Experimental investigation and modelling approach of the impact of urban wastewater on a tropical river; a case study of the Nhue River, Hanoi, Viet Nam

Journal of Hydrology (2007) 334, 347–358

ava i lab le a t www.sc iencedi rec t . com

journal homepage: www.elsevier .com/ locate / jhydrol

Experimental investigation and modelling approachof the impact of urban wastewater on a tropical river;a case study of the Nhue River, Hanoi, Viet Nam

Trinh Anh Duc a,*, Georges Vachaud b, Marie Paule Bonnet c,Nicolas Prieur d, Vu Duc Loi a, Le Lan Anh a

a ICH, VAST, Hanoi, Viet Namb Laboratoire LTHE, HMG, INPG, Grenoble, Francec Laboratoire LMTG, IRD, Toulouse, Franced French Embassy, 57 Tran Hung Dao St., Hanoi, Viet Nam

Received 28 June 2006; received in revised form 3 October 2006; accepted 12 October 2006

00

do

*

KEYWORDSRWQM1;To Lich River;Nhue River;Hanoi;Pollution;Urbanisation

22-1694/$ - see front matte

i:10.1016/j.jhydrol.2006.10

Corresponding author. TelE-mail address: trinhanhdu

r ª 200

.022

.: +84 48c@yahoo

Summary Analyses of water quality and flow regime in combination with laboratory stud-ies and ecological modelling were used to assess the water quality impact of pollutionfrom to To Lich River that drains through Hanoi City and greatly contaminates the NhueRiver. With an average discharge of 26.2 m3/s, the Nhue River receives about 5.8 m3/sof untreated domestic water from the city’s main open-air-sewer – the To Lich River.The studies during 2002–2003 showed high concentrations of BOD (70 mg O2/l), DOC(15 mg C/l), coliform (2.4e6 MNP/100 ml), total phosphorus (3.5 mg P/l), and total nitro-gen (31.6 mg N/l) in the To Lich, while DO level was less than 1 mg O2/l. Such high loads ofuntreated wastewater impacted water quality in the Nhue River where DO decreased attimes to as low as 1 mg O2/l. The accumulation of particulate organic matter andmicro-organisms in the sediments of the Nhue represented substantial sources of nutrientsand sinks for DO. They are also considerable production of dissolved carbon dioxide atconcentrations up to two orders of magnitude higher than pressure. Such pressures(EpCO2) are expected in polluted environments, but the results presented here are newfor Vietnam and much of developing countries. A number of factors linked to field moni-toring and laboratory measurements clearly indicate the importance of autotrophic overheterotrophic biological processes and sediments. An ecological model for managementpurposes has been developed that reliably estimates of the pollutant loads. An opportu-nity was taken to examine the changing impacts and processes when the To Lichwas diverted from the Nhue. The monitoring and modelling of this opportunity showed

6 Elsevier B.V. All rights reserved.

3 61281..com (T.A. Duc).

mailto:[email protected]

1 ‘‘Programme de recherches Francodes eaux en zone urbaine’’ between CeScientifiques, CNRS, France and Vietnand Technology, with the support ofAffairs and Vietnamese Ministry of Sciewww.waterprog-frvn.org.vn.

348 T.A. Duc et al.

low dissolved oxygen levels even if the impact from the To Lich was lessened. Alternativesare proposed to alleviate problems of water quality in the Nhue. It is concluded that thetreatment of the To Lich River’s water is highly recommended; otherwise a reduction toone third of current wastewater discharge is needed to bring water quality back to theenvironmental standard.ª 2006 Elsevier B.V. All rights reserved.

Introduction

Rapid urbanisation and major economic development inVietnam has led to dramatic degradation of the environ-ment and increased health risks due to inefficient process-ing of the increased burden of effluent and solid wastes(Bolay et al., 1997). A major example of this is the Nhue River,a tributary of The Red River, which is polluted by Hanoi City(Fig. 1). With a population of more than three million and nowastewater treatment facility (JICA, 1995), the city’sdomestic wastewater has resulted in severe environmentalissues for the Nhue downstream its confluence with the ci-ty’s main sewer, the To Lich River (JICA, 1995).

This paper examines the water quality issues and remedi-ation strategies for the Nhue based on the efforts of a multi-disciplinary scientific program1 launched in 2001 to assessthe environmental state of the river, especially around Ha-noi, and to develop suitable treatment methods for the pol-luted water. For this, the water quality and flow regime wasused in combination with a 1-D physico-biochemical model(Trinh et al., 2006a). This was done to (1) investigate theecological state of the Nhue River, (2) quantify the impactof wastewater rejected into the river (both point and non-point sources) and (3) evaluate several practicable manage-ment alternatives.

To our knowledge, very few modelling attempts havebeen made to quantify and predict the ecological interac-tions in a similar highly polluted subtropical river in Vietnamand other countries in south East Asia (Quynh et al., 2005;Trinh et al., 2006a; McAvoy et al., 2003). Therefore, interms of ecological modelling, this work is one preliminaryfor ecological modellers to work on highly polluted subtrop-ical river systems, an issue of fundamental importance tothe development of environmental management tools andremediation strategies for such an important region of theworld.

The studied area and the experimental surveys

The river basin studied comprises of two subcatchments,the Nhue and the To Lich (Fig. 1) with respective catchmentareas of 1070 km2 (Cu et al., 2005) and 77.5 km2 (JICA,1995). The To Lich River stems from West Lake in the north-ern part of Hanoi city and it flows southward through Hanoito join the Nhue River downstream. Domestic and industrialwastewater from Hanoi mainly discharges to the To Lich River

Vietnamien sur la pollutionntre National de Recherchesamese Academy of ScienceFrench Ministry of Foreignnce and Technology; http://

without prior treatment and the river is effectively the prin-cipal open-air-sewer of the city. The Nhue River, as abranch of The Red River, takes its source from the Red Riverabout 11 km to the north west of Hanoi and is joined withthe To Lich some 20 km downstream of Hanoi. The mean in-flow to the Nhue River from the Red River is 26 m3/s and ittypically receives around 5.8 m3/s of untreated wastewaterfrom the To Lich River.

Monthly water quality surveys were conducted from Jan-uary 2002 to October 2003 at seven key locations (Fig. 1).Five of these monitoring points (N1, N2, N3, NT1 and NT2)are on the Nhue River at 0 km, 8 km, 15.2 km, 25.2 km,and 33 km from its source, the Red river. There is alsoone monitoring point for the Red river 4 km upstream ofthe Nhue’s source and one monitoring point (TL) in the ToLich River at 800 m upstream of its confluence with theNhue. The information collected during these surveys in-cluded topographic, hydrodynamic, physico-chemical,chemical and biological conditions (Table 1). Water flow

Figure 1 Map of the studied area.

http://eosweb.larc.nasa.gov/


Table

1Ave

rage

monthly

waterqualitysamplingdataco

llectedduringJanuary20

02–Augu

st20

03

Point

DO

(mgO2/l)

pH

Turbidity(NTU)

Conductivity(lS/cm

)Alkalinity(m

gHCO3/l)

Ca(m

g/l)

Min

Mean

Max

Min

Mean

Max

Min

Mean

Max

Min

Mean

Max

Min

Mean

Max

Min

Mean

Max

R5.72

7.08

8.10

6.74

7.55

8.15

4040

413

4315

719

222

598

124

153

25.60

28.06

32.00

N1

4.65

6.82

8.09

6.69

7.57

8.20

2526

064

115

519

722

812

212

815

225

.60

28.29

32.00

N2

2.86

5.76

7.70

6.64

7.37

7.98

2714

664

018

023

933

412

213

717

125

.56

29.08

33.60

N3

0.74

4.88

7.10

6.70

7.28

7.87

1210

555

118

026

643

212

214

619

526

.80

29.74

35.20

TL

0.00

1.35

5.80

6.82

7.35

8.10

1511

953

223

565

896

013

430

140

327

.20

37.14

48.00

NT1

0.32

3.39

6.60

6.74

7.35

8.50

2298

522

199

318

588

122

167

244

24.00

31.23

36.80

NT2

0.54

3.60

6.72

6.40

7.25

8.53

1695

510

202

302

484

122

168

281

25.60

31.26

40.00

NH4(m

gN/l)

PO4(m

gP/l)

DOC(m

gC/l)

BOD(m

gO2/l)

Chlorophyll-a(m

g/m

3)

Coliform

(MNP/1

00ml)

Min

Mean

Max

Min

Mean

Max

Min

Mean

Max

Min

Mean

Max

Min

Mean

Max

Min

Mean

Max

R0.02

0.26

0.77

0.00

90.015

0.020

0.55

2.12

3.87

2.0

6.4

25.2

1.75

4.99

8.39

1.7e

021.53

e05

1.6e

06N1

0.02

0.21

0.85

0.00

20.008

0.015

0.69

2.23

4.47

2.0

7.4

25.2

1.18

5.44

9.26

2.3e

022.77

e04

2.2e

05N2

0.20

0.64

1.80

0.02

00.060

0.123

1.41

2.86

4.80

3.0

9.6

27.3

2.71

8.12

18.14

5.0e

024.84

e05

2.8e

06N3

0.37

0.69

1.57

0.03

60.093

0.194

1.61

3.04

5.47

4.0

10.2

25.2

3.30

10.93

20.86

3.0e

021.31

e05

1.1e

06TL

0.47

11.34

33.45

0.041

1.31

32.435

1.92

7.25

15.26

7.0

37.6

67.5

4.01

33.60

84.87

9.0e

032.92

e05

2.4e

06NT1

0.32

2.68

5.19

0.04

30.294

0.744

1.66

3.80

5.73

5.5

13.3

28.5

2.92

12.37

22.15

7.0e

021.59

e05

1.6e

06NT2

0.30

2.41

5.31

0.04

00.344

1.116

1.62

4.34

7.98

4.0

14.7

44.0

4.01

15.03

23.10

9.0e

022.97

e05

3.0e

06

Experimental investigation and modelling approach of the impact of urban wastewater on a tropical river 349

and river bed topography were determined by using an ADCP(Rio-grand, USA). In situ physico-chemical properties weremeasured with a hydrolab Multi-sonde 4a Surveyor (Hach,USA). Water and sediment upon sampled were pretreatedat sites according to analytical requirements and preservedin dark container at 4 �C for laboratory analysis. Most anal-yses were carried out within a day at Institute of Chemistry,Vietnamese Academy of Science and Technology. The ana-lytical protocols applied for these analyses were extractedmainly from Standard Methods for the Examination of Waterand Wastewater (Clesceri et al., 1999).

In order to construct the hydrodynamic model, dailywater levels for the study period were used based on infor-mation on the Thuy Phuong, the Cau Den, the Dong Quanand the Thanh Liet dams along the river course (Fig. 1).

Model description

A 1-D ecological model was used in this study and details ofthe model are presented by Trinh et al. (2006a). The 41 kmriver length from the Thuy Phuong dam to the Dong Quandam was discretized into 3 reaches (Fig. 1). The Thuy Phu-ong dam (inflow from The Red River) regulates the firstreach, extending from point N1 to point N3, and the CauDen dam. The second reach is limited between the CauDen dam and confluence between the Nhue and To Lich Riv-ers. The third reach extends downstream of the confluenceto the Dong Quan dam (Fig. 1). The lengths of the main stemof the rivers for these reaches are 15.2, 5.0 and 20.8 km,respectively. Inflow of untreated wastewater to the Nhuewas included within the model for the first and secondreaches (JICA, 1995). The concentrations of the various pol-lutants in these inputs were taken as that of the To Lich Riverbased on the fact that the lateral inflow consisted mainly ofuntreated domestic wastewater.

The biochemical module of the model was primarily ta-ken from the RWQM1 model of Reichert et al. (2001). Forthis, the aquatic organisms and materials were subdividedinto four pools: (1) the inorganic materials including nutri-ents and major chemical substances in natural water; (2)the degradable (dead) and inert organic materials in dis-solved and particulate phases; (3) the phytoplankton bio-mass; (4) the heterotrophic and autotrophic bacteria.However, the biochemical module had to use a modifiedversion of the RWQM1 conceptual scheme in three aspects.

Firstly, instead of counting only the benthic micro-organ-ismmicro-organisms (e.g. sessile algae and bacteria), ourmodel takes into account the suspended micro-organismsas a major biomass because the Nhue water is highly turbidand much of this turbidity is associated with suspended mi-cro-organisms. As light extinction is very rapid in the watercolumn, minimal photosynthetic activity occurs at thewater column–sediment interface. Hence, only light-inde-pendent biochemical conversions are taken into account.A previous study on sediment of the Nhue-To Lich River sys-tem showed that sediment–water fluxes of dissolved sub-stances were strong, especially downstream theconfluence between the two rivers (Trinh et al., 2005).Clearly, these exchanges (both dissolved and particulatefluxes) are needed to be incorporated in the model. Becauseof insufficient direct information, an indirect estimation of

Table 2 Parameter values of the Nhue River model

Symbol Description Value Range Unit

(1)k�decay;alg;T Decay rate of phytoplankton at 20 �C 0.06 0.05 1/d(1)k�growth;alg;T Maximum growth rate of phytoplankton at 20 �C 2.12 0.4 1/d(1)k�decay;H;T Decay rate of heterotrophic bacteria at 20 �C 0.06 0.1 1/d(1)k�growth;H;T Maximum growth rate of heterotrophs in aerobic condition at 20 �C 1.34 0.4 1/d(1)k�growth;H;anox;T Maximum growth rate of heterotrophs in anoxic condition at 20 �C 0.23 0.32 1/d(2)k�decay;auto;T Decay rate of nitrifying bacteria at 20 �C 0.05 0.025 1/d(1)k�growth;Auto;T Maximum growth rate of nitrifying bacteria at 20 �C 1.73 0.5 1/d(1)k�hyd;T Maximum hydrolysis rate at 20 �C 2.98 0.5 1/d(1)kads;PO4

First order equilibrium of adsorption/desorption of PO4 10.0 2.0 1/d(4)kads;PO4

Adsorption equilibrium of PO4 35.0 7.0 l/kg(2)KS,H Half-saturation coefficient of degradable dissolves in het. growth 2.0 1.0 gC/m3

(2)KO2 ;H Half-saturation coefficient of oxygen in heterotrophic growth 1.0 0.5 gO/m3

(2)KNH4;H;aer Half-saturation coefficient of NH4 in growth of heterotrophs 0.2 0.1 gN/m3

(2)KNO3 ;H Half-saturation coefficient of NO3 in anoxic heterotrophic growth 0.1 0.05 gN/m3

(2)KP,H,aer Half-saturation coefficient of PO4 in aerobic heterotrophic growth 0.02 0.01 gP/m3

(2)KP,H,anox Half-saturation coefficient of PO4 in anoxic heterotrophic growth 0.02 0.01 gP/m3

(2)(3)KO2 ;auto Half-saturation coefficient of oxygen in nitrifying bacterial growth 0.5 0.25 gO/m3

(2)(3)KNH4;auto Half-saturation coefficient of NH4 in nitrifying bacterial growth 3.5 1.75 gN/m3

(2)(3)KP,auto Half-saturation coefficient of HPO4 in nitrifying bacterial growth 0.02 0.01 gP/m3

(1)KN,Alg Half-saturation coefficient of nitrogen in phytoplanktonic growth 0.1 0.05 gN/m3

(2)KP,alg Half-saturation coefficient of HPO4 in phytoplanktonic growth 0.02 0.01 gP/m3

(5)KO2 ;hyd Half-saturation coefficient of oxygen in hydrolysis 0.2 0.1 gO/m3

(5)Khyd Half-saturation coefficient of organic particulates in hydrolysis 0.03 0.015 –(2)balg Temperature dependant coefficient of phytoplankton 0.046 0.0092 1/�C(2)bH Temperature dependant coefficient of heterotrophic bacteria 0.07 0.014 1/�C(2)bAuto Temperature dependant coefficient of nitrifying bacteria 0.098 0.0196 1/�C(2)bhyd Temperature dependant coefficient of hydrolysis 0.07 0.014 1/�C(2)IK Saturation light intensity 500 250 W/m2

Imax Maximum light intensity at noon from NASA (2003) 550 55 W/m2

(2)YH,growth,aer Yield for heterotrophic growth with dissolved oxygen 0.6 0.12 –(2)YH,growth,anox Yield for heterotrophic growth with nitrate 0.6 0.12 –(2)YH,decay Yield for heterotrophic decay 0.62 0.124 –(2)YAuto,growth Yield for autotrophic growth 0.13 0.026 –(2)YAuto,decay Yield for autotrophic decay 0.62 0.124 –(2)Yhyd Yield for hydrolysis 0.99 –(2)Yalg,decay Yield for algae decay 0.62 0.124 –(2)fI,Auto Fraction of autotrophs becoming inert in autotrophic decay 0.2 0.04 –(2)fI,alg Fraction of algae becoming inert in algae decay 0.2 0.04 –(2)fI,H Fraction of heterotrophs becoming inert in heterotrophic decay 0.2 0.04 –(1)hsed Active sediment layer where bacterial activity take place 0.005 0.0025 M(1)vs Settling rate of particulate matter in water column 3.9 0.173 m/d(1), (2), (3), (4), and (5) indicate the sources from Trinh et al. (2006a), Reichert (2001), Brion and Billen (2000), Garnier et al. (2000), andHenze (1987), respectively.

350 T.A. Duc et al.

sediment biomass was made. Since the suspended sedimentconcentration in the Nhue River decreases continuouslydownstream (Table 1), it indicates that the top sedimentlayer is renewed by a new deposition of suspended solid.Assuming a similar organic/inorganic ratio in the upper partof the sediment as in the water column suspended solids,the biomass of bacteria and OM contents were estimatedusing the equation:

Xorg ¼SorgSSPM

XSPM ð1Þ

X is the concentration in the sediment of either organicmaterial (Xorg), or of inorganic suspended material (XSPM)

(g/m2), Sorg and SSPM the concentration in the water columnof organic material and inorganic suspended solids, respec-tively. Similar equations are applied for sedimentation ofbacterial biomass (Trinh et al., 2006a).

The advantage in this simplification is that it represents aspatial change of sediment impact to water level, whichintensifies near the pollution point sources and eases up inless impacted water areas. This reflects our in situ observa-tions. All detailed conversion processes are represented inTrinh et al. (2006a).

Secondly, the zooplankton is not considered separatelyhere due to data deficiency. Also nitrification is simplifiedas a conversion of NH4 to NO3, not via intermediate NO2.


Thirdly, modification was made to adapt to the physicalprocesses such as exchange of gas as a function of waterand wind speed (O’Connor and Dobbins, 1958; Wanninkhof,1992). The gas exchange modification assisted in the unsteadystate simulation.

A qualitative stoichiometric matrix illustrating all aqua-tic species and bio-physicochemical conversion processestaken into consideration of the Nhue River model is repre-sented in Trinh et al. (2006a). The calibrated values of modelparameters are represented in Table 2.

R R R

RR

R

R

RR R

N1 N1 N1N1 N1

N1

N1

N1N1 N1

N2 N2 N2

N2

N2

N2

N2

N2 N2N2

N3 N3 N3

N3

N3

N3

N3

N3N3

N3

NT1NT1NT1

NT1

NT1NT1

NT1

NT1

NT1

NT1NT2NT2NT2

NT2

NT2

NT2

NT2NT2

NT2

TL

TLTL

TLTL

TL TL

TL

TL

TL

0

20

40

60

80

Jan

Mar

May Ju

lSep Nov

EpC

O2

Figure 2 Excess carbon dioxide partial pressure at differentsampling points in 2002.

Water quality characterisation of the NhueRiver

Trophic conditions

From a management view point, stream classification basedon trophic state provides a criterion for delineating thosesystems most at risk from nutrient and other pollution.According to the USEPA (2000), the trophic state of a riveror stream is classified by using physical factors and nutrientgradients. Besides the physical factors directly influencingthe river trophic status (flow rate, water depth, tempera-ture), nutrients contents, and algal/phytoplankton biomassare primarily used to classify trophic state.

Based on the hydrodynamic conditions, the Nhue Riverdoes not support significant algal growth because its waterwas usually turbid (Janssen-Stelder et al., 2002; Pruszaket al., 2005). The monthly surveys specified an average up-stream water turbidity of higher than 50 NTU (Table 1). Thishigh turbidity slowed down the autotrophic growth espe-cially for the bottom-attached-organisms. Another variableusually used to classify the trophic level of aquatic systemsis chlorophyll-a. However, in flowing water bodies absolutevalue of chlorophyll-a may not really represent the auto-trophic state of the river because most flowing systems dis-perse phytoplankton before high biomass develops (USEPA,2000). Another factor that enriches the downstream chloro-phyll-a in the Nhue is the contribution of effluent. In prac-tice, all of the indicators of the trophic level are poor:even in the nutrient rich condition the water current mayprevent nutrient assimilation and overlap with the biologi-cal production to produce a low trophic state. Nonetheless,the sediment plays a crucial role in this river system (Trinhet al., 2005) and wastewater effluence can easily turn thesystem from autotrophic to a heterotrophic.

Dissolved carbon dioxide provides an indication of thebalance between photosynthesis and respiration by biota.Carbon dioxide also transfers from the water column tothe atmosphere as CO2 gas and precipitation as, for exam-ple, calcium carbonate minerals and in the process removephosphate from solution and act in part as a self cleansingmechanism within the river (House, 1989; Maberly, 1996;Hartley et al., 1996; Neal, 2001a). Dissolved carbon dioxidecan be used as powerful indicator of trophic state of thestudied river (Neal et al., 1998b). Dissolved carbon dioxidehas not been directly determined in this study. However,as introduced by Neal et al. (1998a), the excess carbon diox-ide partial pressure (EpCO2) is used as the alternative fordissolved carbon dioxide as that this parameter can be eas-ily calculated from the pH and Alkalinity which are standard

determinants in environmental studies. In this study, we ap-plied the following formula (Neal et al., 1998a)

EpCO2 ¼ð0:95� AlkÞ � 106�pH

ð6:46� 0:0636� t�CÞ ð2Þ

For this equation, alkalinity (Alk) is that which correspondsto the ‘normal’ measurement of bicarbonate alkalinity ex-pressed in leq/l units and t� is temperature (�C). In Neal’s(1998a) paper, the alkalinity is slightly different to that usedin this study – he used Gran alkalinity: the difference is ex-plained in Neal (2001b) and the equation presented in thispaper is a slight modification to that presented in Nealet al. (1998a).

According to Neal et al. (1998b), an autotrophic/freshsystem tends to maintain an EpCO2 smaller than 1 while het-erotrophic/polluted system has an EpCO2 larger than 1. Asall EpCO2 values calculated for sampling points during2002 are larger than 1 (Fig. 2), the Nhue River system isshown to be heterotrophic/polluted dominated. Indeed,the variation in EpCO2 is remarkable with a range from 1.7to 110 (the mean value is 25.7) and a strong seasonality thatvaries across sites. For example, the variability of EpCO2 atupstream points (R, N1 and N2) changed from lower than 10in winter up to higher than 30 in summer (Fig. 3). It impliesthat the river system was more prone to heterotrophic statein summer than in winter. In the next subsection, this sea-sonality variation will be discussed in detail. Another remarkdrawn from the EpCO2 results is that the environment alongthe river course was clearly deteriorated. For instance, inwinter, EpCO2 increased gradually from 5 upstream to 16downstream. In summer, this tendency was not as obviousas in winter but downstream EpCO2 was generally higherthan upstream one. For some sites downstream EpCO2 werea hundred fold higher than atmospheric pressure and clearlythere is massive respiration and CO2 production occurring inthe rivers and this is characteristic of urban/industrially im-pacted rivers (Neal et al., 1998b). Thus, EpCO2 providesimportant material for this environmental evaluation. More-over, EpCO2 is highly correlated for all adjacent sites andthis shows the importance of biological activity.

In addition to this judgement, the longitudinal variationsof DOC and primary production/respiration ratio (P/R ratio)were also taken into consideration. The inflow and

Feb AprJu

nAug

OctDec

0

1

2

3P

hae

op

hyt

in/c

hlo

rop

hyl

l-a Ups. P/C

Down. P/C

0

10

20

30

Ups. EpCO2

Down. EpCO2

EpC

O2

Figure 3 The phaeophytin/chlorophyll-a ratio and EpCO2 in2002.

352 T.A. Duc et al.

in-stream DOC are related to the trophic index because DOCis an important energy source driving the heterotrophiccommunity and DOC levels can influence riverine responseto algal growth. Streams and rivers enriched with fast-degradable DOC are generally dominated by a heterotrophiccommunity and may be more prone to low oxygen events.The P/R ratio, on the other hand, has long been recognisedas indicating the relative autotrophic (P/R > 1) or heterotro-phic (P/R < 1) characteristic of streams and rivers (Vannoteet al., 1980). Since the primary production and respirationresult in the producing and consuming oxygen, the P/R ratiois quantified by the ratio of the oxygen from primary pro-duction to the oxygen consumed by respiration. In this casestudy, the DOC was experimentally collected while (P/R ra-tio) was simulated with the help of ecological model.

Again, the Nhue River appears as an heterotrophic sys-tem since the P/R decreased from 3 upstream to always lessthan 1 after about 7 km downstream and DOC increasessteadily downstream (Table 1). This feature confirms thatthe Nhue River becomes more polluted-downstream. Fur-ther, the heterotrophic state of the Nhue River is higherthan in the Red River near to their join and this providesadditional evidence for the environmental deteriorationalong the length of the Nhue River.

As pointed out earlier in the paper, if calcite saturationoccurs within the waters there is a potential for phosphateremoval by co-precipitation. In fact, PO4 increased down-stream so the removal of PO4 was very unlikely in studiedriver. To examine this possibility, the calcite saturationwas estimated based on alkalinity, pH, calcium and temper-ature information and the thermodynamic method providedby Neal et al. (1998a). On average, the waters are about sat-urated with respect to calcite, but the saturation varies be-tween about a tent and ten times saturation: log10SICalciteaverages �0.07 with a range of �1.03 to +1.01. Log10SICalciteexhibits a strong linear relationship with pH and this corre-sponds with a change in carbonate concentration as a func-tion of pH, but with near constant calcium and alkalinitylevels and it implies that there is no calcite equilibriumwithin the water column (c.f. Neal et al., 1998c). This lackof equilibrium would be expected for the waters studiedhere as high DOC and phosphate concentrations inhibit pre-cipitation (Neal, 2001a, 2002). Thus, the results indicate

that phosphate self cleansing mechanisms are not importantin the water column of the rivers examined in this study.

Seasonal variation in the river

In North Vietnam, temperature variation approaches 20 �C(the maximum, average, and minimum temperatures are15, 25, and 35 �C, respectively) and solar irradiation in win-ter is only approximately one fifth of the level in summerdue to the monsoon climate (NASA, 2003). As a result, bio-logical activities can vary considerably with seasonal. Asstated in previous subsection, EpCO2 in summer were gener-ally higher than in winter time, showing a possible seasonal-ity in biological activities. To verify this seasonality remark,the ratio of [phaeophytin]/[chlorophyll-a] (P/C) was usedsince this ratio depends mostly on the state of the phyto-planktonic community. Phaeophytin is a degradation prod-uct of chlorophyll-a: a relatively low value of the P/Cratio indicates an active growth of phytoplankton. Indeed,this ratio is frequently used to assess the healthy state ofphytoplankton (USEPA, 2000).

Concentrations of phaeophytin and chlorophyll-a weremeasured using method of Lorenzen (1967) and their ratiosat the upstream (R and N1) and downstream (NT1 and NT2)points were examined. These selections guaranteed thatfresh and polluted waters were both taken into accountand that the measures represent the overall seasonal tro-phic variation of the Nhue. The monthly change of P/C ratioindicates an unhealthy state of phytoplankton in summertime (from April to August) at both the upstream and down-stream positions (Fig. 3). This finding is consistent with theEpCO2 results and other studies linked to a similar climate asthat being considered in this paper (Luong et al., 2005; Tanget al., 2003; Paresys et al., 2005). For the Nhue, there aretwo possible explanations for the unhealthy state of phyto-plankton in the summer. First, the high temperatures in thelater spring/early summer cause an increase of planktonicactivity (both zooplankton and phytoplankton). Conse-quently, in summer, as phytoplanktonic biomass increases,zooplanktonic activity increases and leading to phytoplank-tonic consumption faster than during other times of theyear. Therefore, the high mortality of phytoplankton inthe summer causes a high P/C ratio. Second, in the Red Riv-er water, strong current during the rainy season increaseswater turbidity, disperses phytoplankton and subsequentlyprevents phytoplankton from primary production. Phyto-plankton growth consequently decreases leading to an in-crease of the ratio between phaeophytin and chlorophyll-a.

Assessment of the impact of human activitieson the river system

Since urbanisation and agriculture were intensified, theassessment of anthropogenic impacts to the Nhue is crucial.Two main types of wastewater sources to the Nhue Riverhave been identified: diffuse (non-point) sources inflowalong the upstream reach and point sources for the To LichRiver (Trinh et al., 2006a). Here, the application of experi-mental data and the numerical simulation for characterisa-tion and quantification of these two sources of impact arediscussed.

0

1

2

3

4

5

6

7

0 20 40 60

Discharge (m3/s)

DO

(mg

O2/

l)Opening of Thanh Liet dam

Closing of Thanh Liet dam

Figure 4 Measurements and simulation of DO at point NT1before and after the close of the To Lich effluence; Solid line:simulation of period April–August 2002, Dash line: simulation ofperiod April–August 2003.


The To Lich River’s wastewater impact

The relationship between routinely measured water qualitydeterminands between the different monthly samplingpoints was used to highlight the differences between the up-stream and downstream waters. For this analysis, chemi-cally and biologically conservative water qualitydeterminands provide key indicators of water sources whenthe water sources have contrasting ‘end-member’ concen-trations. The chosen determinands need to be chemicallyand biologically conservative (or approximately so) in orderto avoid complications of, for example, within-river losses.If the variations in the concentrations of these conservativecomponents between two adjacent sampling points are dif-ferent, waters at these points are likely released from dif-ferent sources. In this study, conductivity and hardnessare used. Monthly concentrations showed high correlationfor points N1, N2, and N3. This indicates that the water inthe upstream reach came essentially from the same source(Table 3). Also, the high correlation of Conductivity andHardness between NT1 and NT2 imply an identical watersource downstream the study site. The low correlation be-tween N3 and NT1 implies a dissimilarity between the up-stream and downstream confluence Conductivity andHardness. This difference is clearly essentially due to theTo Lich River’s input. It proves that the impact of the ToLich River is significant to water quality of the Nhue River.

Owing to a Hanoi water management project, the ThanhLiet sluice gate was closed for reconstruction in April 2003and a large portion of wastewater was temporarily divertedto the Yen So reservoir before being pumped to the Red River(Fig. 1). This provided a valuable opportunity to assess theimpact of To Lich wastewater to the Nhue. Two prominentparameters, DO and NH4, were selected to quantify the ToLich River’s impact because among principal domesticwastewater quality indicators, DO and NH4 are consideredas state variables in our ecological model. Here, the DOand NH4 at point NT1 for periods with open sluice gates(April–August 2002) and closed sluice gates (April–August2003) were linked to the river discharges at point N3 (Figs.4 and 5). This was done to examine the impact of the ToLich River’s water at different discharge levels.

Table 3 Correlation matrix of the monthly sampling con-ductivity and hardness between different the observationpoints using the statistical computer program SPSS (2002)

N1 N2 N3 NT1 NT2

N1 Conductivity 1Hardness

N2 Conductivity 0.75 1Hardness 0.88

N3 Conductivity 0.73 0.96 1Hardness 0.75 0.83

NT1 Conductivity 0.54 0.61 0.65 1Hardness 0.54 0.65 0.68

T2 Conductivity 0.48 0.52 0.50 0.94 1Hardness 0.46 0.64 0.67 0.93

Bold: Correlation is significant at the 0.01 level (2-tailed).

After the closure of the sluice gate, the NH4 content re-duced (Fig. 5) but there was no corresponding improvementin DO (Fig. 4). In order to understand this difference, theecological model was employed. The main differences be-tween the gate-open and the gate-closed periods were theTo Lich wastewater inflow and the temperature. The aver-age temperatures during these two periods were 26.8 and29.2 �C, respectively. The average To Lich influence on flowbefore and after April 2003 was 5.82 (m3/s) and 1.10 (m3/s),respectively. The simulation results show that the To Lichclosure had only limited influence the increase of dissolvedoxygen (Fig. 4). The simulated results indicate a tempera-ture sensitiveness of dissolved oxygen simulation that re-flects well with the observed changes. In contrast, forNH4, considerable changes were observed and this matchedwell the model simulations. For instance, at point NT1, be-fore closure, with an average discharge of 26.62 (m3/s), thesimulated dissolved oxygen and NH4 concentrations were4.25 (mg O2/l) and 2.50 (mg N/l), respectively, while afterclosure, the simulated values were 4.88 (mg O2/l) and 0.97(mg N/l), respectively.

0

1

2

3

4

5

6

7

8

0 20 40 60

Discharge (m3/s)

NH

4 (m

g N

/l)

Opening of Thanh Liet dam

Closing of Thanh Liet dam

Figure 5 Measurements and simulation of NH4 at point NT1before and after the close of the To Lich effluence; Solid line:simulation of period April–August 2002, Dash line: simulation ofperiod April–August 2003.

0 5 10 15 20 25 304

8

12

16

20

24

28

DO

C (

mg

C/l)

Time (days)

Aerated TL Unaerated TL Aerated + sediment TL Unaerated + sediment TL Aerated N3 Unaerated N3 Aerated + sediment N3 Unaerated + sediment N3

Figure 6 DOC change of samples collected at point TL andN3.

354 T.A. Duc et al.

Estimation of wastewater lateral inputs by usingconductivity

Our surveys indicated a considerable degrading of the up-stream water quality in the downstream reaches. For in-stance, conductivity increased gradually while pH and DOdeceased along the upstream reach (Table 1). Such resultsdenote a significant input of high conductivity wastewater(Table 1). In polluted rivers, conductivity can often be usedas an indicator to determine the wastewater pollutionsource because it mostly depends on inert variables likebase cations (Na, K, Ca and Mg) and strong acid anions(Cl, SO4, NO3) that are slightly modified by biological activ-ities (USEPA, 2000). The conductivity is also in part con-trolled by calcium and bicarbonate associated withweathering of the bedrock and soils in the area, but thisterm probably provides a constant background on whichthe pollutant components add to the conductivity. In orderto estimate the lateral inflow value, a simplified version ofthe coupled hydrodynamic and biological model of Trinhet al. (2006a) was used. This simplified model examinedconductivity to estimate the lateral inflow; no biologicalconversion process was included. The conductivity of the in-flow was assumed equal to the average conductivity of theTo Lich wastewater during the dry period (800 lS/cm).The data obtained from the monthly samplings of 2002 wereused together with the computer program AQUASIM (Reic-hert, 1998) for the estimation. The lateral input (QLat) wasestimated to be to 0.091 m3/s/km. This value is much high-er than reported earlier for lateral wastewater inflow (JICA,1995) but the present value is probably more realistic be-cause the JICA (1995) report was made prior to the eco-nomic reform that leads to rapid economic growth andpopulation expansion in Hanoi.

02468

1012141618

DO

C (

mg

C/l)

BOD (mg O2/l)

Upstream Downstream DOC

up=0.34*BOD

R2=0.522 DOC

down=0.2*BOD

R2=0.573

0 10 20 4030 50 60 70

Figure 7 Linear relationships between BOD and DOC inupstream and downstream waters.

Fractions of dissolved organic matter (DOC) inwater of the study area

The degradable fraction of DOC was assessed in the labora-tory at ambient temperature (average 29 �C) with samplescollected at fresh (N3) and polluted (TL) points. After col-lection, the samples were divided into four sub-samples toexamine different conditions that might well apply withinthe river. The first sub-sample was incubated without aera-tion, the second was constantly aerated to maintain high le-vel of DO, the third was not aerated but some sediment wasadded, and the fourth was aerated with some sediment. Theincubation time lasted for 28 days. The physico-chemicalconditions were daily recorded. The DOC, BOD, COD, andnutrients were measured on the zero, first, 7th, 14th,21st, and 28th days. The profiles of DOC are shown in Fig. 6.

The biodegradable fraction of DOC (BDOC) at point TL(61.3 ± 1.0%) was about twice that at point N3(31.2 ± 1.6%). The BDOC in the To Lich River’s water wasslightly lower that the results found by Servais et al.(1999) (73.6%) and by Vuillemin et al. (2003) (61.2–76.3%)for raw wastewater. It is explained that in an open-air-sew-er like the To Lich River, biodegradation has already takenbefore sample was collected and reduced fraction of biode-gradable matter. The retention time of domestic wastewa-ter (from the source to the sampling point TL) is from 1 to 3

days. High similarity was found between the BDOC at pointN3 and that in treated wastewater (about 38.9%) (Servaiset al., 1999) and in river water (36.07 ± 14.35%) (Friaset al., 1992). Also, in the exposed irradiation and uncon-trolled high temperature the biodegradation was faster thanthat reported in other publications (Block et al., 1992;Vuillemin et al., 2003). Large portion of BDOC was removedwithin few days (Fig. 6).

The DOC and nutrient (NO3 and PO4) concentrations inthe tests with sediment being present reduced faster thanin tests without sediment. This feature indicates that sedi-ment took a role in nutrient and organic matter removaland the process probably links to the autotrophic and het-erotrophic bacteria abundant in the muddy sediments ofthe river. Indeed, there are usually high correlation be-tween BOD and biodegradable TOC (BTOC), and in lesser ex-tent, between BOD and DOC (Servais et al., 1999).

Within this study, the relationship between the DOC andBOD was examined using the monthly data. Initially, thedata was separated into two groups: the highly polluted(TL, NT1, NT2) and less polluted (R, N1, N2, N3) locations.Then, linear relationships were established between DOCand BOD for these two data groups. The DOC/BOD ratio inpolluted-downstream section (DOC/BOD = 0.2) was lower


than in less polluted-upstream section (DOC/BOD = 0.34)(Fig. 7) as expected. However, it was still higher that the re-sult found by Servais et al. (1999) in the Paris’s raw waste-water (DOC/BOD = 0.13): biodegradation was alreadyprocessed in the polluted zone and the biodegradable or-ganic matter fraction was reduced in our study.

Application of ecological model in calculationof mass balance; examples of the DO and NH4

A mass balance assessment gives indications to what, whereand how the ecosystem can respond, or requires externalsupport, to obtain recovery from a certain pollution status.In this study, the DO and NH4 (two prominent water qualityindicators listed in the Vietnamese water quality standards)at two different locations (N2 and NT1) were selected forthe mass balance computation. The conceptual scheme ofthis ecological model appoints DO fluxes to four mainpools/compartments: exchanged with the atmosphere, con-sumed by bacteria, exchanged with sediment and primaryproduced. Similarly, fluxes of NH4 were directed to fourmain pools/compartments: consumption/release by bacte-ria, primary production, nitrification and sediment ex-change. It was modelled that, at midday, the simulationresults gave a positive DO mass balance (0.53 mg O2/l/d)at N2 and negative DO mass balance at NT1 (�2.36 mg O2/l/d). This result confirmed the heterotrophic status of thedownstream river. In case of the NH4, the corresponding val-ues were �0.167 (mg N/l/d) at N2 and �0.912 (mg N/l/d) atNT1. The constant negative mass balance of NH4 along theriver course implies that NH4 was abundant in the river sys-tem. At the upstream point, the sediment oxygen demand(SOD) was nearly equal to the consumption by bacteria,but at downstream point the SOD was considerably higherthan the consumption by bacteria. This difference provesthat the sediment compartment is critical factor in DOdepletion downstream. This was also the case of NH4 ofwhich nitrification and sediment consumption were similarupstream but different downstream (sediment consumptiontook a larger portion of NH4 than nitrification did). On theother hand, the primary production increases downstreamand this is shown both by the simulation and the increaseof chlorophyll-a (Table 1). There are two reasons for thisfeature. First, due to sedimentation, the downstream reachwas always less turbid than the upstream one. Second, chlo-rophyll-a in the To Lich wastewater is very high, probablydue to a high cyanobacteria (blue green algae) biomass.

Table 4 The total maximum daily load (TMDL), the current waste(P

LA)

Environmentalparameters

Surface water standards for purposes otherthan domestic water supply (mg/l)

BOD <30DO >2SS <80NHþ4 (as N) <1.5NO�3 (as N) <3.4

Bold: Current loads do not meet the Vietnamese standards for surface

The substantial input of blue green algae biomass combinedwith a nutrient enrichment is favourable to an intensifica-tion of primary production downstream the confluence.

Nitrification probably strongly influences the NH4 bal-ance, especially at point N2. At high NH4 concentrationand sufficient dissolved oxygen levels, nitrification is themost important out-sourcing of NH4. At point NT1, the highnitrification rate is attributed to the high NH4 concentra-tions and the large amounts of nitrifiers in the domesticwastewater effluent (Brion and Billen, 2000).

Water quality improvement; cases ofmanagement alternatives

Within the scope of this study, an established guidance/tool, the total maximum daily load (TMDL), is used to quan-titatively evaluate several possible management alterna-tives. In principle, a TMDL is the sum of the individualwaste load allocations (WLA) for point sources and load allo-cations for non-point sources (LA) and natural backgroundwith a margin of safety, MOS (USEPA, 1999). It is describedas TMDL =

PWLA +

PLA + MOS. Here, the

PWLA is identi-

fied as wastewater impact from the To Lich River. ThePLA is equal to total loadings of the Nhue’s upstream input

and lateral wastewater inflow along the Nhue to the conflu-ence point. In this case study, the Margin of Safety is ex-pressed as two causes: increase of population in the studyarea and the evapotranspiration and infiltration. The formerwas taken as 5% of the total discharge since the populationincrease was about 5% per year (Cu and Cham, 2006) andthis calculation was derived the annual average. The latterwas estimated as 2% of the total discharge based on the an-nual fluctuations of precipitation/evapotranspiration andwater table level in the area. Thus, since the calculationdid not take into account these two factors, the MOS is to-tally taken as 7% of TMDL. Also based on the above analysis,the water quality parameters of major concern are BOD,DO, SS, NH4 and NO3. At average flow, the TMDL, the cur-rent

PWLA, and the current

PLA of the considered param-

eters were calculated and represented in the Table 4. Theresults indicate that BOD, NH4, and partly DO have severelyviolated water quality regulation/consents.

The management method herein focuses on reducing theloadings of NH4 and BOD while maintain reasonable levels ofSS and NO3 (DO is not necessarily considered because if thelevels of NH4 and BOD are low, the level of DO will automat-ically be high). Most BOD and NH4 are derived from WLA (the

water load allocation (P

WLA), and the current load allocation

TMDL (ton/d) Current LA Current WLA

70.1 64.2 62.8

5.6 14.9 0.7

224.23 156.0 38.04.2 1.7 6.3

9.5 1.6 0.2

water.

356 T.A. Duc et al.

To Lich River) and therefore management must focus on re-duce the loadings from WLA. In this case study, the modelwas used to evaluate the applicability of management alter-natives by calculating levels of parameters at defined posi-tion; 5 km downstream the impact zone. The idea behindthe choice of 5 km point is to let the system stabilise aftermixing of two different water masses (Trinh et al., 2006b;McAvoy et al., 2003). Since the loading five km downstreamof the confluence has changed compared with it at the con-fluence due to biogeochemical activities (e.g. NH4 is oxi-dised and reduced or BOD reduces), the model is neededto precisely calculate this loading change at different man-agement alternatives.

Three management alternatives practicable in this studysite are proposed, the rationale for the analysis for thethree approaches were as follows.

1. Treatment of the To Lich wastewater. A waste watertreatment plant (WWTP) was constructed at the ThanhLiet dam to treat the wastewater of the To Lich River’swater before rejecting to the Nhue River. One factorneeds to be established for this management alternativeis the change in amounts of materials and micro-organ-isms before and after treatment by WWTP (treatmentefficiency). Such a change is evaluated as the contentratios between treated and untreated wastewater andshown in Table 5 – the ratios were manipulated fromthe study of Servais et al. (1999). Moreover, becausesewer system in Hanoi collected both wastewater andrainwater, the scenario was assessed under dry and rainyconditions. For the dry period, the treatment volumewas set up as 90% of the total water input. Correspond-ingly, during the rainy period treatment volume wastaken as 50% of the total river water input. To Lich dis-

Table 5 Fractions of contents of nutrients, organic matter and orin the treatment scenario

No. Variable Rat

1 Phytoplankton 1.02 Dissolved degradable organic matter 0.33 Particulate degradable organic matter 0.04 Inert particulate organic matter 0.15 Inert dissolved organic matter 0.96 Heterotrophic bacteria 0.1

Table 6 Water quality indicators at different management alter

Parameters Unit Sim. treatedof TL water dry

Sim. treaTL water

BOD5 mg/l 30.6 25.2DO mg/l 6.1 4.9SS mg/l 59.7 62.4NHþ4 (as N) mg/l 1.4 1.4NO�3 (as N) mg/l 2.9 1.2

Bold: The indicators do not meet the Vietnamese standards for surfac

charges were set as 5.82 and 15 m3/s in dry and rainyconditions, respectively. The seasonal variation is notconsidered here because in this river portion, hydrologyis strictly regulated by human and therefore barelydependent from hydrology of the watershed as a whole.

2. Reduce the discharge of the To Lich water. Based on thecalculation of current load allocations and water qualitystandards (Table 4), the discharge reduction by one thirdwas selected for calculation since it would apparentlyreduce NH4 closely to the allowable loadings.

3. Increase the upstream discharge of the Nhue River (pol-luted water flushing). For this alternative, the upstreamdischarge was set at of 50 (m3/s) – this value is the max-imum possible discharge not causing flooding in thewatershed. The increase of discharge dilutes the pollu-tant contents to meet standards.

Because the water discharges applied for these manage-ment alternatives are different, the simulation results areshown in concentration units instead of loading for an easycomparison (Table 6).

The results of this exercise indicate that the first andsecond methods would lead to similar outcomes since mostparameters met the standards. For the third alternative,the two parameters NH4 and SS did not meet the standards.Therefore, load allocations were computed for the first twomanagement alternatives (three scenarios) and representedin Table 7. For the treatment in dry period, the WLA of NH4

and BOD were equal to only one third and a half of the pres-ent WLA, respectively. For the other two scenarios, afterthe loadings were normalised by discharge for comparingwith present average flow condition, the WLA of NH4 andBOD of these two scenarios were equal or less than a halfof the present WLA. Therefore, in order to meet the water

ganisms between treated and untreated wastewater employed

io No. Variable Ratio

0 7 Nitrifying bacteria 0.650 8 NH4 0.539 9 NO3 33.99 10 PO4 0.994 11 pH 1.003 12 DO 1.00

natives simulated by the ecological model

ted ofrainy

Reduce TL disch. Ups. disch. 50 m3/s

33.8 30.45.7 6.0

58.2 96.7

1.4 1.6

0.7 0.67

e water.

Table 7 Allocations of loadings at acceptable managementalternatives for two parameters NH4 and BOD; MOS is takenimplicitly as 7% of TMDL

Treated indry period

Treated inrainyperiod

DecreaseTo Lich’sflow

BOD NH4 BOD NH4 BOD NH4

TMDL (ton/d) 85.8 3.8 90.5 5.0 83.4 3.4LA (ton/d) 49.8 1.2 44.2 0.5 58.8 1.4WLA (ton/d) 30.0 2.4 39.9 4.2 18.8 1.8


quality standards, the point source loadings of the pollu-tants BOD and NH4 should be diminished by at least twiceof the current loadings, respectively.

Conclusion

This study confirms the heterotrophic state of the Nhue Riv-er and that the pollutant impact of the To Lich River input.A combination of monitoring, laboratory experimentationand model simulation indicate massive respiration/organicmatter breakdown downstream the confluence betweentwo rivers.

The study provides a new analysis of water quality inrelation to dissolved carbon dioxide levels, that has notreally been used in the region previously, but which canbe calculated relatively simply using commonly determinedwater quality parameters (pH and alkalinity). This analysisshows that the waters are greatly oversaturated with re-spect to CO2 and provides strong indication of the massiverespiration/organic matter breakdown and the dominanceof autotrophic processes in the polluted parts of the NhueRiver. It is evident that the abundance of DOC preventedcalcite precipitation within the water column thoughlog10SICalcite were sometimes above the saturation level,leading to no self cleansing process in the system. There-fore, the measurement of EpCO2 using pH and alkalinityanalysis needs to be considered an integral part of riverineenvironmental water quality and biological studies.

In addition, in a non linear hydrodynamic system, thechoice of chlorophyll-a/Phaeophytin ratio for evaluatingthe healthy state of suspended phytoplankton proves to besimple but effective. Remarkably, this evaluation togetherwith pH and alkalinity analyses are easily applied to thecomplex and polluted hydrosystems in developing countrieswhere there are important constraints on urbanisation andindustrialisation while advance analytical techniques aresomehow limited.

The assessment on DOC degradation has identified theimportance of the sediment to removal of pollutants intropical river water but this area of research also deservesfurther attention. They are, for instances, the quantifica-tion of bacterial biomass in water and sediment and the ef-fect of tropical conditions on the degradation rate.

With the help of the ecological model, the pollutant loadreduction requirements were determined according to theTMDLs and allocated to each parameter. The requirementsshow that the current environmental state failed to meetthe TMDL goals for BOD, DO, and NH4. From this evaluation,

the following were concluded and recommended to satisfy/achieve the TMDLs from the current environmental state:

1. The existing runoff and flow regime of the Nhue River canno longer cope with the untreated domestic wastewaterloads of more than 3 million people in Hanoi city today.

2. The organic matter in the To Lich’s wastewater need tobe minimised and a WWTP for To Lich water treatment ishighly recommended.

3. If recommendations 1 and 2 are not used, two third ofthe current To Lich water discharge should be divertedto other water bodies, for instance the Red River, inorder to stop pollution in the Nhue water. This diversionwould not alter the environmental state of the Red Riverbecause the very high discharge in the Red River (average3577 m3/s; Quynh et al., 2005) would greatly dilute ToLich domestic wastewater.

This evaluation implies that the management alterna-tives should be proposed and evaluated based on the waterquality criteria, practical aspects, and a calibrated modelwhich then identify the best management practice (BMPs)among possible alternatives, allowing the TMDL programto be complemented more effectively.

It is concluded that there are straight-forward means ofalleviating the problem and that this is shown by a combina-tion of measurement (with hydrochemical interpretationsuch as end-member mixing analysis), laboratory experi-mentation of rates of change in pollution with contaminatedsediment loading and modelling.

Acknowledgements

I gratefully acknowledge Prof. Peter Reichert for someexplanations on the technique described in Reichert(2001). Gratitude is sent to Prof. Colin Neal for highlightingsome points in composing this article. The work presentedin this paper was supported by the Vietnamese Academyof Science and Technology (VAST, Vietnam) and Centre Na-tional de Recherche Scientifique (CNRS, France).

References

Block, J.C., Mathieu, L., Servais, P., Fontvieille, D., Werner, P.,1992. Indigenous bacterial inocula for measuring the biodegrad-able dissolved organic carbon (BDOC) in waters. Wat. Res. 26(4), 481–486.

Bolay, J.C., Cartoux, S., Cunha, A., Thai, T.N.D., Bassand, M.,1997. Sustainable development and urban growth: Precarioushabitat and water management in Ho Chi Minh City, Vietnam.Habitat Int. 21 (2), 185–197.

Brion, N., Billen, G., 2000. Wastewater as a source of nitrifyingbacteria in river systems: the case of the river Seine downstreamfrom Paris. Wat. Res. 34 (12), 3213–3221.

Clesceri, L.S., Greenberg, A.E., Eaton, A.D., 1999. StandardMethods for the Examination of Water and Wastewater. (20thed.) ESBN 0-87553-235-7. APAH 1015 15th Street, NW Washing-ton, DC 20005.

Cu, N.V., Cham, D.D., 2006. Planning and implementation of theNhue-Day River survey (In Vietnamese). Final report, Ministry ofNatural Resources and Environment.

358 T.A. Duc et al.

Frias, J., Ribas, F., Lucena, F., 1992. A method for the measure-ment of biodegradable organic carbon in waters. Wat. Res. 26(2), 255–258.

Garnier, J., Billen, G., Palfner, L., 2000. Understanding the oxygenbudget and related ecological processes in the river Mosel: theRIVERSTRAHLER approach. Hydrobiologia 410, 151–166.

Hartley, A.M., House, W.A., Leadbeater, B.S.C., Callow, M.E.,1996. The use of micro-electrodes to study the precipitation ofcalcite upon algal biofilms. J. Colloid Interface Sci. 183, 498–505.

House, W.A., 1989. Kinetics of crystallisation of solids from aqueoussolutionsComprehensive chemical kinetics reactions at theliquid–solid interface, 28. Elsevier, Amsterdam (Chapter 3).

Janssen-Stelder, B.M, Augustinus, P.G.E.F., van Santen, W.A.C.,2002. In: Winterwerp, J.C. and Kranenburg, C., (Eds.), Sedi-mentation in a coastal mangrove system, Red River delta,Vietnam. Fine Sediment Marine Dynamics in the Marine Environ-ment. 455–567.

JICA, 1995. The study on Urban Drainage and Wastewater DisposalSystem in Hanoi city, Final report.

Lorenzen, C.J., 1967. Determination of chlorophyll and pheo-pigments: spectrophotometric equations. Limnol. Oceano. 12,343–346.

Luong, V.C., Yang, Yi., Lin, C.K., 2005. Cove culture of marble goby(oxyeleotris marmorata bleeker) and carps in tri an reservoir ofVietnam. Aquaculture 244, 97–107.

Maberly, S.C., 1996. Diel, episodic and seasonal changes in pH andconcentrations of inorganic carbon in a productive lake. Fresh-water Biol. 35, 579–595.

McAvoy, D.C., Masscheleyn, P., Peng, C., Morrall, S.W., Casilla,A.B., Lim, J.M.U., Gregorio, E.G., 2003. Risk assessmentapproach for untreated wastewater using the QUAL2E waterquality model. Chemosphere 52, 55–66.

NASA, 2003. NASA Langley Research Center Atmospheric SciencesData Center. (<http://eosweb.larc.nasa.gov/>).

Neal, C., House, W.A., Down, K., 1998a. An assessment of excesscarbon dioxide partial pressures in natural waters based on pHand alkalinity measurements. Sci. Total Environ. (210/211),173–185.

Neal, C., House, W.A., Jarvie, H.P., Eatherall, A., 1998b. Thesignificance of dissolved carbon dioxide in major lowland riversentering the North Sea. Sci. Total Environ. (210/211), 187–203.

Neal, C., Harrow, M.L., and Williams, R.J. 1998c. Dissolved carbondioxide and oxygen in the River Thames: Spring-summer 1997.Sci. Tot. Environ. 210/211, 205–218.

Neal, C., 2001a. The potential for phosphorus pollution remediationby calcite precipitation in UK freshwaters. Hydrol. Earth Syst.Sci. 5 (1), 49–58.

Neal, C., 2001b. Alkalinity measurements within natural waters:towards a standardised approach. Sci. Tot. Environ. 265, 110–115.

Neal, C., 2002. Calcite saturation in eastern UK rivers. Sci. Tot.Environ. (282/283), 311–326.

O’Connor, D.J., Dobbins, W.E., 1958. Mechanisms of reaeration innatural streams. Trans. Am. Soc. Civ. Eng. 123, 641–684.

Paresys, G., Rigart, C., Rousseau, B., Wong, A.W.M., Fan, F.,Barbier, J.P., Lavaud, J., 2005. Quantitative and qualitativeevaluation of phytoplankton communities by trichromatic chlo-rophyll fluorescence excitation with special focus on cyanobac-teria. Wat. Res. 39 (5), 911–921.

Pruszak, Z., Pham, V.N., Szmytkiewicz, M., Nguyen, M.H., Ostrow-ski, R., 2005. Hydrology and morphology of two river mouthregions (temperate Vistula Delta and subtropical Red RiverDelta). Oceanologia 47 (3), 365–385.

Quynh, L.T.P., Billen, G., Garnier, J., Thiery, S., Fezard, C., Minh,C.V., 2005. Nutrient (N, P) budgets for the Red River basin(Vietnam and China). Glob. Biogeochem. Cyc. 19,GB2022.

Reichert, P., 1998. AQUASIM 2.0 User Manual. Swiss FederalInstitute for Environmental Science and Technology (EAWAG).

Reichert, P., Borchardt, D., Henze, M., Rauch, W., Shanahan, P.,Somlyody, L., Vanrolleghem, P., 2001. River water quality modelno 1 (RWQM1): II. Biochemical process equations. Wat. Sci.Tech. 43 (5), 11–30.

Servais, P., Garnier, J., Demarteau, N., Brion, N., Billen, G., 1999.Supply of organic matter and bacteria to aquatic ecosystemsthrough waste water effluents. Wat. Res. 33 (16), 3521–3531.

SPSS, 2002. SPSS for windows 11.5. SPSS Inc. Headquarters, 233 S.Wacker Drive, 11th floor, Chicago, Illinois 60606.

Tang, D.L., Kawamura, H., Lee, M.A., Dien, T.V., 2003. Seasonaland spatial distribution of chlorophyll-a concentrations andwater conditions in the Gulf of Tonkin, South China Sea. Rem.Sens. Environ. 85, 475–483.

Trinh, A.D., Akamatsu, Y., Ikeda, S., Le, L.A., Vu, D.L., 2006b.Application of a 2-D numerical model for computing suspendedsediment and nutrients transports to Nhue River (Hanoi, Viet-nam). Ann. J. Hydr. Eng. JSCE 50, 19–24.

Trinh, A.D., Bonnet, M.P., Vachaud, G., Chau, V.M., Prieur, N., Vu,D.L., Le, L.A., 2006a. Biochemical modelling of the Nhue River(Hanoi, Vietnam); practical identifiability analysis and parame-ters estimation. Ecol. Mod. 193 (3–4), 182–204.

Trinh, A.D., Bonnet, M.P., Vachaud, G., Prieur, N., Vu, D.L., Le,L.A., 2005. Investigation of sediment impact to water quality inthe Nhue River (Hanoi, Vietnam) by two prominent variables DOand NH4. Adv. Nat. Sci. (Vietnam) 6 (2), 135–142.

USEPA, 1999. Protocol for Developing Nutrient TMDLs, EPA 841-B-99-007.

USEPA, 2000. Nutrient Criteria Technical Guidance Manual, Riversand Streams. EPA-822-B-00-002 (<http://www.epa.gov/water-science/criteria/nutrient/guidance/rivers/index.html>).

Vannote, R.L., Minshall, G.W., Cummins, K.W., Sedell, J.R.,Cushing, C.E., 1980. The river continuum concept. Can. J. Fish.Aquat. Sci. 37, 130–137.

Vuillemin, M.H.T., Dispan, J., Mouchel, J.M., Servais, P., 2003.Biodegradable fraction of organic carbon estimated under oxicand anoxic conditions. Wat. Res. 37, 2242–2247.

Wanninkhof, R., 1992. Relationship between wind speed and gasexchange over ocean. J. Geophys. Res. 97, 7373–7382.


http://www.epa.gov/waterscience/criteria/nutrient/guidance/rivers/index.html

http://www.epa.gov/waterscience/criteria/nutrient/guidance/rivers/index.html

Experimental investigation and modelling approach of the impact of urban wastewater on a tropical river; a case study of the Nhue River, Hanoi, Viet Nam

Documents