Influence of the water table initial depth when pumping an unconfined aquifer

Robert P. Chapuis

In the proceedings of: GeoRegina 2014: 67th Canadian Geotechnical Conference

Session: Hydrogeology and Groundwater

ABSTRACT: of the water table initial depth when pumping an unconfined aquifer Robert P. Chapuis1 1 Department of Civil, Geological and Mining Engineering, École Polytechnique de Montréal, Montréal ed by the initial osition of the water table. This is a key parameter, which is not explicitly considered in existing theories. itiale de surface libre. C'est un paramètre clé qui n'est pas considéré de façon explicite dans les théories existantes. INTRODUCTION te. However, steady-state was reased-form solutions, a realistic numerical anasis was necessary. The main numerical results are presented. ABSTRACT This study was triggered by two old pumping tests with the same pumping well, fully penetrating an unconfined aquifer of uniform sand. The first pumping test had not reached steady-state after three days. The second pumping test, at the same rate, was planned to last one week or more, but reached steady-state after one day. In addition, the two tests provided different drawdown curves, and thus different values for the hydraulic parameters. Such results looked strange. However, they may be explained by a long list of differences between how the problem is handled by theories and how it should be. A finite element method solving equations for saturated and unsaturated seepage was used recently to try to clarify the old results. The differences between the drawdown curves of the two tests were fully explainp RÉSUMÉ Cette étude a été déclenchée par les résultats de deux essais de pompage dans le même puits totalement pénétrant une nappe libre de sable uniforme. Le premier essai n'avait pas atteint le régime permanent après trois jours. Le second essai, au même débit, était prévu pour durer une semaine ou plus, mais atteignit le régime permanent en un jour. De plus, les deux essais ont donné des courbes de rabattement différentes, et donc des valeurs différentes pour les paramètres hydrauliques. Ces résultats semblaient bizarres. Cependant, on peut les expliquer par une longue liste de différences entre la façon dont le problème est traité par les théories et comment il devrait l'être. Une méthode d'éléments finis, pour l'écoulement saturé et non saturé, a été utilisée récemment pour tenter de comprendre les vieux résultats. Les différences entre les courbes de rabattement des deux essais ont été expliquées par la position inla 1 Over twenty five years ago, the author was asked to try to explain the strange results of two pumping tests using the same pumping well of radius rw = 0.11 m, fully screened in an unconfined aquifer of uniform sand underlain by non-fissured clay. The sand aquifer is horizontal and about 7 m thick. Two 5-cm monitoring wells were located at radial distances, r, of 4.96 m and 15.5 m from the pumping well. They had 1.5-m long screens at the bottom of the sand layer. The pumping well was located between two parallel ditches, 1.3-1.5 m deep, draining the flat area (Fig. 1). The ditch water levels were controlled by gates. The first pumping test did not reach steady-state after 72 hours, and therefore the long-term drawdown was unknown. Therefore, a longer second pumping test was planned to last until it would reach steady-stached after one day. After briefly presenting the theories for transient and steady-state conditions, the usual analysis of drawdown data is presented for the two tests. The two sets of data yielded different and questionable values for the hydraulic parameters. In addition, the transient and steady state interpretations provided different values for the radial (horizontal) hydraulic conductivity, kr. To try to understand the reasons for such strange results, a list of differences was established between how the problem is handled by existing theories and how it should be. Since the real problem is physically described by highly non linear differential equations that cannot yield clolyditch 1.5 m deepditch 1.5 m deepPWMW1MW2almost flat areax (m)-60-50-40-30-20-100102030405060y (m)0102030405060 Figure 1: Pumping and monitoring wells in a flat area between two parallel ditches. ed to interpret the doWs) for either ansient or steady-state conditions. 2 CONSTANT RATE PUMPING TESTS, THEORIES Several theories have been proposdrawwn data of monitoring wells (Mt

RÉSUMÉ: ence of the water table initial depth when pumping an unconfined aquifer Robert P. Chapuis1 1 Department of Civil, Geological and Mining Engineering, École Polytechnique de Montréal, Montréal ed by the initial osition of the water table. This is a key parameter, which is not explicitly considered in existing theories. itiale de surface libre. C'est un paramètre clé qui n'est pas considéré de façon explicite dans les théories existantes. INTRODUCTION te. However, steady-state was reased-form solutions, a realistic numerical anasis was necessary. The main numerical results are presented. ABSTRACT This study was triggered by two old pumping tests with the same pumping well, fully penetrating an unconfined aquifer of uniform sand. The first pumping test had not reached steady-state after three days. The second pumping test, at the same rate, was planned to last one week or more, but reached steady-state after one day. In addition, the two tests provided different drawdown curves, and thus different values for the hydraulic parameters. Such results looked strange.

Access this article:
Canadian Geotechnical Society members can access to this article, along with all other Canadian Geotechnical Conference proceedings, in the Member Area. Conference proceedings are also available in many libraries.

Cite this article:
Robert P. Chapuis (2014) Influence of the water table initial depth when pumping an unconfined aquifer in GEO2014. Ottawa, Ontario: Canadian Geotechnical Society.

@article{GeoRegina14Paper273,author = Robert P. Chapuis,title = Influence of the water table initial depth when pumping an unconfined aquifer,year = 2014}