ABSTRACT: Artesian pressure was observed in the contact between two basaltic flows in the foundation of the Taquaruçu dam-Brazil, at depths of 4 to 5 meters. A series of inflow and outflow water tests were performed with small heads and compared with conventional water loss tests. Differences in the flow regimens were quite evident showing the importance of working at low heads to identify limit head for laminar flow.
RESUME: Dans la discontinuite A/B à la profondeur de 4 a 5 m, dans les roches basaltiques des fondations du barrage de Taquaruçu (Sao Paulo, Brèsil), on a observe d'ecoulement artesiène. Plusieurs essays ont ete realizes dans cette discontinuite. Ces essays on ete du type d'injection et de drainage. Les resultats ont ete mise en rapport aux essays d'eau conventionales (type Lugeon). Les differences ont montree l'importance d'utilization des basses pressions des essays pour identifier le limite d'ecoulement laminaire.
ZUSAMMENFASSUNG: "Artesianism" ist im kontakt A/B (4 bis 5m Tief) der basaltiches unterguind des Taquarucu Dammes abgeschlossen. Injektioneversuche und draiñageversuche wurden gemachen und die ergebnissen mit lugeonversuche vergleicht. Es sind groβe unterschieden in den, strom festgeschtellt, was verlaugt die benutzung von versuchen unter kleinen drucken.
1. INTRODUCTION The Taquarucu dam of Companhia Energetica do de Sao Paulo (CESP), under construction, situated at Paranapanema river, West region Paulo-Brazil. The dam is 30 m high and has a length of 2 000 m (Figure 1). The flow through the rock foundation at the concrete structures is mainly affected by the contact of basaltic flows A and B. This contact is continuous with a large oppening (milimetric to centimetric, with an average value of 0,50 cm) practically without any fill material and constitutes a preferencial horizontal flow feature. The contact is 4 to 5 meters below the surface within the foundation level of the Spillway and Intake Structures of the dam. During the excavations the initial flow into the area was about 12 000 l/min. dropping to 6 000 l/min after stabilization. This very high discharge has caused some problems for the execution of concrete structures. During the excavations and at different drainage conditions the A/B contact was tested with special attention and details. Multiple stage tests both of inflow (infiltration) and outflow (drainage). associated with conventional water loss tests were performed. The multiple stage with low heads made possible an accurate analysis of the flow characteristics at the basaltic feature. Plots of flow (Q) versus effective pressure or head (Hο) and plots of Q/Hο allowed a much better interpretation of the flow characteristics and of the transfer of laminar to transitional and turbulent flow. A good estimative of both critical head (Hcr) and corresponding critical flow (Qcr) that characterizes this change of flow could be done. These techniques have shown a great advantage in the prediction of flows into the excavation area that is normally under estimated when test results of conventional water loss tests area used.
2. FIELD TESTS Three types of tests were run in the contact A/B. Conventional water loss tests, using however a larger number of pressure; inflow tests or infiltration (injection of water with very low pressures) and outflow tests or drainage also with low pressures. The conventional water loss tests followed known techniques similar to Lugeon test (ABGE, 1975). Infiltration and drainage tests were performed by increasing the pressure test by means of longer casing (above the top of the rock) to test the artesianism flow or by reducing the pressure test using shorter tubes (below the artesian head). Figure 3 shows a sketch for this kind of test. The discharge Q, for each head was carefully recorded. The duration of each test was about 10 minutes for each pressure used. The test was repeated three times or more until the flow became approximatelly constant. Four drainage and nine infiltration tests were executed in differents points of the contact. Conventional water loss tests were also run in the exploratory phase of the project. Multiple stages tests were performed according to Rissler (1978) and also reported by Cruz and Quadros (1983) in connection with Nova Avanhandava dam in Brazil.