Volume 15, numéro 2, 2002

FR :

Cet article traite des problèmes liés à l'étude des équilibres calco-carboniques en limitant leur présentation à l'aspect purement graphique. Dans cette optique, on a développé un nouvel abaque original en choisissant comme coordonnées pH et [CO₃ total].

La construction d'un tel abaque ne fait appel qu'à 2 lois fondamentales, à savoir : a) la relation du produit de solubilité du carbonate calcique :

K_s=Y²₂[Ca²⁺[CO^2-₃]

Cette dernière peut, moyenant l'introduction d'une variable auxiliaire :

∆=[Ca²⁺]-[CO₃ total]

se transformer en une relation équivalente plus commode pour le calcul des courbes constitutives de l'abaque.

b) une deuxième relation, prenant en compte les ions H+ soit libres, soit engagés dans des combinaisons avec le groupe carbonate ou avec le solvant H₂ O. Dans ce but, on introduit une concentration totale appelée acidité potentielle totale (en abrégé APT) définie de la sorte :

APT=2[H₂CO₃]+[HCO^-₃]+[H₂O]+[H⁺]

En outre, on utilise une variable dérivée de APT à laquelle on a donné le nom d'acidité potentielle carbonatée et le symbole. Sa définition est la suivante :

δ=2[H₂CO₃]+[HCO^-₃]

Les 2 variables et APT jouissent en commun de la propriété intéressante de rester invariantes lors de la mise en équilibre de l'eau selon l'équation :

CaCO₃+H⁺⇌Ca²⁺+HCO^-₃

La variable peut également être considérée comme pratiquement invariante dans le domaine des eaux naturelles.

Cette propriété est mise à profit pour construire 2 réseaux de courbes constitutives de l'abaque, à savoir :

a) celles satisfaisant à la condition : APT=constante.

Ces dernières seront appelées "courbes d'iso-acidité

b) celles répondant à la condition : ∆=constante.

On les désignera sous le nom de "courbes d'iso-équilibre"

L'article s'attache à décrire le mode d'utilisation de cet abaque tout en mettant en valeur sa facilité d'emploi pour l'évaluation, tant qualitative que quantitative du comportement d'une eau à l'égard du carbonate calcique.

En particulier, la technique permet de visualiser les caractéristiques essentielles de l'eau telles que pH équilibrant, pH stabilisant et indice de stabilisation, caractéristiques dont la signification est précisée dans l'article.

L'abaque en question, en parallèle avec un logiciel qui peut lui être associé, constitue ainsi un outil précieux tant pour caractériser une eau que pour procéder ensuite aux traitements chimiques les plus connus.

Le cas du mélange d'eaux peut également être avantageusement traité par le biais de cet abaque. De nombreux exemples d'application aux traitements de l'eau sont également largement développés dans la publication. Ceux-ci devraient aider grandement ceux qui sont confrontés aux divers problèmes de l'industrie de l'eau.

L'analyse graphique constitue par ailleurs une excellente introduction à la gestion informatisée plus complète de ces problèmes et dont le traitement ne pouvait trouver place dans le cadre limité de cette publication.

Le lecteur désireux de poursuivre ces développements ou d'avoir accès au programme informatique complémentaire à cette étude, trouvera les indications nécessaires dans les références bibliographiques figurant en fin de l'article.

EN :

Although many good methods exist for computing calcium-carbonate equilibria, we considered that it was still possible to improve the representation to the problem and for this reason we have developed a new method of calculation. In this paper we have restricted ourselves to the graphical aspect of the question, referring the reader to the original bibliography at the end of the paper and as well we have not taken into account the kinetic aspects of the reactions involving the dissolution or precipitation of CaCO₃.

From this perspective, an original graph has been designed choosing for co-ordinates the pH (abscissa) and [CO₃ total] (ordinate). The construction of such a graph requires only two fundamental laws, i.e.

1) The solubility product relationship of calcium carbonate. This relationship, with the introduction of an auxiliary variable:

∆=[Ca²⁺] - [CO₃ total]

can be transformed itself into an équivalent relationship, more convenient for the computation of the various curves composing the graph.

2) A second relationship, taking into account the H+-ions in their free or bound state (with the C0₃ group or the H₂0 solvent). For this purpose, one introduces a total concentration, called potential total acidity (in abbreviated form PTA), the définition of which is:

PTA=2[H₂CO₃]+[HCO^-₃]+[H₂O]+[H⁺]

In addition, one uses also another variable derived from PTA to which we gave the name : carbonated potential acidity with the following definition:

δ=2[H₂CO₃]+[HCO^-₃]

Both variables PTA and ∆ have jointly the interesting property of remaining constant during the evolution of the water towards equilibrium according to the equation:

CaCO₃+H⁺⇌Ca²⁺+HCO^-₃

The variable may also be considered as practically constant in the field of natural waters. We have taken advantage of this property to build up two networks of curves that make up the overall graph, i.e., a) the curves satisfying the condition : PTA=constant (these curves are named " iso-acidity curves "), and b) the curves whose characteristics are : =constant, and which are named " iso-equilibrium curves ".

The paper describes how to use the graph and brings out its ease of use for both the qualitative and quantitative prediction of the behaviour of water in contact with calcium carbonate. In particular, the graph illustrates such essential water characteristics as equilibrium pH, stabilisation pH and stabilisation index, characteristics whose meanings are defined in the paper. The graph is shown to be a useful tool for the characterisation of waters as well as for the evaluation of most chemical treatments. The case of the mixture of different waters is also treated. Numerous examples for using the graph under different circumstances are developed in the paper, which should be useful for those whose job is connected with industrial water treatment. The graphical method here described, in conjunction with an associated computer programme, also constitutes an excellent introduction to the more complete computing methods developed elsewhere but which could not be treated in the limited frame of this publication.

The graph turns out to be a useful tool for the characterisation of waters as well as for the evaluation of different chemical treatments, such as softening, carbonation, decarbonation, acidification, alkalinisation, aeration, etc. The case of the mixture of different waters is also treated. Numerous examples for using the graph under different circumstances are developed. The paper shows how it is possible, on the basis of the diagram, to predict the evolution of the water when different reagents are added, when the temperature changes or when one mixes different waters. The diagram should be a useful aid for those people whose job is connected with industrial water treatment.

A computing program is necessary to draw the curves, to obtain the water characteristics with a better accuracy than can be read from the graphs. This programme, named Carbeq+, is commercially available (see below for the bibliographical reference). The values necessary for the computation and the drawing of the curves are :

1. temperature of analysis and temperature of computation

2. pH, alkalinity, calcium concentration

3. the ionic strength, which can be calculated either from the knowledge of the complete analysis of the water, or if that one is not available, by using an empirical relation based on either the conductivity or the salt content.

The program computes then the equilibrium pH, the stabilisation pH and the stabilisation index and draws the graphs described in the paper. The graphical method here described, in conjunction with the associated computer programme, constitutes also an excellent introduction to the more complete computing methods developed elsewhere but which could not be treated in the limited frame of this publication. Note that the case of the influence of the gaseous carbon dioxide phase on the equilibrium is not treated in this paper, but it is taken into account in the Carbeq+ programme mentioned above.

FR :

Dans les périmètres irrigués du Maroc, la pollution nitrique diffuse des eaux souterraines augmente le risque de détérioration de la qualité des ressources hydriques et engendre un risque sanitaire pour la population rurale s'approvisionnant le plus souvent directement de l'aquifère. La présente étude, réalisée dans le périmètre du Tadla (3600 km²), située au centre du Maroc, vise à faire l'état des lieux de la contamination des eaux souterraines par les nitrates et à identifier les causes potentielles de cette pollution.

Le suivi de la concentration des eaux souterraines en nitrates a été réalisé durant la période août 1996 - avril 1998 dans un réseau de cent puits. Les paramètres du milieu physique et des pratiques culturales ont été déterminés soit par des mesures au champs, des analyses au laboratoire ou par enquête auprès des agriculteurs dans un réseau de 40 sites. Les résultats sont présentés sous forme de cartes thématiques par le SIG.

Le niveau actuel de la pollution nitrique des eaux souterraines dans le périmètre irrigué du Tadla commence à devenir critique. De part la pollution ponctuelle mise en évidence aux alentours de quelques agglomérations urbaines et à l'aval des sucreries, les zones dont les teneurs en nitrates des eaux souterraines dépassent 50 mg NO_3- /l sont localisées en aval des Béni Amir et en aval et à l'Est des Béni Moussa Est. Béni Moussa Ouest reste en général moins polluée que Béni Moussa Est. Les apports excessifs en eau et en engrais azotés pour les cultures pratiquées (blé, betterave, cultures maraîchères, arboriculture et luzerne), sont les causes directes de la pollution diffuse. Par ailleurs, il est certain que la texture du sol (argileuse, argilo-limoneuse et limono-argileuse), la teneur en argile (29% à 55%) et tous les paramètres qui leurs sont corrélés notamment la réserve utile du sol, réduisent le risque de lixiviation des nitrates en profondeur et assurent en conséquence un contrôle naturel de la pollution nitrique des eaux souterraines. Néanmoins, tous les autres facteurs étudiés notamment le niveau des nappes faible par rapport au sol (<4m-12m), la vitesse d'infiltration du sol en milieu saturé généralement élevée (20-450 mm/h), les teneurs assez importantes du sol en matière organique (1.5%-3.5%) et en azote total (0.10%-0.15%), joueraient positivement sur la vulnérabilité de la nappe à cette pollution. De même, le pouvoir de minéralisation élevé des sols couplé aux conditions écologiques favorables à une minéralisation intense de la matière organique native du sol, assez caractéristique des régions irriguées à climat méditerranéen, seraient en partie responsable de cette pollution. Néanmoins, Aucune corrélation n'a été mise en évidence entre la teneur en nitrate des eaux souterraines et les combinaisons linéaires des variables du milieu physique. Ceci serait dû à l'homogénéité du milieu physique et à l'accumulation de la pollution suivant le sens d'écoulement des nappes de l'amont vers l'aval hydraulique. Vu le niveau d'intensification agricole que connaît cette région, la situation risquerait de s'aggraver rapidement induisant, en dehors de toute mesure de prévention, la détérioration de la qualité des eaux souterraines dans cette région.

EN :

In Morocco, diffuse ground water pollution by nitrates in irrigated areas has caused an increase in the risk of water quality deterioration. This has generated a health risk in rural areas, since most of the rural population get their drinking water supply from the aquifer. The present study was carried out in the Tadla region. The plain of Tadla is situated in the centre of Morocco, in the Oum Erbia river basin. It covers a surface area of 3600 km². The Oum Erbia river (160 km long) divides this plain into two large irrigated areas (Beni Amir in the east (3300 ha) and Beni Moussa in the south (6900 ha)), with different hydrological characteristics. The irrigation water comes from the Oum Erbia river, the Bin El Ouidane dam and from two important waters tables: Beni Amir (190 Mm³) and Beni Moussa (250 Mm³). The objectives of this study were to quantify the degree of ground water contamination by nitrates and to identify the potential causes of this pollution.

The monitoring of nitrate concentrations in ground water was carried out during the period of August 1996 to April 1998, in 100 wells in the region of study. The physical environment and agricultural practices were determined either by field measurements, laboratory analyses or by investigations with farmers at 40 selected sites. The results are presented by thematic maps using a geographical information system.

The level of ground water nitrate pollution is becoming a serious problem. The point pollution is localized around or downstream of urban built up areas such as Fquih Ben Saleh, Beni Mellal, Souk Sebt and Oulad Ayad. This pollution, caused by household wastes, is likely aggravated by waste dumping from the sugar refineries (SUBM, SUTA and SUNAT) as well as by waste water reuse in agriculture. Other zones with nitrate concentrations more than 50 mg NO₃ -/L were observed in wells situated downstream from Beni Amir and east of Beni Moussa. West Beni Moussa was generally not as polluted as east Beni Moussa, but an increase in ground water nitrate levels was observed in some downstream wells in west Beni Moussa. The diffuse nitrate pollution has been attributed to land use, irrigation and fertilization practices. These practices may be responsible for nitrate leaching beyond the root zone during the cropping season and may generate large quantities of residual inorganic nitrogen at harvest time. This represents a potential risk of nitrate ground water pollution in the next draining period. The physical characteristics of soil texture (clay, clay-silt and silty clay), percentage clay (29 % - 55 %) and their correlated parameters (especially the useful storage capacity) act to reduce the risk of nitrate leaching and ensures a control on ground water nitrate pollution. However, other observations including: a low water level table compared to the soil surface, particularly in the middle of the irrigated area and around the edge of the Oum Erbia river (<4 - 12m); a high soil infiltration rate in unsaturated soil, which varied from 20 -100 mm/h (permeable soils) to 200 - 450 mm/h (extremely permeable soils); a high level of organic matter in the soil (1.5% - 3.5%) and; total nitrogen contents (0.10% - 0.15%) would increase the aquifer vulnerability to ground water nitrate pollution. Similarly, the high soil mineralization capacity in the 0-60 cm soil layer (100 - 600 kg N /ha), combined with the ecological conditions, favoured an intense mineralization of the native organic matter of soil. This high mineralization capacity, which is characteristic of irrigated Mediterranean areas, could be partly responsible for this pollution. However, no correlation was found between ground water nitrate contents and linear combinations of physical environmental variables. This could be due to the homogeneity of the physical environment and to the movement of water in the saturated soil, in the direction of ground water flow, inducing the accumulation of nitrate pollution in the hydraulic downstream area of the water table.

Considering the degree of agriculture in this area, the situation could worsen rapidly, thus inducing significant ground water deterioration. In attempts to avoid this situation, many important measures at the regional scale are necessary, in order to rationalize agricultural practices and to increase the farmers' awareness of ground water nitrate pollution. These urgent preventative actions should be adopted in order to combine agricultural productivity and durability with water economy and water quality preservation. Some programs are in progress within this region. The objectives are essentially to reduce nitrate excess and to introduce more efficient irrigation techniques such as the pivot (frontal ramp) recommended for cereals and sugar beet, and localized irrigation system proposed for vegetable crops and arboriculture. Also, since the Oum Erbia river basin has also been subjected to a clean-up of urban and industrial pollution, it would be interesting to carry out a more precise study of nitrogen dynamics and nitrate leaching at the plot scale, especially for the most common practiced crops in the region (wheat, sugar beet and vegetable crops). This would advance the understanding and management of nitrogen cycles in the crop-water-soil system.

Finally, the database included as part of this study may be integrated in a water and nitrate movement simulation model. This will allow the prediction of nitrate leaching losses and identify, through scenario analysis, the links that should be acted on, in order to reduce the risk of ground water nitrate pollution in this region.

FR :

Des recherches ont été menées dans la Sierra Madre Occidentale du Mexique afin de déterminer l'impact des modifications physiques du milieu (surpâturage et déboisement) sur le fonctionnement hydrodynamique des bassins. L'étude des régimes hydrologiques des deux principaux cours d'eau (le río Ramos avec un bassin de 7.100 km² et le Sextin de 4.700 km²) composant le haut Nazas montre une variation possible du régime hydrologique qui pourrait être liée aux dégradations de l'espace. Cette modification du régime n'est pas visible dans les tendances annuelles des pluies et des écoulements. De ce fait, d'autres indicateurs sont utilisés dans la recherche de tendances. En utilisant des algorithmes ordinaires et un modèle global simple (le modèle Nazas) nous testons et proposons ici trois indicateurs du régime hydrologique:

1. la comparaison entre les coefficients d'écoulement de base et de crue montre une possible diminution de l'écoulement de base;

2. Les temps de réponse des bassins montrent une probable diminution. Cela signifie que la transformation de la pluie en débit pourrait devenir de plus en plus rapide

3. le degré de participation de l'humidité préalable du milieu montre une possible diminution des temps de ressuyage du sol.

Si les résultats sont peu significatifs statistiquement dans cette zone, tous les indicateurs montrent une possible modification du régime hydrologique. L'enjeu est de taille ; la sierra approvisionne en eau plus du quart de la superficie du Mexique. Les modifications du régime hydrologique pourraient avoir des conséquences graves sur le développement.

EN :

Environmental impacts can alter the runoff regime of river basins. However, if the transformations are progressive, the hydrologic repercussions may not be always clearly visible. The influence that the physical parameters of the environment (soil surface features, vegetation, soil occupation, etc.) have on the hydrodynamic functioning of river basins is a current research concern in hydrological sciences. Generally these studies conclude that a decrease in vegetative cover leads to an increase in runoff, irrespective of the climate of the region.

For decades, the Western Sierra Madre has suffered progressive degradation by overgrazing and deforestation. This subtropical climate has a long dry season during the year (8 to 9 months), and the soils are generally thin (mostly Phaeozems and Lithosol). The vegetative cover of this zone is significantly degraded. Forests were reduced by more than 50 % from 1972 to 1992 as a result of deforestation. Overgrazing also has impacted the landscape, leaving stony slopes and crusted surfaces observable all over the mountain.

Due to both overgrazing and deforestation, increased runoff coefficients and erosion rates have been observed at the plot and micro-plot scale in this region. The response is different at the basin scale, due to the spatial variability of the hydrodynamic behaviour of the slopes. The aim of this paper is to determine if over-exploitation of the terrestrial system could have hydrological consequences. New methods and indexes are proposed, which are able to highlight the hydrological responses of the basins. Examples are given for two watersheds of the Upper Nazas basin.

An examination of the two main rivers of the High Nazas (the Rio Ramos with a basin of 7100 km2 and the Sextin of 4700 km2) shows a likely variation of the hydrologic regime, which could be linked to spatial degradation. This modification of the regime is not visible in the annual trends of rain and runoff. As a result, ordinary algorithms and a simple global model (the Nazas model) were used to describe the annual trends. Three indicators of hydrologic regime are proposed here :

1. A comparison between the base-flow and flood-flow based on the algorithm of Gustard et al. (1989), which allows the separation of flood flow from base flow;

2. The lag time (or response time) of watersheds, the time necessary for the rain to cause an increase in river flow;

3. Changes in the water content of soil. Infiltration and evapo-transpiration rates are considered to be the main elements of water balance and it is generally assumed that they increase with increasing vegetation cover. This allows a greater volume of water to stay in the soil. When there is no soil water content assessment, the Antecedent Precipitation Index (API) can be used to compensate. This takes into account the volume of the previous rainfall events, the time passed since this event, and a parameter Alpha of time water depletion.

Changes in land use were evaluated by remote sensing. The data collected were used in a calibration field campaign, to verify the hypothesis that there is a strong increase in overexploitation of the environment.

The results of the hydrological analysis were not always statistically significant in this zone. However, the changes are concomitant and all of the chosen indicators show a modification of the hydrologic regime. The analysis of 30 years of data shows that there is an evolution of the hydrologic behaviour of the watersheds. The following observations were made :

- A reduction of the base-flow was noticeable during the years 1970-1979 to 1990-1998. It appears that the floods have more and more importance in the hydrographs of the watersheds.

- A reduction in response time, which indicates that the transformation of the rainfall into runoff has become more and more rapid.

- The contribution of previous humidity in the environment in the formation of runoff also supports the earlier observations. It would seem that the drying time of the soil decreased. It appears to be a consequence of a reduction in soil water content, due to erosion and the degradation of vegetation.

The Western Sierra Madre supplies water to more than a quarter of the total area of Mexico. Changes in the hydrologic regime could have serious consequences on development in this region. The southern part of the country is mountainous, and 54% of the fields suitable for cultivation are located in the north of Mexico, but this region receives only 7% of the total water available in the whole country. The purpose of this research was to illustrate the importance of forests in the maintenance of the sustainability in the great "Laguna" irrigation district.

FR :

La détermination du débit de crue d'une période de retour donnée nécessite l'estimation de la distribution des crues annuelles. L'utilisation des distributions non paramétriques - comme alternative aux lois statistiques - est examinée dans cet ouvrage. Le principal défi dans l'estimation par la méthode des noyaux réside dans le calcul du paramètre qui détermine le degré de lissage de la densité non paramétrique. Nous avons comparé plusieurs méthodes et avons retenu la méthode plug-in et la méthode des moindres carrés avec validation croisée comme les plus prometteuses.

Plusieurs conclusions intéressantes ont été tirées de cette étude. Entre autres, pour l'estimation des quantiles de crue, il semble préférable de considérer des estimateurs basés directement sur la fonction de distribution plutôt que sur la fonction de densité. Une comparaison de la méthode plug-in à l'ajustement de trois lois statistiques a permis de conclure que la méthode des noyaux représente une alternative intéressante aux méthodes paramétriques traditionnelles.

EN :

Traditional flood frequency analysis involves the fitting of a statistical distribution to observed annual peak flows. The choice of statistical distribution is crucial, since it can have significant impact on design flow estimates. Unfortunately, it is often difficult to determine in an objective way which distribution is the most appropriate.

To avoid the inherent arbitrariness associated with the choice of distribution in parametric frequency analysis, one can employ a method based on nonparametric density estimation. Although potentially subject to larger standard error of quantile estimates, the use of nonparametric densities eliminates the need for selecting a particular distribution and the potential bias associated with a wrong choice.

The kernel method is a conceptually simple approach, similar in nature to a smoothed histogram. The critical parameter in kernel estimation is the smoothing parameter that determines the degree of smoothing. Methods for estimating the smoothing parameter have already been compared in a number of statistical papers. The novelty of our work is the particular emphasis on quantile estimation, in particular the estimation of quantiles outside the range of observed data. The flood estimation problem is unique in this sense and has been the motivating factor for this study.

Seven methods for estimating the smoothing parameter are compared in the paper. All methods are based on some goodness-of-fit measures. More specifically, we considered the least-squares cross-validation method, the maximum likelihood cross-validation method, Adamowski's (1985) method, a plug-in method developed by Altman and Leger (1995) and modified by the authors (Faucher et al., 2001), Breiman's goodness-of-fit criterion method (Breiman, 1977), the variable-kernel maximum likelihood method, and the variable-kernel least-squares cross-validation method.

The estimation methods can be classified according to whether they are based on fixed or variable kernels, and whether they are based on the goodness-of-fit of the density function or cumulative distribution function.

The quality of the different estimation methods was explored in a Monte Carlo study. Hundred (100) samples of sizes 10, 20, 50, and 100 were simulated from an LP3 distribution. The nonparametric estimation methods were then applied to each of the simulated samples, and quantiles with return period 10, 20, 50, 100, 200, and 1000 were estimated. Bias and root-mean square error of quantile estimates were the key figures used to compare methods. The results of the study can be summarized as follows :

1. Comparison of kernels. The literature reports that the kernel choice is relatively unimportant compared to the choice of the smoothing parameter. To determine whether this assertion also holds in the case of the estimation of large quantiles outside the range of data, we compared six different kernel candidates. We found no major differences between the biweight, the Normal, the Epanechnikov, and the EV1 kernels. However, the rectangular and the Cauchy kernel should be avoided.

2. Comparison of sample size. The quality of estimates, whether parametric or nonparametric, deteriorates as sample size decreases. To examine the degree of sensitivity to sample size, we compared estimates of the 200-year event obtained by assuming a GEV distribution and a nonparametric density estimated by maximum likelihood cross-validation. The main conclusion is that the root mean square error for the parametric model (GEV) is more sensitive to sample size than the nonparametric model.

3. Comparison of estimators of the smoothing parameter. Among the methods considered in the study, the plug-in method, developed by Altman and Leger (1995) and modified by the authors (Faucher et al. 2001), turned out to perform the best along with the least-squares cross-validation method which had a similar performance. Adamowski's method had to be excluded, because it consistently failed to converge. The methods based on variable kernels generally did not perform as well as the fixed kernel methods.

4. Comparison of density-based and cumulative distribution-based methods. The only cumulative distribution-based method considered in the comparison study was the plug-in method. Adamowski's method is also based on the cumulative distribution function, but was rejected for the reasons mentioned above. Although the plug-in method did well in the comparison, it is not clear whether this can be attributed to the fact that it is based on estimation of the cumulative distribution function. However, one could hypothesize that when the objective is to estimate quantiles, a method that emphasizes the cumulative distribution function rather than the density should have certain advantages.

5. Comparison of parametric and nonparametric methods. Nonparametric methods were compared with conventional parametric methods. The LP3, the 2-parameter lognormal, and the GEV distributions were used to fit the simulated samples. It was found that nonparametric methods perform quite similarly to the parametric methods. This is a significant result, because data were generated from an LP3 distribution so one would intuitively expect the LP3 model to be superior which however was not the case. In actual applications, flood distributions are often irregular and in such cases nonparametric methods would likely be superior to parametric methods.

FR :

Dans la majorité des pays, il existe des lois strictes pour réglementer la qualité de l'eau provenant des systèmes de traitement d'eaux usées. Ces spécifications légales sont essentiellement influencées par des questions telles que la santé publique, l'environnement et les facteurs économiques.

Les objectifs fondamentaux des procédés de traitement des eaux usées sont d'atteindre, avec l'utilisation d'énergie et à des coûts opérationnels minimaux, une concentration de matière biodégradable et de nutriments suffisamment basse dans les effluents et une production minimale de boues.

Les systèmes de traitement des eaux usées sont de grandes dimensions et complexes. Ils sont aussi sujets à des variations importantes dans le flux d'entrée et dans la composition de l'eau à l'entrée, qui ne sont pas bien connues. Le procédé est multivariable, avec beaucoup de couplages croisés et nonlinéarités importantes. La dynamique dépend de la variabilité des flux d'entrée et de la complexité des phénomènes physico-chimiques et biochimiques. Le comportement dynamique démontre une énorme variation de temps de réponse (de quelques minutes jusqu'à plusieurs jours). Ces problèmes, combinés aux objectifs les plus importants du traitement des eaux usées, donnent lieu à une demande de techniques de commande avancées, qui peuvent conduire à une réduction du volume à traiter, une diminution importante dans l'utilisation des produits chimiques, et une possibilité d'économie d'énergie et une diminution des coûts d'opération.

Dans ce travail, un " benchmark " (modèle de référence) d'un système complet de traitement des eaux usées a été développé, pour évaluer, à partir de simulations, la performance des différentes stratégies de commande proposées, y compris les techniques de respirométrie ("respirometry"). Ce travail s'aparente au Programme Européen d'Action COST (COST 624), et au projet "Respirometry in Control of the Activated Sludge Process (IWA Respirometry Task Group").

Le "Benchmark" représente un procédé de prédénitrification de la boue activée pour éliminer la matière organique et l'azote des effluents domestiques. Le simulateur est basé sur des modèles largement acceptés par la communauté internationale et il a été implanté dans un environnement Matlab/Simulink. La topologie du système et le développement complet du simulateur sont présentés dans ce travail. L'effet des conditions initiales et des caractéristiques du flux d'entrée (valeurs moyennes) sont analysés, aussi bien qu'un test en boucle ouverte. Les stratégies suivantes ont été sélectionnées en guise d'illustration de l'application de la commande automatique dans le "benchmark" (seulement avec commande proportionnel-intégral monovariable): commande basée sur la concentration d'oxygène dissous ("DO concentration-based control"), commande par respirométrie (commande par biomasse active et commande par taux de respiration bactérienne), et commande par concentration de nitrate (commande par dosage externe de carbone et recyclage du flux interne). Le "benchmark" est continuellement mis a jour et sa prochaine version va incorporer des fonctions d'optimisation en temps réel (on line) pour le procédé.

EN :

This paper presents a benchmark for a plant for biological treatment of wastewater, in order to evaluate, through simulations, different strategies of control. The benchmark represents the process of activated sludge in a configuration with pre-denitrification, for the treatment of domestic effluents, by removal of organic matter and nitrogen. It is based on models widely accepted by the international community and it is implemented in the Matlab/Simulink environment. The techniques of control here presented are typically used to control the process, such as: dissolved oxygen (DO) concentration-based control, respirometry-sbased control and nitrate concentration-based control, employing the classical proportional-integral (PI) controller.

FR :

Malgré la présence de nombreux barrages au Québec, peu d'études ont été consacrées à l'analyse des impacts de ces ouvrages sur les régimes hydrologiques des cours d'eau. La présente note a pour but d'analyser les impacts d'un barrage sur le régime hydrologique de la rivière Matawin en comparant le régime hydrologique de la rivière en amont (1390 km²) et en aval (4070 km²) du barrage pendant une période de 60 ans (1930-1990) et sur trois échelles temporelles distinctes. A l'échelle interannuelle, l'influence du barrage se manifeste par une persistance plus marquée des effets des épisodes humides ou secs ainsi qu'une hausse ou une baisse des débits moyens annuels respectivement durant ces périodes. Mais cette succession et cette persistance n'ont pas affecté significativement la stationnarité de la série hydrologique. Aux échelles mensuelles et saisonnière, l'influence du barrage se manifeste par une inversion du régime hydrologique caractérisée par une hausse des débits hivernaux et une baisse des débits printaniers. Cette influence se traduit aussi par une baisse significative du débit maximum mensuel mais une hausse du débit minimum mensuel. Il en résulte une diminution du coefficient d'immodération. A l'échelle journalière, le barrage modifie la période d'occurrence des débits extrêmes minimums et maximums. Il provoque la diminution significative des débits extrêmes minimums et maximums. Mais l'écrêtement des crues est modéré pour les débits de récurrence=10 ans. L'impact le plus significatif du barrage de Matawin est sans nul doute l'inversion du régime hydrologique dont les conséquences morphologiques et biologiques ne sont pas encore documentées dans la littérature scientifique canadienne. Cette inversion résulte du faible écoulement hivernal et d'une forte production de l'énergie électrique pendant la saison froide.

EN :

Few studies have characterized the effect of dams on the hydrologic regime of rivers in Quebec. This is rather strange given the large number of hydroelectric dams that have been constructed in the province. To shed some light on the environmental impact of these dams, this paper aims at describing and quantifying the effect of the Matawin River dam on the hydrologic regime of the river on an annual, seasonal and daily basis. The Matawin River is located north of the St-Lawrence River and is a tributary of the St-Maurice River. The Matawin dam was built in 1929 by Shawinigan Water and Power Co. mainly to supply the Gabelle hydroelectric dam on the Saint-Maurice River. The dam is 26 m high and the storage capacity of the reservoir is 348,000,000 m³ when full. The catchment area of the dam is 4070 km².

To assess the effect of the dam, we used various statistical methods to compare discharge time series over 60 years as measured at two gauging stations on the river. One of the stations is located upstream whereas the other one is located downstream from the Matawin dam. The upstream drainage basin covers an area of 1390 km2. No major tributaries are found between the two gauging stations, thus allowing us to ascertain the effect of the Matawin dam on the natural hydrologic regime of the river at different time scales.

On the annual scale, no difference in the mean annual discharge is observed upstream and downstream from the dam. The specific discharges upstream and downstream from the river's dam are respectively 17.2 and 17.1 l/s/km2. This is to be expected because the reservoir is used neither for irrigation nor for derivation. However, analysis of the interannual variability of mean annual discharges, using the Hanning low pass filter, reveals that wet and dry periods are far more persistent downstream than upstream from the dam. This persistence can be seen to occur within the two dry periods of 1930-1960 and of 1980-1990 and during the wet period from 1965 to 1980. This persistence does not affect the stationarity of the discharge time series downstream from the dam as no significant changes are detected from Mann-Kendall and Pettitt statistical tests.

On the monthly and seasonal scale, the comparison of the time of occurrence of maximum and minimum discharges shows a strong inversion within the hydrologic regime. Upstream from the dam, the maximum and minimum discharges are measured respectively during the spring and the winter. Downstream from the dam, the regime is inverted, with the maximum and minimum discharges being measured in winter and spring respectively. This inversion is closely associated with the production of hydroelectricity during the cold winter season when large amounts of water are released from the reservoir. Furthermore, is worth noticing that the monthly and seasonal coefficients of maximum discharge are lowered downstream from the dam whereas those for the minimum discharge remain similar.

On a daily basis, the comparison of dates of occurrence for the lowest annual discharge downstream and upstream from the dam shows these are found at different times of the year. Upstream from the dam, most of the minimum daily discharges are measured in August and September whereas downstream from the dam, they largely occur during April. On the other hand, the maximum daily discharges are recorded almost exclusively in April and May upstream from the dam but can occur throughout the year downstream from it, with a marginally larger number in January.

These results are relevant for the assessment of the environmental impacts of dams on rivers in the province of Quebec. For example, the inversion of maximum and minimum discharges is likely to have an important impact on the winter habitat characteristics by increasing the area of suitable habitat, but also by increasing the likelihood of sediment being transported during periods where usually only sporadic transport events occur. In the future, it would be crucial to understand the exact effect of the inversion on the morphological and biological components of the river dynamics.