Abstracts
Résumé
Sous le climat à faible pluviosité du Tchad, les altérites sont dénoyées et seul le socle fracturé est aquifère. Le taux d'échec des forages atteint 60% car les fractures ont une répartition très discontinue comme le montre leur organisation fractale. Cela entraîne la coexistence de secteurs productifs et stériles. A l'échelle kilométrique, on peut ajuster le nombre de sondages de reconnaissance par village en fonction des caractéristiques climatiques, topographiques et géologiques. On définit ainsi des zones de productivité forte, où le taux de succès atteint 79%, moyenne et faible. On peut alors affecter chaque village d'un " potentiel d'investigation " qui est le produit du nombre de sondages par leur profondeur prévisionnelle.
A l'échelle locale, une analyse en composantes principales des paramètres de forage montre que la présence d'eau souterraine est liée aux caractéristiques du socle fracturé et non altéré. Une analyse discriminante fournit une " équation de productivité " qui permet de prévoir 90% des résultats en cours de foration: dès que le forage a traversé une dizaine de mètres de socle non altéré, elle permet de définir une profondeur limite d'investigation dépendant des caractéristiques intrinsèques de chaque site. Elle est surtout applicable dans les zones les moins productives où l'on observe systématiquement un surcreusement inutile des forages négatifs.
On dispose ainsi d'une stratégie de prospection alliant le nombre et la profondeur des forages. Elle permet de limiter la profondeur des forages implantés sur des sites peu productifs et de reporter le métré ainsi récupéré sur des sites plus prometteurs.
Mots-clés:
- Hydrogéologie,
- socle,
- granite,
- forage,
- taux de succès,
- analyse statistique,
- Afrique
Abstract
The aim of this study is to define a new strategy for groundwater prospection in sahelian basement aquifers. At present, the number and the depth of boreholes are fixed a priori in the project document: these parameters are the same for all the villages, regardless of their environmental context. In fact, during the drilling campaign, we systematically observe a useless overdrilling of negative boreholes that affects the cumulative drilled length of the project (Table 1). This is particularly important in granitic basement areas under the low rainfall sahelian climate: water is difficult to find because of low success rates, and the driller needs to ensure no groundwater indication appears a few meters under the fatal 60 m depth.
An illustration of this methodology is proposed for the Guéra, Ouaddaï and Biltine provinces of eastern Chad (Figure 1). This 150 000 km2 area is situated at the border of the Chadian basin from 10 to 15 ° north latitude at elevations of 400-700 m, with an annual rainfall between 200 and 600 mm. The geology is represented by precambrian granitoïds. Tectonics are well developed with many fractures, faults and photolineations from metric to multi-kilometric scales.
In Chad, weak recharge rates imply that the weathered rock reservoir is unsaturated and the aquifer is constituted by the fractured granitic basement. Thus, the overall success rate of 500 boreholes is only 42%. The unequal distribution of fractures leads to the presence of productive and barren adjacent areas with significantly different success rates. A statistical analysis of photolineations shows their fractal distribution with a dimension around 1.57, similar to the 1.59 dimension obtained in fractal fracture models (Figure 2). Fracturation is a main component of hydrogeological knowledge in basement areas and its variations between the villages can explain the different potentials of basement productivity: we must consequently adjust an "investigation potential" depending of the characteristics of each area. The proposed strategy of prospection determines the number of boreholes to drill and their specific depth.
At the kilometer scale, the total number of boreholes can be adjusted according to climatic, topographical and geological characteristics (Table 2). We show that only four parameters can explain a range of success rate from 0 to 79% in different villages (Table 3): altitude, average rainfall, petrography and fracturation intensity (measured in situ). Thus we can define the investigation potential which is the previous depth divided by the theoretical success rate of the area including the village. It is interesting to notice that the success rate in the high productivity class is similar to the average value obtained in more rainy basement countries of West Africa: for example 79% in south-west of Burkina Faso or 73% in Togo.
At the local scale, a principal components analysis on 12 drilling parameters was performed. It shows that the appearance of groundwater is mainly correlated to parameters describing the unweathered fractured rocks (Figure 3). A discriminant analysis was then performed on four of these parameters: thickness of unweathered drilled basement, depth of first water arrival, number of water arrivals and hammer velocity in the unweathered basement. This yields a "productivity equation" which allows one to anticipate 90% of the borehole results (Table 4). According to this equation, we can define a maximum investigation depth based on the geological characteristics of each borehole site.
The last section presents the complete strategy of groundwater basement prospection and two examples applied to Chad. For an average aquifer depth of 60 m, the investigation potential of each village depends on its productivity class: it varies theoretically from 10 boreholes (i.e. 600 m) in low productivity area to 1.3 boreholes (i.e. 76 m) in high productivity zones (Table 5). This potential must then be distributed among the different sites according to the result of their productivity equations.
The village of Getgéré is situated in a particularly unproductive zone (see Table 3) where about ten boreholes are statistically needed to obtain a positive result: its investigation potential is supposed to be 300 m. Four negative boreholes were drilled from 62 to 75 m with a total depth of 261 m. In fact, the productivity equation showed all these sites were unproductive from drilled depths of 28 to 40 m deep (Table 6): the same result could have been obtained with only 130 m drilled; 131 m were uselessly consumed. With this excess drilled length, we could have drilled eight extra shallower boreholes and increased the probability of success in obtaining a productive well.
The village of Eroua is situated in a productive area where the success rate is 79%: its investigation potential is 60 / 0.79=76 m. The first borehole was negative at depth of 74 m, but the productivity equation already indicated this result after only 38 m of drilling. At the second site, a positive borehole was obtained at 41 m depth where the equation foresaw 43 m. Finally, the cumulative drilled length was 115 m and the investigation strategy would have permitted the transfer of 34 m to another more promising site.
Keywords:
- Hydrogeology,
- basement,
- granite,
- borehole,
- success rate,
- statistical analysis,
- Africa