**3.3 Water quality**

58 Current Issues of Water Management

**50.3 44.7 <sup>5</sup>**

Fig. 14. Estimated average rivers flows and simulated values for a scenario corresponding to

Alto 7 Rabagão

Tojal

**20**

CÁVADO

HOMEM

Paradela

4

3 Beredo

Location

6

**Vilarinho das Furnas**

FT\_ETAR52 FT\_ETAR36

Caniçada

**16**

**19**

FT\_ETAR37

Alto 2 Cávado

RIO CÁVADO

**3.2 3.2 <sup>11</sup>**

**5.5 6.1 <sup>12</sup>**

8

**21**

9

Borralha

Rabagão

Ponte do Bico

FT\_ETAR29

**17**

FT\_ETAR38 FT\_ETAR13

Borralha

10

5 Cavadas

**61.3 61.4 <sup>7</sup>**

the annual mean values of the available monitoring period

Venda Nova

FT\_ETAR11

Febras Pontes

Penide Ruães

**22**

FT\_ETAR28 FT\_ETAR25

Fig. 15. River Cávado model scheme: WWTP discharges

**23**

FT\_ETAR23 FT\_ETAR50 FT\_ETAR51

**64.0 64.2 <sup>8</sup>**

Cabril

11

**56.4 50.8 <sup>6</sup>**

**68.6 68.7 <sup>9</sup>**

Cabreira

13

12

**25**

FT\_ETAR5

Covo

**27**

FT\_ETAR40

**24**

FT\_ETAR7 FT\_ETAR41

FT\_ETAR8

Milhazes

FT\_ETAR27

**26**

Salamonde

Toco

Caveiro

FT\_ETAR9

FT\_ETAR45 FT\_ETAR44 FT\_ETAR43

FT\_ETAR46

**28**

FT\_ETAR24

14

**5.5 5.5 <sup>1</sup>**

**Valor estimado (m<sup>3</sup>**

Estimated value (m3/s) Simulated value (m3/s)

**Valor simulado (m<sup>3</sup> /s) Trecho**

**/s)**

1

FT\_ETAR39

Gerês **15**

**18**

FT\_ETAR31 FT\_ETAR33 FT\_ETAR34 FT\_ETAR35 FT\_ETAR32 FT\_ETAR30

**12.6 12.6 <sup>2</sup> 29.0 28.9 <sup>3</sup>**

**36.4 36.5 <sup>4</sup>**

**10.7 10.7 <sup>10</sup>** Simulations of water quality were based on hydrodynamic simulations previously presented and on the consideration of different characteristics for the discharges of waste water treatment plants (WWTP), industries, livestock units, and other contributions associated with different tributaries. The establishment of different values for these pollutant discharges results in different simulation scenarios.

In order to present the potential of the developed modelling system different scenarios were defined for average annual and monthly rivers flows discharges and considering the WWTP efficiency in compliance with their treatment schemes (scenarios 1 to 13). Additionally, for the wettest month (January) and the driest month (August) the effects of discharges from WWTP considering extreme values of its efficiencies were worked out (scenarios 14 to 17). Finally, it was considered three scenarios (scenarios 18 to 20) associated with the failure of each one of the three main WWTP: Frossos, Esposende and Vila Frescaínha (Table 1).


Table 1. Water quality modelling scenarios

Web-Based Decision Support Framework for

**189 <sup>47</sup> <sup>10</sup>**

**904 <sup>464</sup> <sup>10</sup>**

0.9 0.3 <sup>10</sup>

0.9 0.5 <sup>11</sup>

1.4 0.8 <sup>2</sup>

1.2 0.9 <sup>3</sup>

**315 <sup>154</sup> <sup>11</sup>**

**5894 <sup>771</sup> <sup>2</sup>**

**1609 <sup>1148</sup> <sup>3</sup>**

**55 <sup>17</sup> <sup>11</sup>**

**684 <sup>85</sup> <sup>1</sup>**

**2837 <sup>800</sup> <sup>1</sup>**

FCB (MPN/100 mL) Scenario 1

1.3 0.8 <sup>1</sup>

BOD (mg/L) Scenario 1 <sup>5</sup>

**970 <sup>84</sup> <sup>2</sup>**

**490 <sup>125</sup> <sup>3</sup>**

**107 <sup>26</sup> <sup>4</sup>**

Fig. 16. Water quality model results in the vicinity of monitoring stations for scenario 1

**21**

**726 <sup>91</sup> <sup>4</sup>**

**14 <sup>14</sup> <sup>12</sup>**

1.0 1.1 <sup>4</sup>

0.8 0.8 <sup>12</sup>

**<sup>52</sup> <sup>9</sup>** SB (MPN/100 mL) Scenario 1

**238 <sup>28</sup> <sup>5</sup>**

**367 <sup>61</sup> <sup>5</sup>**

0.9 1.2 <sup>5</sup>

1.2 1.2 <sup>6</sup>

**84 <sup>79</sup> <sup>6</sup>**

**39 <sup>36</sup> <sup>6</sup>**

**<sup>20</sup> <sup>12</sup> <sup>19</sup>**

**38 <sup>34</sup> <sup>7</sup>**

**44 <sup>39</sup> <sup>7</sup>**

1.0 1.0 <sup>7</sup>

> **32 <sup>32</sup> <sup>8</sup>**

0.9 0.9 <sup>8</sup>

**<sup>18</sup> <sup>8</sup>**

**54**

**59 <sup>51</sup> <sup>9</sup>**

1.3 1.3 <sup>9</sup>

Simulated value Estimated value Location

Water Resources Management at River Basin Scale 61

The qualitative characteristics of the pollutant discharges from WWTP were estimated for the following parameters: biochemical oxygen demand (BOD5), dissolved oxygen (DO), and total coliform bacteria (TCB), faecal coliform bacteria (FCB) and streptococci bacteria (SB).

Figure 15 shows the approximate location of all wastewater treatment plant specified in the water quality model.

The loads associated with industrial pollutant discharges were estimated. It was used a simple method based on the industrial activities to compute its waste water discharges. However, most of the industries effluents are treated in WWTP.

Another pollutant sources estimated for each simulation was the livestock farms discharges. The effluent loads were estimated considering the number of heads of cattle per farm.

In the upstream open boundary of the modelled rivers it was considered that the water presents characteristics of unpolluted water, with zero values of concentrations of BOD5 and bacteriological variables. For dissolved oxygen, it was considered a value close to the saturation concentration (10 mg/L) and a water temperature of 10 °C.

The initial conditions of the model were established by considering the values of average concentrations at the monitoring stations for each water quality variable.

Results were obtained for the twenty scenarios, using a simulation time of seven days for each scenario.

It should be stressed that the results prove to be more sensitive to the values adopted for the concentrations of wastewater discharges than to the calibration coefficients. Once that data series related to those discharges are available it is possible to improve the performance of the water quality model through the fine tune of the calibration coefficients.

Table 2 presents the obtained simulated results for scenario 1 at the last simulation time and the average values monitored at monitoring stations. It appears that for the rivers Homem and Cávado (where monitoring data is available) the water quality profile related with organic matter discharges, inferred from the results of concentrations of BOD5, presents nearly uniform values (around 1 mg/L). It was possible to achieve good approximation between the monitored values and model results for this scenario.

The installed treatment capacity of waste water (mainly from domestic sources) in the river basin, based on the removal of organic matter is reflected in the monitored values, resulting in low concentrations of BOD5.

Regarding the results of bacteriological variables the differences between simulated and observed values in scenario 1, are more significant in the regions located in the most urbanized areas.

The differences between measured and simulated values arise primarily from the uncertainty associated with quantification of bacterial loads (a sensitivity analysis was performed concerning adopted calibration parameters, and it was concluded that these parameters cannot justify the differences found). The values used in the estimation of these loads (mean values of bacterial loads) are highly variable and may justify the simulated behaviour (underestimation of bacterial loads).

The qualitative characteristics of the pollutant discharges from WWTP were estimated for the following parameters: biochemical oxygen demand (BOD5), dissolved oxygen (DO), and total coliform bacteria (TCB), faecal coliform bacteria (FCB) and streptococci bacteria (SB). Figure 15 shows the approximate location of all wastewater treatment plant specified in the

The loads associated with industrial pollutant discharges were estimated. It was used a simple method based on the industrial activities to compute its waste water discharges.

Another pollutant sources estimated for each simulation was the livestock farms discharges. The effluent loads were estimated considering the number of heads of cattle per farm.

In the upstream open boundary of the modelled rivers it was considered that the water presents characteristics of unpolluted water, with zero values of concentrations of BOD5 and bacteriological variables. For dissolved oxygen, it was considered a value close to the

The initial conditions of the model were established by considering the values of average

Results were obtained for the twenty scenarios, using a simulation time of seven days for

It should be stressed that the results prove to be more sensitive to the values adopted for the concentrations of wastewater discharges than to the calibration coefficients. Once that data series related to those discharges are available it is possible to improve the performance of

Table 2 presents the obtained simulated results for scenario 1 at the last simulation time and the average values monitored at monitoring stations. It appears that for the rivers Homem and Cávado (where monitoring data is available) the water quality profile related with organic matter discharges, inferred from the results of concentrations of BOD5, presents nearly uniform values (around 1 mg/L). It was possible to achieve good approximation

The installed treatment capacity of waste water (mainly from domestic sources) in the river basin, based on the removal of organic matter is reflected in the monitored values, resulting

Regarding the results of bacteriological variables the differences between simulated and observed values in scenario 1, are more significant in the regions located in the most

The differences between measured and simulated values arise primarily from the uncertainty associated with quantification of bacterial loads (a sensitivity analysis was performed concerning adopted calibration parameters, and it was concluded that these parameters cannot justify the differences found). The values used in the estimation of these loads (mean values of bacterial loads) are highly variable and may justify the simulated

However, most of the industries effluents are treated in WWTP.

saturation concentration (10 mg/L) and a water temperature of 10 °C.

concentrations at the monitoring stations for each water quality variable.

the water quality model through the fine tune of the calibration coefficients.

between the monitored values and model results for this scenario.

water quality model.

each scenario.

in low concentrations of BOD5.

behaviour (underestimation of bacterial loads).

urbanized areas.

Fig. 16. Water quality model results in the vicinity of monitoring stations for scenario 1

Web-Based Decision Support Framework for

river Cávado: Ponte Nova Barcelos.

in the dry season, for all simulated variables.

worst situation in terms of impact on water quality.

diffuse sources.

Barcelos.

waters.

Water Resources Management at River Basin Scale 63

Results for this scenario showed that the bacteriological pollution in the river Cávado stretch downstream the confluence of the river Homem it is a reality that the wastewater treatment solutions have failed to solve in this basin. The bacteriological pollution also occurs during the dry season which decreases the likelihood of contamination come from

Adopting the same conditions of pollutant discharges of scenario 1 but now considering the average monthly rivers flows discharges, with the simulated scenarios 2 to 13 it is possible to assess the influence of seasonal flow variation on water quality behaviour. Figure 17 shows results in these scenarios in a specific location of the river Cávado: Ponte Nova de

The results reveal a strong influence of the seasonal river flow regime variation on water quality concentrations. The variations reach values of around 70% of the concentration values in the wettest months in the case of bacteriological variables and about 40% for BOD5. Identical results can be achieved with simulations involving different time scales with the ODeCav System. It is particularly interesting to evaluate the influence of hydropower generation plants operational rules on water quality for the rivers Cávado and Homem, since for these two rivers the flow regime is strongly influenced by the operations of those hydraulic structures. Also the influence of hourly variations in flow regimes resulting from energy production on water quality of rivers Cávado and Homem, can be easily assessed once the pollutant discharges in the rivers are known with a similar temporal resolution.

With the scenarios 14 to 17 it is possible to evaluate the impact of different performances of the WWTP assuming different river flow regimes (January and August) in the resultant water quality. Figure 18 depict the obtained results in these scenarios in a location of the

The results show a greater sensitivity of the river receiving waters to wastewater discharges

Although the efficiencies of the WWTP considered in each of the scenarios have been defined from literature values, there is a great variability in the resulting receiving waters concentrations, especially for bacteriological variables. For the presented location, the concentrations of bacterial variables when treatment plants operate with a minimum efficiency it is approximately four times the concentration when treatment plants operate at maximum efficiency. The developed modelling system reveal to be very important for the evaluation of alternative wastewater treatment investments since it allows, in a simple way, to anticipate the effects of these structures to improve water quality in river receiving

Finally, it is presented in Figure 19 results for the Frossos WWTP rupture scenario (scenario 18). The rupture is simulated for a situation of average runoff and therefore not reflect the

The presented results refer to two distinct locations: one in the vicinity of the discharge (immediately downstream) of the WWTP and the other on a stretch away. Based on this kind of results protective measures can be planned and the affected water uses can be

anticipated, minimizing the impact of this potential accidental situation.

Fig. 17. Concentrations results for BOD5 and FCB for scenarios 2 to 13 at Ponte Nova de Barcelos in river Cávado

Fig. 18. Concentrations results for BOD5 and FCB under scenarios 14 to 17 at Ponte Nova Barcelos in river Cávado

**JAN FEV MAR ABR MAI JUN JUL AGO SET OUT NOV DEZ**

**JAN FEV MAR ABR MAI JUN JUL AGO SET OUT NOV DEZ**

Fig. 17. Concentrations results for BOD5 and FCB for scenarios 2 to 13 at Ponte Nova de

Coliformes totais Caudal

FEB APR MAY AUG SEP OCT DECFCB Discharge

**Janeiro Agosto**

**January August**

Fig. 18. Concentrations results for BOD5 and FCB under scenarios 14 to 17 at Ponte Nova

**minimum average maximum WWTP efficiency:**

BOD5

CBO5 Caudal

Discharge

**Discharge (m3/s)** 

AUG SEP OCT DEC

**Discharge (m3/s)** 

**0.0 0.2 0.4 0.6 0.8 1.0 1.2**

**0.0**

**0.0 0.2 0.4 0.6 0.8 1.0 1.2**

**BOD5 (mg/L)** 

**(**

**FCB (MPN/100 mL)** 

 **)**

**500.0**

**1000.0**

**1500.0**

**FCB (MPN/100 mL)** 

Barcelos in river Cávado

Barcelos in river Cávado

**2000.0**

**2500.0**

FEB APR

**BOD5 (mg/L)** 

Results for this scenario showed that the bacteriological pollution in the river Cávado stretch downstream the confluence of the river Homem it is a reality that the wastewater treatment solutions have failed to solve in this basin. The bacteriological pollution also occurs during the dry season which decreases the likelihood of contamination come from diffuse sources.

Adopting the same conditions of pollutant discharges of scenario 1 but now considering the average monthly rivers flows discharges, with the simulated scenarios 2 to 13 it is possible to assess the influence of seasonal flow variation on water quality behaviour. Figure 17 shows results in these scenarios in a specific location of the river Cávado: Ponte Nova de Barcelos.

The results reveal a strong influence of the seasonal river flow regime variation on water quality concentrations. The variations reach values of around 70% of the concentration values in the wettest months in the case of bacteriological variables and about 40% for BOD5.

Identical results can be achieved with simulations involving different time scales with the ODeCav System. It is particularly interesting to evaluate the influence of hydropower generation plants operational rules on water quality for the rivers Cávado and Homem, since for these two rivers the flow regime is strongly influenced by the operations of those hydraulic structures. Also the influence of hourly variations in flow regimes resulting from energy production on water quality of rivers Cávado and Homem, can be easily assessed once the pollutant discharges in the rivers are known with a similar temporal resolution.

With the scenarios 14 to 17 it is possible to evaluate the impact of different performances of the WWTP assuming different river flow regimes (January and August) in the resultant water quality. Figure 18 depict the obtained results in these scenarios in a location of the river Cávado: Ponte Nova Barcelos.

The results show a greater sensitivity of the river receiving waters to wastewater discharges in the dry season, for all simulated variables.

Although the efficiencies of the WWTP considered in each of the scenarios have been defined from literature values, there is a great variability in the resulting receiving waters concentrations, especially for bacteriological variables. For the presented location, the concentrations of bacterial variables when treatment plants operate with a minimum efficiency it is approximately four times the concentration when treatment plants operate at maximum efficiency. The developed modelling system reveal to be very important for the evaluation of alternative wastewater treatment investments since it allows, in a simple way, to anticipate the effects of these structures to improve water quality in river receiving waters.

Finally, it is presented in Figure 19 results for the Frossos WWTP rupture scenario (scenario 18). The rupture is simulated for a situation of average runoff and therefore not reflect the worst situation in terms of impact on water quality.

The presented results refer to two distinct locations: one in the vicinity of the discharge (immediately downstream) of the WWTP and the other on a stretch away. Based on this kind of results protective measures can be planned and the affected water uses can be anticipated, minimizing the impact of this potential accidental situation.

Web-Based Decision Support Framework for

**5. Acknowledgment** 

**6. References** 

Water Resources Management at River Basin Scale 65

years, the use of all the potentialities of this kind of platforms in practical situations under different water management problems constitutes a major challenge for its evaluation. Moreover, water authorities once decide to use this kind of management tools will certainly see improved their analysis capabilities, strengthening their technological skills for the adoption of more sustainable water management policies. Especially it is adequate for developing big projects since it facilitates collaborative studies in one common platform and

The authors thank to Comissão de Coordenação e Desenvolvimento Regional do Norte,

Berlekamp, J., Lautenbach, S., Graf, N., Reimer, S. & Matthies, M. 2007 Integration of

Borowski, I. & Hare, M. 2007 Exploring the gap between water managers and researchers.

De Kok, J. L., Kofalk, S., Berlekamp, J., Hahn, B. M. & Wind H. 2008 From Design to

European Commission 2000 Directive of the European Parliament and of the Council

Price, R. K. 2000 Hydroinformatics and urban drainage: an agenda for the beginning of the

Ravesteijn, W. & Kroesen O. 2007 Tensions in water management; Dutch tradition and

Rekolainen, S., Kämäri J. & Hiltunen M. 2004 A conceptual framework for identifying the

Thomann R. V., and J. A. Mueller. 1987. Principles of Surface Water Quality Modeling and

need and role of models in the implementation of the Water Framework Directive.

Chapra, S. 1997. Surface water-quality modeling. The McGraw-Hill Companies, Inc.

DELTARES 2009 SOBEK software. Available from:< http://delftsoftware.wldelft.nl/>. Dudley, J., Daniels, W., Gijsbers, P. J. A., Fortune, D., Westen, S. & Gregersen, J. B. 2005

MYSQL 2009 MySQL database server. Available from:<http://www.mysql.com/>.

21st century, Journal of Hydroinformatics, vol. 2, No. 2, 133-147.

European policy. Water Science and Technology 56(4): 105–111.

International Journal of River Basin Management 1(4):1–6.

Control. Harper and Row, Inc., New York.

MONERIS and GREAT-ER in the decision-support system for the Elbe river basin.

Difficulties of model-based tools to support practical water management. Water

Application of a Decision-support System for Integrated River-basin Management,

Applying the Open Modelling Interface (OpenMI), Proceedings of the MODSIM

2000/60/EC Establishing a Framework for Community Action in the Field of Water Policy, Official Journal 2000 L 327/1, European Commission, Brussels. Horn, A., Rueda, F.J., Hörmann, G. & Fohrer, N. 2004 Implementing river water quality

modelling issues in mesoscale watershed models for water policy demands — an overview on current concepts, deficits, and future tasks. Physics and Chemistry of

the modelling results became much more transparent for all project partners.

Águas do Ave, SA and Águas do Cávado, SA, for the financial support.

Environ Model Softw22(2):239–247.

Resources Management 21 (7), 1049-1074.

2005 conference, Melbourne, Australia.

the Earth 29, 725–737.

Water Resources Management, (23), 1781-1811.

Fig. 19. Concentrations results of BOD5 and FCB for the scenario 18 (rupture of Frossos WWTP) at two different locations

It should be noted that the results presented in this chapter are not exhaustive. All simulations are made available with the installation of ODeCav System and can be found on the web based platform.

The presented examples illustrate the potential of the developed system for management of water quality at river Cávado basin.
