130 mg·L-1-N.However,as seen in Fig.2, average reduced nitrate concentration in this period was 34 mg·L-1in the heterotrophic reactor corresponding to 24%nitrate removal.This situation was probably due to high nitrate concentration and low HRT.

3.1.2.Autotrophic reactor performance

The partially denitrified water was treated by sulfur-based denitrification in an autotrophic reactor.During the entire study, the average effluent nitrate concentration was below 2 mg·L-1as shown in Fig.2.In the first 3 periods,the autotrophic reactor influent and effluent nitrate was (12.2 ± 7) and (1.87 ± 2) mg·L-1N,respectively.However,the reduction of the HRT did not result in a performance decrease.

In the period 4,the heterotrophic reactor was operated in ethanol limiting conditions.While influent nitrate concentration was increased to (145 ± 5) mg·L-1, ethanol concentration in the feed was remained at(46±4.7)mg·L-1.The autotrophic effluent nitrate during this period was (2.2 ± 0.86) mg·L-1-N.Despite the increased influent nitrate concentration, the effluent nitrate concentrations of the autotrophic membrane bioreactor were not affected by the influent nitrate.During this period, HRT was 6.4 days in the autotrophic reactor.It has been reported in the literature that a higher nitrate removal rate can be obtained with lower HRTs.In our previous study, up-flow anaerobic denitrifying column reactors were operated for simultaneous nitrate and perchlorate reduction and the denitrification ratio was 0.3 g·L-1·d-1while HRT was 2 hours in elemental sulfur based autotrophic reactors.In the same study, methanol based denitrification rate was 0.6 g·L-1·d-1while HRT was 1 hour [11].In another study, upflow anaerobic column reactors were operated in serial mode.Effluent of heterotrophic reactor was pumped to the autotrophic reactor and 2.4 g·L-1·d-1and 0.86 g·L-1·d-1denitrification ratios were obtained in heterotrophic and autotrophic reactors, respectively[11].Sahinkaya et al.used a sulfur-based autotrophic membrane bioreactor, and the denitrification ratio was 0.24 g·L-1·d-1with an HRT of 5 hours.Despite reported high nitrate reduction rates, high effluent quality is desired to encourage public acceptance for drinking water treatment.Membrane filtration can be applied to provide high effluent quality, as it rejects not only microorganisms but also microbial products and most suspended solids.

The Influent nitrate concentration was(29.9±4.9)mg·L-1-N in the period 5 and this nitrate was completely reduced in this period.Despite the high COD concentration in the effluent of heterotrophic reactor in the last period, the nitrate removal rate remained at 25% and therefore (107 ± 20) mg·L-1-N was entered into the autotrophic reactor together with(87 ± 18) mg·L-1COD.However, complete denitrification was achieved in the autotrophic reactor.Based on influent and effluent COD concentrations, it could be noted that autotrophic reactor in this period worked as a mixotrophic reactor.Nitrite was only observed in the heterotrophic reactor effluent and its concentration was(3.81±3.52)mg·L-1-N and autotrophic reactor effluent nitrite was below the detection limit except few measurements.

3.2.Alkalinity

3.2.1.Heterotrophic reactor performance

According to Reaction (2), 3.57 mg CaCO3is produced in the ethanol-based denitrification of each gram-N.During the first three periods, 12.7 mg·L-1-N was removed in the heterotrophic reactor, which theoretically produces 45.6 mg·L-1CaCO3.The influent average alkalinity was (133 ± 17) mg·L-1CaCO3and theoretically, this value was expected to reach 178 mg·L-1CaCO3.Measured average alkalinity concentration was (182 ± 37) mg·L-1which was in agreement with the theoretically calculated level(Fig.3).Due to low ethanol fed in the 4th period,alkalinity production remained low in this period.In this period,the effluent nitrate concentration of the heterotrophic reactor was(44.9±5.6)mg·L-1NO3--N, and the alkalinity concentrations were in accordance with the reduced nitrate.In the 5th period, the amount of heterotrophically reduced nitrate was increased and average measured alkalinity in this period was increased to(210 ±19) mg·L-1CaCO3which was in good agreement with the theoretical value (200 mg·L-1CaCO3).In the last period, the real underground water was fed to the system; however, only 24% of influent nitrate was reduced in the heterotrophic reactor (from 141 to 107 mg·L-1-N).Theoretically calculated and measured alkalinity concentrations in the last period were 290 and 284 mg·L-1CaCO3, respectively.

Fig.3.Sulfate and alkalinity variations throughout the study.

3.2.2.Autotrophic reactor performance

Elemental sulfur based autotrophic denitrification process is an alkalinity consuming reaction which consumes 4.57 mg CaCO3for each mg reduced-N(Reaction(1)).During the first three periods, the autotrophically reduced nitrate concentration was(13 ± 4) mg·L-1NO3--N, and therefore, the theoretical alkalinity requirement was approximately 60 mg·L-1CaCO3.The autotrophic reactor average influent and effluent alkalinity concentrations in these periods were(181±29)and(85±42)mg·L-1CaCO3,respectively.The autotrophic reduced-N was(42±6)mg·L-1and the alkalinity consumption in the autotrophic reactor was(107 ± 36) mg·L-1CaCO3(Fig.3).Alkalinity consumption in the period 5 was(72.5±23)mg·L-1CaCO3,which was 67%lower than the previous period.This decrease in alkalinity consumption can be explained by the increased COD concentration in the 5th period.During this period, COD was increased from (46 ± 4.7) mg·L-1to(138 ± 8.7) mg·L-1(Table 1).Hence, the amount of heterotrophically reduced nitrate is also increased.In the last period, autotrophic reactor influent average alkalinity concentration was(285 ± 59) g·L-1, and it was decreased to (74 ± 43) mg·L-1at the effluent.The decrease in alkalinity during this period indicates that 46 mg·L-1-N should be autotrophically reduced.However, it appeared that in the last period much more nitrate (100 mg·L-1-N) was removed in the autotrophic reactor.It is thus concluded that the ethanol-based denitrification process might have continued in the autotrophic reactor.Throughout the study, influent, heterotrophic effluent and autotrophic effluent pHs were 8.28 ± 0.37, 8.64 ± 0.47 and 7.92 ± 0.76, respectively.

3.3.Sulfate production in the autotrophic reactor

According to reaction(1),7.56 mgis produced for each mg-N reduced.The measured and theoretical sulfate concentrations are shown in Fig.3.For the first three periods, the effluent average sulfate concentration was (206 ± 41) mg·L-1while the theoretical value for these periods was 150 mg·L-1.In the following period, the effluent sulfate increased to(372 ± 42) g·L-1with an increase in the autotrophically removed nitrate.Increasing the influent COD concentration in the 5th period decreased the amount of autotrophically reduced nitrate and the effluent sulfate concentrations decreased to(215±99)mg·L-1.In the last period,up to 800 mg·L-1sulfate was produced due to high nitrate reduction in the autotrophic reactor.There might be a number of reasons for higher sulfate concentrations than theoretical values.Oxygen leaking into the MBR during feeding or the sampling-operation may have caused the elemental sulfur to react with oxygen.Another cause may have been elemental sulfur disproportionation[15].Average total sulfide concentrations in heterotrophic and autotrophic reactors throughout the study were ＜1 mg·L-1.

3.4.Chemical oxygen demand

In the system, the influent COD was adjusted to reduce a portion of the influent nitrate concentration.Heterotrophic reactors provide much higher denitrification rates than autotrophic reactors.However, the disadvantage of these systems is the organic matter which can be remained in the effluent without being used in the reactor.In the sequential system, this COD can be retained in the autotrophic reactor.During the first 4 periods, 50 mg·L-1;in the 5th period,138 mg·L-1COD were added to feed.The average effluent COD was (17.5 ± 10) mg·L-1for the first 4 periods and(23.8±11)mg·L-1for the 5th period.In the last period,the influent-N was increased from 50 mg·L-1to (145 ± 5) mg·L-1and the COD was also increased to 490 mg·L-1.During this period, an increase in the effluent of heterotrophic reactor, COD concentration was observed with an average value of (87 ± 18) mg·L-1.

3.5.Membrane filtration in the autotrophic reactor

Polyethersulfone membrane (PES) was used throughout the study.Initially, the flux was kept at 2.9 L·m-2·h-1for 114 days without any chemical washing.In this period, membranes were cleaned by a sponge once in every 10-12 days (Fig.4a).When the flux was increased to 8 L·m-2·h-1on day 115, more frequent(onceevery 2 days) and intensive washing was required.In this period, chemical washing was also applied.The chemical washing was carried out by soaking the membranes 1 hour in 1%NaClO and then 1 hour in acid solution(pH 2 with H2SO4).Irrecoverable clogging was observed when the flux was increased to 8 L·m-2·h-1.Despite chemical washing, initial TMP after washing gradually reaches 0.5 MPa.This initial pressure was due to the blockage resistance due to pore blockage.The changes in the clogging rate are given in Fig.4b.Throughout the entire operation, 2 different fouling rate trends were experienced.In the first 114 days,the rate of fouling was(0.27±0.0097)MPa·d-1.In the following days,when the flux was increased to 8 L·m-2·h-1, the rate of fouling was increased to (1.45 ± 0.3950) MPa·d-1.Depending on reactor operation and the cleaning type in the previous fouling period,approximately 2.50 MPa·d-1fouling rates were also observed.

Fig.4.Trans membrane pressures and fouling rates of the autotrophic reactor (a:TMP values, b: Fouling rates).

4.Discussion

A biological method for nitrate removal from underground waters is preferred because of their advantages such as low costs,not require expensive catalysts (electrochemical reduction) and not produce secondary pollution (i.e.brine).Single heterotrophic or autotrophic denitrification processes have been used successfully in a series of studies for nitrate removal[10,16,17].However,both processes have some drawbacks such as alkalinity and sulfate production in the sulfur based autotrophic process and risk of organic contamination in the heterotrophic process.The mixotrophic processes established by the combination of auto and heterotrophic processes may help to eliminate the disadvantages of both processes [18].One of these advantages of the mixotrophic process is the elimination of the alkalinity needs of the systems as sulfur based autotrophic denitrification requires 4.57 mg CaCO3for each mg-N reduced.Such a need in the autotrophic process can be produced in the heterotrophic process.The influent average alkalinity throughout the study was (132 ± 30) mg·L-1CaCO3,which was sufficient to reduce theoretically 28 mg·L-1-N in the autotrophic process.Although higher than 28 mg·L-1-N was reduced during the study(50 mg·L-1-N in the 4th period and 145 mg·L-1-N in the 6th period), no pH decrease was observed due to alkalinity production in the heterotrophic reactor.

Designing as a continuous heterotrophic-autotrophic sequential system was provided an advantage at this point.Nitrate was firstly reduced in the heterotrophic process and alkalinity was produced at a rate of 3.57 mg CaCO3·(mg NO3-N)-1is a practical approach,56% of influent nitrate should be reduced in heterotrophic and 44% in autotrophic process to balance the alkalinity if sulfate is not concern.

Another advantage of the system is that the unoxidized COD in the effluent of the heterotrophic reactor could be retained in the autotrophic reactor.In the last period of the study, influent(490±20)mg·L-1COD was not oxidized completely in the heterotrophic reactor and residual COD was removed in the subsequent autotrophic reactor.In this period, 491 mg·L-1influent COD was reduced to (87 ± 18) and (2.9 ± 1.2) mg·L-1, in the effluent of heterotrophic and autotrophic reactors, respectively.

In the literature,there are examples of heterotrophic reactors in which all nitrate is reduced by organic electron sources.However,there is a risk of effluent contamination with organic residues.For example, in a column reactor operated with methanol, 25 mg·L-1-N was treated at different loading rates and an average of 2.4 mg·L-1DOC was measured at the effluent [11].In another study, average effluent DOC concentration was decreased to around 1 mg·L-1in a mixotrophic reactor where methanol was supplemented with elemental sulfur[9].Organic matter remaining in the effluent can cause microbial growth in water distribution networks.In addition, the organic material used should not threaten public health and be acceptable to society while providing the expected reduction efficiency.While methanol has toxic effects,ethanol can create extra public resistance for some societies.In heterotrophic autotrophic sequential systems,un-oxidized organic electron donors in heterotrophic reactors can be oxidized in the following autotrophic reactor.Hence, effluent organic concentration generally much lower than that in only heterotrophic processes [18].

Autotrophic denitrification using inorganic electron sources is a unique process and which can be used as single processes or in combination with organic supplementation as mixotrophic denitrification.In this way,utilization of inorganic compounds reduces the risk of organic contamination in the effluent, while reducing the risk of disinfection by-products such as trihalomethane in disinfection with chlorine.Common electron sources used in autotrophic denitrification include hydrogen gas and reduced sulfur compounds (elemental sulfur, sulfur and thiosulfate).In addition,Zero-valent iron (Fe0), Ferrous iron, Arsenite (As3+), Manganese(Mn2+) are among the other electron sources reported [19].Electron sources have a serious effect on reaction rates.It is reported that H2and reduced inorganic sulfur compounds(elemental sulfur(S0), sulfide (S2-) and thiosulfate (S2), have a high denitrification rates while lower rates were reported with other inorganic compounds such as arsenite [20-23].In the selection of electron donor,parameters such as substrate bioavailability of microorganisms, microbial affinity, energy efficiency, reactor type, reaction rate, efficiency should be evaluated.According to all these parameters, inorganic electron sources were reviewed in the excellent studies [19].

Sulfur-based autotrophic systems have advantages such as lower cost of elemental sulfur and the on-demand release in the aquatic environment.However, sulfate production is another drawback of the process.Especially,in some cases,nitrate and sulfate may be found together in the groundwater.For example, in a study conducted in 2015 at Harran Plain, up to (425.7 ± 36.1) mg·L-1sulfate was measured in some groundwater samples and the average sulfate for the whole plain was 82 mg·L-1[3].The presence of nitrate together with high concentrations of sulfate reduces the possibility of using sulfur-based autotrophic reactor operations as for drinking purposes, sulfate should be less than 250 mg·L-1.In such sensitive systems, reduced nitrate amount should be shifted towards the heterotrophic reactor by adding organic carbon to the influent.In the elemental sulfur based denitrification, it is inevitable to observe higher than theoretical levels of sulfate.For example, despite all the precautions taken, dissolved oxygen may remain in the feed.In addition, an increase in sulfate concentrations can be observed as a result of elemental sulfur disproportionation.In the first three periods, while the autotrophically reduced nitrate was(12.2±8)mg·L-1-N,mean effluent sulfate concentration (produced + influent sulfate) was (182 ± 55) mg·L-1,which was 38 mg·L-1above the theoretical level.Similar studies have also reported high sulfate concentrations up to 75 mg·L-1above the theoretical sulfate mainly because of oxygen leakage into the system [1].Therefore, a safety factor should be applied when setting the amount of nitrate to be removed in the autotrophic reactor.

5.Conclusions

The effluent sulfate values can be controlled by heterotrophic autotrophic sequential denitrification system in the treatment of high nitrate concentration waters.In the study, 50 mg·L-1NO3--N was completely removed while controlling the sulfate by adjusting the influent ethanol concentration.In the sequential system, alkalinity produced in the first heterotrophic process was used in the subsequent autotrophic process to eliminate the need for external alkalinity.Real groundwater samples having 145 mg·L-1N was completely removed,and COD and-N in the heterotrophic reactor effluent were removed in the autotrophic reactor when high performance was expected from the system.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was supported by Harran University Scientific Research Projects Coordination Unit (HUBAP, project no 18018).

Credit Authorship Contribution Statement

Cemile S?eyma ARZUM YAPICI:Conceptualization, Investigation,Methodology.Dilan TOPRAK:Data curation.Müjgan YILDIZ:Investigation,Software.Sinan UYANIK:Visualization,Writing - Review & Editing, Resources.Yakup KARAASLAN:Writing - Review & Editing.Conceptualization,Methodology, Funding acquisition, Writing - Original draft.