Simulation of denitrification in groundwater from Chaohu Lake Catchment,China

Ji-zhong Qin*,Xiu-xun WngLei MLong-ping Wng,Jin-kui Liu,Zhng-xin Yng

aSchool of Resources and Environmental Engineering,Hefei University of Technology,Hefei 230009,China

bAnhui Institute of Geo-Environment Monitoring,Hefei 230009,China

Abstract The eutrophication of Chaohu Lake in China is mainly attributed to nitrate in flow from non-point sources in the lake catchment.In this study,biological nitrate reduction from groundwater in the Chaohu Lake Catchment was investigated under laboratory conditions in a continuous upflow reactor.Sodium acetate served as the carbon source and electron donor.Results showed that a carbon-to-nitrogen(C/N)molar ratio of 3:1 and hydraulic retention time(HRT)of 8 d could achieve the most rapid nitrate nitrogen()depletion(from 100 mg/L to 1 mg/L within 120 h).This rate was confirmed when field groundwater was tested in the reactor,in which a removal rate of 97.71%was achieved(from 60.35 mg/L to 1.38 mg/L within 120 h).Different levels of the initial concentration(30,50,70,and 100 mg/L)showed observable in fluence on the denitrification rates,with an overall average removal efficiency of 98.25%at 120 h.Nitrite nitrogen()accumulated in the initial 12 h,and then kept decreasing,until it reached 0.0254 mg/L at 120 h.Compared with the initial value,there was a slight accumulation of 0.04 mg/L for the ammonia nitrogen()concentration in the ef fluent,which is,however,less than the limit value.These results can provide a reference for evaluating performance of denitrification in situ.

Keywords:Chaohu Lake Catchment;Denitrification;Eutrophication;Nitrate;Wetland

1.Introduction

Nitrate contamination in groundwater has become an environmental problem across the world(Qian et al.,2015;He et al.,2016).High nitrate concentrations in drinking water can cause problems such as infantile methemoglobinemia(Haugen et al.,2002)and gastric cancer(Moreno et al.,2005).In addition,the main factor causing the eutrophication of water bodies is nitrate flowing into lakes through groundwater.Therefore,remediation of nitrate-contaminated groundwater has caught the attention of scientists all over the world.

Remediation technologiesincludetheion exchange method,the reverse osmosis process,the electrodialysis process,and biological denitrification.Of these technologies,biological denitrification is a simple,highly selective,and economical method,which has been proven to be one of the most cost-effective processes in removing nitrate nitrogen by transforming it into inert nitrogen gas(Ovez et al.,2006).Biological denitrification,which is ubiquitous in nature,is an anaerobic process requiring carbon sources to serve as electron donors(Rocca et al.,2007).Previous studies have shown that denitrification below the root zone is often limited by the low availability of carbon(C)or other electron donors.Liquid carbon substrates(i.e.,methanol,molasses,sodium acetate,or glucose)were selected as electron donors for heterotrophic denitrificans.Woodchips and sawdust are the most frequently employed solid carbon substrates in field-scale projects(Schmidt and Clark,2012).

However,denitrification presents complex spatio-temporal patterns and is controlled by many factors(Bernard-Jannin et al.,2016).Green et al.(2016)found that the rates of oxygen and nitrate reduction are key factors in determination of the chemical evolution of groundwater.Yuan et al.(2013)reported thatcarbon-to-nitrogen (C/N)molarratiosof 2.5-4.0 did not affect the nitrate reduction rate,but signi ficantly affected the nitrite accumulation rate.C/N molar ratios of 2.5 and 3.0 were favorable for nitrite accumulation,and a C/N molar ratio greater than 3.5 would lead to a decrease in nitrite accumulation in a sequencing batch reactor.Capua et al.(2017)reported that temperature had no significant effects on the denitrification efficiency between 3°C and 20°C.Removal of nitrate is in fluenced by the hydraulic retention time(HRT).Liu et al.(2017)showed that the removal rate of nitrate was higher than 70%,when the HRT decreased from 8 h to 6 h.Studies of amending external electron donors to enhance denitrification have been attempted in both laboratories and field.However,there are some problems in simulation or in situ research,such as selection and use of the carbon source.Soluble carbon is convenient,but secondary pollution may be caused by overuse.Insoluble carbon,wheat straw(Aslan and Cakici,2007),and sawdust(Shen et al.,2011)have been selected as carbon sources,and may cause the ef fluent from the reactor to be colored and odorous in spite of their high efficiency in removing nitrate.Of these carbon sources,sodium acetate has performed best in sustaining rapid nitrate reduction in water(Bilanovic et al.,1999;Lee et al.,2001).

Chaohu Lake,located in Hefei City,is one of the five largest freshwater lakes in China.Eutrophication is the main problem currently faced by the lake(Fu et al.,2006;Qian et al.,2007;Wei et al.,2008).In our previous surveys of well water around Chaohu Lake,the nitrate concentration in the groundwater was 24.80-60.35 mg/L.Nitrate in riparian areas can be efficiently removed,mainly through denitrification from shallow groundwater flowing from uplands towards rivers(Bouwman et al.,2013).Nitrogen-enriched groundwater has recently been proposed as an important direct contribution to the eutrophication of Chaohu Lake.Removing nitrate from drinking water can both diminish health risks to the human population and prevent nitrogen from entering Chaohu Lake through groundwater.

Inordertoreducethecontributionofnitratetothenitrogenof groundwater in thelake catchment,we simulated the removalof nitratefromgroundwaterinthe aquifer,anddevelopedaninsitu technology for the treatment of nitrate-contaminated groundwater.The denitrifying process was investigated with groundwater samples collected from the lakeshore area of Chaohu Lake.Acontinuousup- flowmodesandyreactorwasestablished and evaluated for nitrate removal in the laboratory.Biodenitrification at different HRTs,C/N molar ratios,and initial concentrations of nitrate with sodium acetate as a carbon source for this system were assessed.The results can provide a theoretical basis for in situ application to the treatment of nitrate pollution in the Chaohu Lake Catchment.

2.Materials and methods

Groundwater and soil samples were collected from the Nanfei River area within the Chaohu Lake Catchment.The geographical location of the study area is shown in Fig.1.The Nanfei River,which flows through Hefei City(a city with a population of 3 million),is the main source of recharge to Chaohu Lake.Groundwater in this area has been contaminated by nitrate from agricultural and residential activities(Qian et al.,2007).The baseline quality of the samples was analyzed,and the results are summarized in Table 1.

Fig.1.Geographical location of study area.

Table 1 Denitrification in field groundwater.

Sediment samples were collected from a natural wetland near Chaohu Lake,at a depth of 1.5 m below the ground surface.They were used as the inoculum for denitrifying bacteria in the laboratory reactors.

10 g of sediment sample was added to 500 mL of liquid medium(20 g of C4H4KNaO6·4H2O,2 g of KNO3,0.5 g of K2HPO4,0.5 g of KH2PO4,and 0.2 g of MgSO4·7H2O,dissolved in 1000 mL of deionized water with a pH value of 7.2),which was used for denitrifying bacteria.After 48 h of enrichment at 30°C under anaerobic conditions,the medium was brie fly shaken and the aqueous phase was used as the inoculum for the reactor,which was filled with the same liquid medium(sterile)and allowed to stand for 2-3 d before it received any in fluent.

The reactor was made of a polymethyl methacrylate cylindrical column(with a height of 45 cm and an inner diameter of 11 cm) filled with pre-washed sand(with an average diameter ranging from 0.5 to 1.0 mm)to a height of 40 cm(Fig.2).The column's ef fluent port was at a height of 40 cm,and the top of the column was connected with a vent pipe equipped with an automatic cutout valve.A peristaltic pump was used to regulate the in fluent flow rates.Experiments were conducted under room temperatures ranging from 22°C to 25°C.

Fig.2.Schematic design of reactor.

Sets of various HRTs and C/N molar ratios were tested in the reactor to evaluate their in fluence on denitrification.The reactor was completely rinsed with deionized water after each set of experiments.Initial in fluent was groundwater containingwith concentrations from 30 to 100 mg/L(spiked if needed for different concentrations)and sodium acetate at C/N molar ratios of 1-7.In fluent was pumped into the reactor at flow rates of 1.2,2.4,and 4.8 mL/min,achieving HRTs of 8 d,4 d,and 2 d,respectively.The control reactor was not amended with sodium acetate.

Experiments were performed for 120 h for each set of samples,and the beginning of experiments was marked as 0 h.Chemical oxygen demand(COD),nitrate nitrogen(),nitrite nitrogen(),and ammonia nitrogen()in ef fluent were analyzed.Water samples were filtered through 0.45-μm membrane filters before analysis.andwere detected with ion chromatography(IC,Dionex DX-100,Sunnyvale,California,USA).TheconcentrationwasmeasuredwiththeNessler method.The COD concentration was measured with the potassium dichromate titration method(Wei,2003).Total organic carbon(TOC)was measured with the combustion oxidation-non-dispersive infrared absorption method.The pH was monitored throughout the study.

3.Results and discussion

The baseline groundwater quality is summarized in Table 1.Theconcentration was found to be 60.35 mg/L and theconcentration was found to be 0.167 mg/L at the beginning.Nowas detected(＜0.0030 mg/L).The concentration of TOC in the groundwater varied seasonally and the measured COD levels ranged from 4 to 22 mg/L.The pH value was found to be 7.2.

The resultsdescribe the variations of the,andconcentrations with different HRTs,as shown in Fig.3.At an initialconcentration of 30 mg/L and a C/N molar ratio of 3 with acetate(510 mg/L)as the substrate,was almost completely removed when the HRT was 2 d or longer(Fig.3(a)).The in fluence of HRT onremoval rates during the test was more significant at the early stage,andcould be removed faster with a longer HRT.For example,80%of(from 30 to 6 mg/L)was removed within 24 h at 8 d of HRT.The removal rate was 57%(from 30 to 13 mg/L)at the same time when the HRT was 2 d.This difference diminished in samples collected over 48-120 h(Fig.3(a)).It appeared that lower flow rates(with longer HRTs)might be favorable toremoval in the reactor,especially during the first 24 h.

Fig.3.Effects of HRT on nitrate removal.

Experiments were performed for 120 h to determine the effect of the C/N molar ratio on denitrification,with C/N ratios set at 1,3,5,and 7.The control group contained no substrate.The pro files ofconcentrations are summarized in Fig.4(a).In the control group,approximately 18.73%removal ofoccurred over 120 h.When the C/N molar ratio was 1,theremoval rate was 72.84%,which was signi ficantly higher than that in the control group but significantly lower than those with the C/N molar ratio of 3 or higher(98.25%,97.86%,and 97.86%for the C/N molar ratios of 3,5,and 7,respectively).No significant difference inremoval rates was observed when the C/N molar ratio increased to above 3(Fig.4(a)).Trends ofandaccumulations corresponded with thedepletion in a vastly different pattern(Fig.4(b)and(c)).Therefore,we determined a C/N molar ratio of 3 to be suitable to sustain effective nitrate removal when sodium acetate was used as the substrate.

The in fluence of the initialconcentration in the in fluent on denitrifying rates was studied in the reactor with an HRT of 8 d and C/N molar ratio of 3,which produced the highest denitrifying rate in previous tests.As shown in Fig.5,although the initialconcentration was different in each set of tests,nitrate was almost entirely removed within 120 h in all the treatment sets.Although theremoval rate in the in fluent,which contained 100 mg/L of,appeared to be highest at the beginning(Fig.5(a)),the differences diminished with time and thelevels in all sets were finally indistinguishable.This result indicates that the reactor is less sensitive to in flowconcentration within the range of 30-100 mg/L,which happens to be the typicalconcentration detected in the Chaohu Lake Catchment.Therefore,the existing concentrations ofloading are not expected to impact the denitrifying efficiency of the wetland.

Fig.4.Effects of C/N ratio on nitrate removal.

Fig.5.Nitrate removal at different initial concentrations.

Groundwater collected from the Chaohu Lake Catchment had aconcentration of 60.35 mg/L.Sodium acetate(619 mg/L)was added to the groundwater to reach a C/N molar ratio of 3.The reactor's HRT was maintained at 8 d.The pH values generally remained constant in the neutral range(7.2-7.6)throughout the tests.was removed rapidly,as shown in Table 1.Theremoval rate reached 46%at 12 h and 65%at 24 h.Within 120 h,theconcentration decreased from 60 mg/L to 1 mg/L,achieving a removal rate of 98%.was initially detected at a maximum of 0.4 mg/L at 12 h,after which it decreased.in the ef fluent was detected at 0.0254 mg/L at 120 h.Theconcentration increased slightly from 0.167 mg/L in the in fluent to about 0.2 mg/L in the ef fluent,which was expected to be readily assimilated biologically as a preferential nitrogen source for heterotrophic bacteria.

To determine the role of lakeside wetlands as a reactive barrier to nitrate in flow into the Chaohu Lake Catchment,the simulated reactor was enhanced by sodium acetate,which was also detected previously in the regional groundwater.The results show thatcan be removed effectively in such a system,and theremoval rate reached 98.25%at the C/N molar ratio of 3 within 120 h.This rate is higher than that reported by Lee et al.(2001),and close to the value reported by Sumino et al.(2006).Results from this study demonstrate that nitratecontaminated groundwater in the Chaohu Lake Catchment can be treated by enhanced denitrification,a measure that can be implemented in the wetlands near Chaohu Lake.

Tests of different HRTs indicate that lower flow rates or longer retention times may be favorable toremoval in the reactor,especially during the first 24 h.A C/N molar ratio of 3 was determined to be the optimal dose for acetate amendment to achieve the highest denitrifying rate.Background denitrification(18.73%ofremoval during 120 h)was probably due to the indigenous substrate contained in the soil samples during the initial reactor inoculation and groundwater collected from the field.Water velocity had little effect on the total nitrate removal rate over 120 h,and it did not appear to be an important parameter in the system performance.This observation is consistent with that reported by Aslan and Tu¨rkman(2005).These data will be incorporated in the field design for the modification and enhancement of lakeside wetlands to achieve the optimal removal of,which would otherwise flow into Chaohu Lake and contribute to the eutrophication of the lake.

Thequalityoftheef fluentsamples wasanalyzedbasedonthe Standard for Groundwater Quality(GB/T 14848-2017)of China and was in compliance with the Class-III water quality in regard to nitrogenous constituents(:20 mg/L,:1mg/L,:0.5 mg/L).Resultsfromthis study demonstrate that enhancing the denitrifying process,for example by adding substrates to the existing wetlands near Chaohu Lake,holds the promise of effectively removing nitrate from the contributing groundwater in the lake catchment.Field-scale tests in the lakesidewetlandsareongoingandexpectedtoobtainoperational parameters that can help achieve the optimal nitrate removal from the in flow groundwater in the region.

4.Conclusions

Based on the results and discussion above,the following conclusions are drawn:

(1)A C/N molar ratio of 3:1 and HRTof 8 d can achieve the most rapiddepletion(from 100 mg/L to 1 mg/L within 120 h).

(3)Different levels of the initialconcentration showed significant in fluence on the removal rates.Tentativeaccumulation was observed but the concentrations decreased to near 0.003 mg/L shortly afterwards.A low level of(up to 0.03 mg/L)was also observed as a result of nitrate reduction.

(4)Biodenitrification of groundwater from a well near the Chaohu Lake Catchment achieved effective removal of nitrate in the sandy reactor by adding sodium acetate at the C/N molar ratio of 3.in the ef fluent reached the standard level of groundwater.

Acknowledgements

The authors thank Song Jin from the University of Wyoming for his help.